Recent developments in applications of MRI techniques for foods and agricultural produce—an overview

Abstract

For the detection of defects, evaluation of internal quality and analysis of internal structures in food and biological materials, there is a number of image acquisition techniques available so far. Among them, MRI has the advantage of allowing a variety of measurements to be made that not only contribute to the evaluation of maturity and quality parameters in fruits and vegetables and other food materials but also improve the understanding of underlying physiological processes. So, the ability of MRI to function in a completely non-destructive and to encode molecular dynamics through different contrast mechanism has encouraged developments of many applications in various fields. An overview of MRI as non-destructive detection in quality of food and agricultural produce and its application in postharvest sorting and processing have been presented in this paper. This paper elaborated principle of MRI, function of its components, quality detection methods, and discussed recent research and applications. Since, no paper yet published on MRI, covered so wide applications as included in this work, its importance being increased more.

Introduction

Methods to measure quality and quality-related attributes have been developed over centuries, with instrumentation appearing in the past eighty years. Most recently, the emphasis has been on developing sensors for real-time, non-destructive sorting (Abbott 1999). Since, in the food industry, some quality evaluation is still performed manually by trained inspectors, which is tedious, laborious, costly and inherently unreliable due to its subjective nature. Increased demands for objectivity, consistency and efficiency have necessitated the introduction of computer-based imaging techniques and recently, computer vision employed image processing techniques have been developed rapidly, which can quantitatively characterize complex size, shape, colour and texture properties of foods (Du and Sun 2004). However, a computer vision system is useful for surface inspections, in many specialized investigations, the food technologists and scientists frequently need to see an internal view of the sample (Abdullah 2008). Since, to the human observer, the internal structures and functions of the food and agricultural product’s body are not generally visible. So, various medical technologies have been applied in agriculture and its allied sectors to identify the defects, blemishes and abnormalities by the researchers. According to them, by these technologies, images can be created through which the food processor can look into the bodies and guide during quality evaluation and inspection operation. One of the most powerful, non-destructive analytical techniques that enable studies of the molecular structure through measurement of the interaction of an oscillating radio-frequency electromagnetic field with sample immersed in a strong magnetic field is nuclear magnetic resonance (NMR).

The formation of images by NMR, known as magnetic resonance imaging (MRI) latter on, was first suggested by Damadian (1971), but the technique now used stems from work by Lauterbur (1973) who coined the term ‘zeugmatography’. Subsequently, images were produced by Mansfield et al. (1974) at Nottingham University. Although NMR had been used for spectroscopy for many years, following its discovery thirty years ago (Bloch 1946; Purcell et al. 1946), and initially it’s possible medical value was not realized until Damadian (1971) found that rat tumors in vitro had different NMR characteristics compared with normal tissue.

However, the development of alternative applications of MRI based on the principles of nuclear magnetic resonance (NMR) that has ability to work non-destructively and can be used to evaluate important quality attributes of food products non-invasively. So, slowly but powerful steps of advancement of this new non-invasive technology not only strengthened in medical imaging but invaded in other area of research also, such as agricultural sciences especially in food industries.

Furthermore, the magnetic resonance imaging is an imaging technique primarily used to obtain high quality images of the inside of the object in two or three dimensions. It has achieved general acceptance as a powerful tool for the diagnosis and assessment of clinical conditions following the widespread introduction of medical imaging systems during the 1980s. According to researchers, the magnetic resonance imaging is a window to the agricultural materials and foodstuff’s body that allows to survey internal condition of bodies. However, no image window reveals everything. Different imaging modalities reveal different characteristics of the biological body. With each method, the range of image quality and structure visibility can be considerable, depending on characteristics of the imaging equipment, skill of the operator, and compromises with factors such as radiation exposure and imaging time. In this review paper we have discussed principle of MRI and recent developments in the applications of MRI in food processing as well as in other area of agriculture. Since, no review paper has been published recently on the MRI and its application in Food. It provides a great opportunity to learn the basic principle of MRI and introductory knowledge of its application to the researchers. In this paper, we have reviewed its potential of application as well as different parts of MRI system.

Principle of MRI

The MRI phenomenon and its utilization in exploring the properties of food materials generally depend on the inherent magnetic properties of the certain atomic nuclei in a magnetic field. For instance, the hydrogen nucleus, a proton, behaves as a spinning charged particle; it possesses angular momentum and generates a polar field (Kauten and McCarthy 2005). Since, the structure of food ranges from pure liquids to solutions, gels, emulsions, foams, suspensions and solids with different content of water or from the chemical point of view; food is composed of water, carbohydrates, lipids and proteins. Different NMR techniques can be applied in food studies as a consequence of physical and chemical variety, including solution and solid state spectroscopy as well as NMR imaging (Jankowski 2003). However, the sensitivity of nucleus for investigation by an NMR spectroscopy depends on the magnitude of its magnetic moment, the natural abundance of nuclei and the strength of the applied magnetic field. Food molecules contain NMR-active nuclides, e.g. ¹H, ¹³C, ³¹P, ²³Na. Hydrogen ¹H has the highest receptivity at natural abundance, while carbon ¹³C the lowest (receptivity of 1.7 × 10⁻⁴ relative to hydrogen). In addition to this, Table 1 lists the spin and isotopic composition for several elements commonly found in biological systems. The ¹H nucleus, consisting of a single proton, is a natural choice for probing the body using MR techniques, for several reasons. It has a spin of 1/2 and is the most abundant isotope for hydrogen. Its response to an applied magnetic field is one of the largest found in nature. Since the body is composed of tissues that contain primarily water and fat, both of which contain hydrogen, a significant MR signal can be produced naturally by normal tissues (Brown and Semelka 2010a, b).

Table 1 Constants for selected nuclie of biological interest

Nuclear magnetic resonance (NMR) imaging gives density of protons or hydrogen nuclei of the body at resonant frequency and based on the association of each spatial region in a sample with a characteristic nuclear magnetic resonance frequency, by imposing an external magnetic field. Since, without the external magnetic field, the magnetic moment would point in all directions at random, and there would be no magnetization. However, in the presence of large magnetic field, hydrogen nuclei will preferentially align their spin in the direction of magnetic field. This is known as Larmor effect, the frequency at which nucleus process around the axis (Fig. 1) is termed the Larmor frequency (Abdullah 2008), defined by:

$$ \nu =\gamma {B_0} $$

(1)

where, γ is a constant (the magnetogyric ratio) whose value depends upon which nucleus is being investigated, and B₀ is the strength of the external magnetic field. In a 4.7 T (4.7 × l0⁴ G) imaging system, γ is 4,257 Hz Gauss⁻¹ and v is about 200 MHz for aqueous protons (Clark et al. 1997). The other commonly used various terms in magnetic resonance imaging techniques are given in Table 2.

Fig. 1

Precession of a spinning top. As a top spins on its axis it also rotates or precesses around the vertical axis. This may be compared with the precession of the magnetization associated with protons, which occurs at a radiofrequency around the Z axis (Henderson 1983)

Table 2 Commonly used MR terms

Larmor effect implies a transfer of energy from the spin system to another system of lattice and characterized by an exponential relaxation law with time constants T₁and T₂, which are also known as the spin–lattice excitation and spin–spin relaxation times, respectively (McCarthy 1994). In some instances, MRI research was conducted to measure moisture and its migration in food systems; other researches demonstrated the potential of using MRI to investigate physical or biological properties of food products; still some other work sought to estimate the yield.

The extension of standard NMR to imaging is accomplished by imposing a linear magnetic gradient on the external magnetic field. The resonance (Larmor) frequencies of protons thus vary according to their positions along the gradient axis as:

$$ \nu =\gamma \left( {{B_0}+r{G_r}} \right) $$

(2)

where r (cm) is the position along the axis of the gradient G_r (Gauss cm⁻¹). So, frequencies can be measured and from the frequencies the position of resonating nuclei can also be determined.

Moreover, being an analytical tool, MRI is based on the absorption and emission of energy in the radio frequency range of the electromagnetic spectrum. The generation of magnetic resonance images can be controlled by the radio frequency pulse sequence used for excitation of the nuclear spins. Selection of the optimal pulse sequence parameters can result in enhanced images, with good contrast between a region of interest and its surrounding area (McCarthy 1994).

Furthermore, according to researchers, by applying a radio-frequency (RF) field at the resonant frequency, the magnetic moments of the spinning nuclei lose equilibrium and hence radiate a signal which is a function of the line integral of the magnetic resonance signature in object. This radiation reflects the distribution of frequencies, and a Fourier transform of these signals provides an image of the spatial distribution of the magnetization (Rinck 2001).

Principle components of MRI system

A MRI system consists of several components that perform different functions. Figure 2a is a generic illustration of an MRI system. The essential components of an MR imaging system include: (1) a large magnet which generates a uniform magnetic field, (2) smaller electromagnetic coils to generate magnetic field gradients for imaging, and (3) a radio transmitter and receiver and its associated transmitting and receiving antennae or coils. In addition to these fundamental components, a computer is necessary to coordinate signal generation and acquisition and image formation and display. Figure 2b shows essential components of an MR imaging system.

Fig. 2

a Block diagram of a typical MRI system (Abdullah 2008). b Block diagram of an MRI system (Jin 1989)

Main magnets

The magnet dominates the MRI system. It is the biggest, most expensive, and perhaps most demanding and controlling aspect of an MRI system. The ideal MRI magnet produces a perfectly homogeneous magnetic field. This can be accomplished theoretically by producing a spherically shaped surface current that varies with a sinusoidal spatial distribution (Jin 1998). The main magnet is needed to generate a strong magnetic field (B₀) that is uniform over the volume of interest. It is always desirable to have a strong magnetic field since it provides better SNR (signal to noise ratio) and better resolution in both frequency and spatial domains. When magnets are unable to provide the desired B₀ field uniformity for high-resolution spectroscopy, shim coils are employed to improve the field uniformity. Clinical imaging systems typically have field strengths no stronger than 3 T. However, some experimental systems reportedly have 4 T or 7 T main magnets.

In commercial MRI, the magnetic field ranges from 0.5 to 2.0 T (compared with Earth’s magnetic field of less than 60 μT). T₁ is typically of the order of 0.2–2 s, and T₂ ranges from 10 to 100 ms. According to Planck’s equation E = hf, for a field strength of 1.5 T, f corresponds to radiowaves with a frequency of 60 MHz (Abdullah 2008). However, most of the NMR research in food science is performed using high-resolution or solid state instrumentation (magnetic field of 5.9–21.1 T) (Brown and Semelka 2010a, b). According to them, for industrial applications, the low-cost and benchtop instruments are used, for instance, the MINISPEC series made by Bruker (0.4 T). An MRI technique, operating at 2 T, applies mainly the 31-cm bore magnets (Jankowski 2003). For instance, Barreiro et al. (1999) used magnet bore size of 40 cm, both with gradient stack in place, the bore was reduced to 26 cm and a home-built high pass birdcage coil (n = 8) with an inner diameter of 10.4 cm and a length of 14.0 cm was constructed to adapt the coil to the maximum size of apples during mealiness assessment.

Gradient coil

The gradient coils are used to encode spatially the positions of the nuclear spins in the MRI system by varying the value of the local magnetic field causing precessional frequencies to vary as a function of their positions (Jin 1989). A gradient coil set is made up of 3 coils that produce the x, y and z-gradients. The field produced by the gradient coil is required to be linear over the imaging volume. Gradient coils are turned on and off rapidly for imaging applications. Some desirable performance criteria for gradient coils are: low inductance (short switching time), high current to gradient ratio (high current efficiency), minimal interaction with any other equipment, which would otherwise result in eddy currents (minimal resistance), low power consumption, and good gradient field uniformity (gradient linearity over a large volume). Over the last two decades new methods of gradient coil design have been developed, and a combination of these methods can be a mixture of them trying to avoid bad effect on tissues of biological material and discomforts to patients in biomedical. The Helmholtz, Maxwell and Golay coils pair are commonly used in MRI systems (Table 3).

Table 3 Commonly used gradient coils in MRI systems

RF transmit and RF receive coils

The RF coils in MRI that excite the magnetization are referred to as transmit coils. Separate receiver coils receive the signals from the relaxing nuclei spins. A single coil can be used to transmit and receive signals, but separate coils are often employed to improve the SNR. The use of separate coils depends on properties that are desirable for each mode of operation. Transmit coils are required to produce a uniform B1 field over the region of interest, while receive coils are required to insure uniformity of the received signal over the volume being imaged. But ongoing developments in the RF subsystem are probably the most profound changes in MRI system development in the last several years. In the traditional block diagram of an MRI system, the large body coil supplied the relatively uniform perturbing RF pulse and another smaller volume or surface coil detected the time varying relaxation signature, spatially encoded by the gradient subsystem. It was a significant advance when the concept of multiple receiving elements, the phased array coil, was carefully studied (Roemer et al. 1990) and implemented commercially. Shortly before Roemer et al. (1990), another paper by Carlson (1987), demonstrated that it was possible to use the spatial response characteristic of coil as another spatial encoding mechanism. Ten years later parallel imaging i.e. Simultaneous Acquisition of Spatial Harmonics (Sodickson and Manning 1997) and sensitivity encoding (Pruessman et al. 1999) were developed. These two methods proposed to accelerate image acquisition by reducing the number of phase encodes needed and use the coil spatial response characteristics to synthesize the missing k-space lines (SMASH: Simultaneous Acquisition of Spatial Harmonics) or remove the aliasing ambiguity (SENSE: Sensitivity Encoding) in the image domain.

For the better results, the scanner MRI system has bore diameter of a few tens of centimeters, superconducting coil used to generate static magnetic field, and RF coils generally used to transmit radio-frequency excitation into the material to be imaged which excites a component of magnetization in the transverse plane which can be detected by a RF reception coil. The signals are transduced and conditioned prior to image reconstruction. Consequently, MRI scanners generate images with sub-millimeter resolution of virtual slices through the sample and the thickness of the slices is also of the order of a millimeter. In addition to this contrast resolution between materials depends strongly on the strength of the magnetization, T₁, T₂, and movement of the nuclei during imaging sequences. The most striking artifacts appear when the magnetic field is disturbed by ferromagnetic objects. Other artifacts, such as ringing, are due to the image reconstruction algorithm and sensor dynamics. A basic block diagram of a typical MRI data-acquisition system is shown in Fig. 3a and b.

Fig. 3

Image of water-core apple. a Photograph prior to MRI experiments b MR images along the length of the sample (Marigheto and Hills 2005)

Assessments of agricultural produces

Owing to the fact that MRI provides rapid, direct, and, most importantly, non-invasive, non-destructive means for the determination of not only the quantity of water present but also the structure dynamic characteristic of water, this relatively new imaging technique has become useful for food engineering (Abdullah 2008; Hills 1995). There are numerous applications of MRI, since water is the basic building block of many food materials.

Fruits and vegetables

Much effort nowadays is directed towards increasing the consumption of fruits and vegetables because of their health benefits. The attractiveness of horticultural crops with to respect the purchasing behaviour of consumers is affected by visual appearance, but the expected internal quality is of equal importance. The NMR imaging technology can provide high resolution images of internal structure of intact fruits and vegetables and is suitable for detecting internal defects and various quality factors such as bruises, dry regions, warm damage, stage of maturity/ripening, and presence of voids, seeds and pits (Chen et al. 1989 ) Since, MRI is considerably faster than conventional histology on the basis of the type of employed pulsed sequence that can afford a superior realization of the sample’s of 3-D characteristics, and is able to distinguish tissues with dissimilar relaxation properties, in contrast, the use of conventional histological technologies for structural studies of soft tissues fruits (Williamson et al. 1992). However, there are some difficulties and associated artifacts. Since, this is not suitable for detecting pit fragments of pitted fruits and split pits in whole fruits and the variation of experimental parameters, such as echo delays, resolution, thickness of scanning slices, can affect the enhancement of specific features of the specimen (Chen et al. 1989 ).

Furthermore, since, fruits and vegetables are capable of absorbing radiofrequency (RF) energy and being stimulated in a (¹H) MRI experiment due to presence of a multitude of proton-containing species. The H atoms associated with water molecules selectively targeted for imaging in biological material because of their high abundance (Goodman et al. 1993; Pope et al. 1991). Goodman et al. (1993) and Pope et al. (1991) stated that this relative high abundance gives a large signal-to-noise ratio and consequently allows shorter imaging times that is an important consideration in the acquisition of MR images. However, where they are localized or present in high concentrations, it is possible to image the constituent H atoms in other metabolites such as lipids and sugars (Clark et al. 1997). So, to achieve satisfactory signal-to-noise ratios, compounds should be present in high concentrations or localized in specific organs or organelles. The applications of MRI techniques in assessments of some fruits and vegetables have been reviewed in this paper.

Fruits

Apples

A nuclear magnetic resonance (NMR) imaging method in radiology can be used to obtain images of several internal diseases and its distribution in apples and pineapples. The uneven distribution of mobile water and its NMR relaxation times, spin–lattice (T₁) and spin–spin (T₂) produce signals that are used to reconstruct cross-sectional images. From the variation of signal intensity in the images, detailed internal structures of apple fruit, including petal bundle, endocarp, outer limit of carpel, dorsel bundle of carpel, cortex of receptacle, pith of receptacle and seeds, can be discerned. Based on NMR images, the normal and watercored tissues can be clearly differentiated on the images. Primarily, the watercore occurred area was noted in ±20 mm from the center of the fruit by Wang et al. (1988) and the most affected area reported by them was between 5 and 10 mm from the center, toward the stem end. Since, watercore is a physiological disorder affecting apples in which intercellular spaces are filled with liquid. So, proton magnetic resonance imaging (¹H-MRI) was used by Clark et al. (1998) to investigate its postharvest amelioration in ‘Fuji’ apple (Malus domestica Borkh.) and noticed, curing leads to fruit of normal appearance and taste, and observed loss of watercore. The severity of the disorder and its distribution within the fruit can be quantified by digitizing affected and unaffected areas in each image and the condition can be ameliorated. The recent developed MRI techniques can monitored the curing process of watercore in Fuji apples that indicated in 7 to 15 weeks after harvest. The curing process, however, depends on the intensity and severity of the disorder in a particular growing season. The distribution and severity of watercore at individual locations within affected samples can also be determined by using multi-slice images of single fruit obtained by screening at various times during fruit development, since the high signal intensity associates with affected areas within the fruit. A (physical) slice can be taken from the most damaged region and can be used for MRI with the standard spin–echo imaging sequence. The change in proton density in the water-core (Fig. 3) affected regions gives significant image contrast that can be exploited in on-line detection of watercore (Marigheto and Hills 2005).

However, in nuclear magnetic resonance images of bruised apples, the bruised tissue regions appear brighter than intact tissue regions. McCarthy et al. (1995) imaged bruised apples using different pulse sequence techniques to analyze temporal changes in image contrast in bruised and unaffected areas of flesh. In their study they reported that with the time, contrast between bruised and non-bruised regions was found in increasing trend. By comparing spin-echo and gradient-recalled-echo images, they concluded that due to the difference in contrast between these two tissues, it could attribute to change in diamagnetic susceptibility. That is, fluid from ruptured cells displaces air in the intercellular space thereby reducing the total area of interfaces across which magnetic field gradients occur. The rate at which this process proceeded should have been dependent on the nature of the cellular structure in the tissue, and the extent to which damage had occurred. In apples, bruising can be identified as areas of water-soaked tissue, which have high signal contrast, while voids characterized by the absence of signal (Chen et al. 1989). Maturation processes may also result in an increase in free water, as well as other metabolites associated with quality, such as oil and simple carbohydrates. Free water and translucency in over-ripe pineapple flesh was clearly detected by MRI (Chen et al. 1989) implying a quality control role for commercial imaging systems.

Nuclear magnetic resonance (NMR) relaxometry and imaging have been evaluated as a technique to measure mealiness. Barreiro et al. (1999) used magnetic resonance techniques to assess mealiness in apples. Since, magnetic properties of some atomic nuclei is the basis of magnetic resonance imaging, when placed in a magnetic field, the natural magnetic dipoles of the nuclei reorient themselves along the magnetic field. After excitation they return to their equilibrium position. The rate at which this happens can be expressed by two relaxation times (T₁ and T₂) and is a function of the texture of the material (Smith and Lange 1998). Barreiro et al. 1999 found that the variability of the T₂ values not only inside an apple was larger than that between the apples but a difference between the average T₂ value of fresh apples and that of apples stored in mealiness-enhancing conditions was also noticed. In their study, they concluded that minimum T₂ values were shown to be significantly lower for mealy apples than for fresh apples, indicating that a more desegregated structure and a lower juiciness content lead to lower T₂ values. Furthermore, they also found that the T₂ images of mealy apples showed a regional variation of contrast which was not shown for non-mealy apples. The magnetic resonance images of mealy apples also showed a regional variation of contrast which was not the case for non-mealy apples (Barreiro et al. 2000). This variation of contrast was similar to the NMR images of apples with internal breakdown although the contrast was smaller. Figure 4 shows T₂ maps of the non mealy, mealy and internal breakdown fruits tested with MRI technique.

Fig. 4

T₂ maps and histograms corresponding to the a non mealy b mealy and c internal breakdown apple fruits (Barreiro et al. 1999)

Letala et al. (2003) obtained images by magnetic resonance imaging (MRI) of the apple varieties Idared, Redspur and Topaz during ripening and storage. They analysed images by texture analysis (TA) to determine the correlation between TA parameters and firmness, soluble solids content (SSC) and titratable acids, and they reported that all varieties differ significantly in SSC, acidity and the firmness of fruits. Chayaprasert and Stroshine (2005) used a low-cost, low-field (5.55 MHz) MR sensor to detect internal browning in apple. According to them it is possible to use the MR sensor and conveyor system for on-line sorting of apples with internal browning at conveyor speeds below 100 mm/s if precise control of the conveyor speed and apple position at the time of interrogation can be maintained. Good performance at faster conveyor speeds may be possible if improvements such as better pre-magnetization are implemented.

So, MRI is a nondestructive method, can asses quality of apples and use as a tool for measurement of spatial distribution of water in biological material and its physical properties. In addition to this, it can also be used to detect mechanical damage and internal diseases of apples. Suchanek and Olejniczak (2008) carried out experiments on a low field (0.088 T) MRI system. They studied temporal evolution of the internal browning process in apples by recording MRI images, using the spin echo (SE) sequence and obtained proton density and T₂ weighted images by varying the echo time (TE) and repetition time (TR). In order to shorten the acquisition time, fast imaging sequences, like FSE (Fast Spin Echo) and FLASH (Fast-Low-Angle-Shot), were also used by them and reported that these provided similar spatial information about the disease affected tissue of the fruit after comparing the obtained images.

Citrus

Citrus are one of the most important fruits in worldwide trade and a high percentage of production in countries such as United States, China, Spain and Mexico, is earmarked for export and therefore fruit quality is required to be carefully assessed. Regarding the commercial strategy, inspection needs to perform on a large amount of fruit with minimum time allocation. So, on-line internal quality measurement in the range between 2 and 10 fruits per second is typically a key issue for inspection (Barreiro et al. 2004). Nuclear magnetic resonance (NMR) is a versatile technology, which has emerged as a powerful analytical technique that probes the physical state of matter and provides both spatial and chemical information (Gambhir et al. 2004).

One of the main concerns for citrus producers is the presence of seeds within oranges and mandarins. For consumer demand, seeds are especially undesirable in mandarin. Concerning MRI detection of seeds in citrus, the acquisition of MR images where seeds are distinguishable from juice segments is the most promising procedure to be applied. Hernandez-Sanchez et al. (2006) inspected mandarins using magnetic resonance imaging (MRI) in order to detect the presence of seeds and reported that it is feasible to perform a straightforward export of models developed under static conditions towards on-line conditions. They found correct classification 88.9 % for seedless and 86.7 % seed-containing fruits under MRI stationary conditions while 92.5 % and 79.5 % for the seedless and seed-containing category, respectively, under MRI dynamic conditions. The performance under on-line conditions was evaluated by imaging the fruits while conveyed at 54 mm/s. In their study they noticed that the transverse relaxation time-weighted fast low angle shot images (703 ms acquisition time) were acquired to enhance contrast between seeds and pulp, effectively. The acquired axial MR fast images (780 and 390 ms acquisition time) achieved a reasonable contrast for seeds within citrus under static conditions (Hernandez et al. 2005). A large decrease in image quality was found in comparison to the static ones when conveying fruits at a rate of 54 mm/s. In orange also, Hernandez-Sanchez et al. (2006) used MRI to detect seeds. Wang et al. (2007) investigated with low-field NMR water content changes and moisture migration behavior of different ways of handling naval orange during storage and found that the navel orange corruption had a great relationship with the water migration. Thus, to obtain high-quality contrast between seeds and flesh, MRI is feasible technique, although time consuming (7 s) and so only available for off-line conditions at juicy stage. But, a compromise between spatial resolution, contrast and acquisition time has to be reached in order to enhance the interest of MRI application for on-line purposes Blasco et al. (2003).

For better results, two different types of fast MRI sequences were investigated: a gradient echo and a spiral–radial, with 484 ms acquisition time for the former compared to 240 ms for the latter by Barreiro et al. (2008) to acquire MR images of the internal structure of mandarins for non-destructive seed identification. They got most promising results after post-processing of images with two-dimension histogram variance segmentation technique and reported that the image features including perimeter, compactness, maximum distance to the gravity centre, and aspect ratio can be employed in a linear discriminant function, by which seed identification of mandarins could be achieved with 100 % accuracy using radial–spiral sequence and 98.7 % accuracy with gradient echo images.

Kiwifruits

During long-term storage or transport, water loss of kiwifruit is a serious problem as kiwi epicarp is very permeable to water mass transfer. The rate of water loss and the effects on the cell metabolism of postharvest fruits and vegetables have been reported by Hsiao (1973). With 0.5 % water loss, cell wall enzymes activity is already increased, and a further increase of water loss accelerates respiration and ethylene production, together with the loss of volatiles. In addition to water loss, very small injuries such as the ones produced by kiwi brushing for the market, trigger ethylene biosynthesis hastening the ripening process (Massantini et al. 1995). Also bruising, occurring during harvest and postharvest practices including the shipping, hastens kiwi fruits ripening increasing ethylene production and fruit temperature plays an important role in symptom appearance (Mencarelli et al. 1996). Thus kiwifruits require high care during harvest and postharvest handling and the application of magnetic resonance imaging (MRI) can be a useful tool to follow non-invasively slight structural modification occurring in kiwifruit during ripening which could trigger a stress ethylene response. Using MRI, Burdon and Clark (2001) investigated the water status in kiwifruits and noticed that the rate of water loss and changes in water parameters depended on the initial water status of the fruit; relaxation data, T₁ and T₂, highlighted large differences among core, inner and outer pericarp, but no significant changes were found in any MRI measurement during the study. Successively, Newman and Redgwell (2002), studying the nuclear spin relaxation experiments, showed that pectin retained in the cell wall became “softened” in the early stage of ripening, prior to solubilisation later in ripening. Recently, Taglienti et al. (2009) evaluated the effect of storage conditions on minimal structural changes of kiwi tissue using MRI and concluded that the MRI appeared a useful tool to identify minimal texture changes linked with water mobility. This minimal texture changes could be the reason of the rapid softening of kiwifruits during storage even at 0° C probably through a wound ethylene response. Moreover to the use of MRI as sensing tool for determination of internal morphology, several research have been carried out mainly concerning to the identification of the evolution of tissues during post-harvest ripening/storage and the evaluation of the overall quality by measuring quality related parameters. For instance, Vaientini et al. (2009) observed that the MR imaging is capable of distinguishing kiwifruits treated with PGR (plant growth regulator) from those non treated, as well as to discriminate the PGR, i.e. auxin and citoquinine.

Mango

MR and MRI can “see inside” a piece of fruit by picking up areas of increased water or voids in the internal tissue. These discrepancies in the internal tissue highlight defects such as bruising, chilling injury and insect damage, researchers found. For instance, damage from heat treatment can appear as “air filled cavities,” while increased water in the fruit from the center to the skin’s surface indicated gradual ripening, studies showed. In addition to this, MRI has also been used to evaluate damage and ripeness in fruits like apples and pineapple, but further work is necessary to determine how MRI technology could track maturation and ripening in mangos, researchers said (Slaughter 2009). According to Mazucco et al. (1993) areas of increased free (unbound) water or voids in the internal tissue of fruits are readily detectable by MRI and allow detection of internal defects such as bruising, chilling injury, and insect damage. For example, Joyce et al. (1993) demonstrated that MRI could be used as a nondestructive method to detect heat treatment induced injury to mesocarp tissue in ‘Kensington Pride’ mangoes because the injured areas contained air filled cavities in the tissue and the response of mango to damaging heat treatment includes impaired starch degradation and development of internal cavities within the mesocarp. Using MRI to monitor temporal changes over 11 days after heat-treatment, Joyce et al. (1993) observed these symptoms in mango. According to them, their images show that heat treatment-induced lesions started to develop on the day of treatment. By 4 days later, areas corresponding to damaged pulp had reached maximum size. Furthermore, Joyce et al. (2002) conducted another but limited study of MRI on four intact ‘Kensington Pride’ mangoes during ripening and reported that water activity in the flesh increased in an outward-moving flux as the fruit ripened, and they observed that the MRI signal in the middle part of the flesh increased during ripening. Their study was based on image data by MRI. Although MRI is being explored for many foods, yet no work on mango has been reported. This technique should be explored for dry matter content determination, internal bruises, fruit fly, soft tissue, and maturity assessment of mango fruits. Processes leading to quality defects, such as bruising, freezing and pathogen invasion, and changes occurring within damaged regions of tissue, have been reported for a number of crops. These regions subsequently collapsed as the fruit ripened.

Pear, peach and pomegranate

Researcher has already been shown that the MRI can be used as an effective technique for internal quality assessment in a wide variety of fruit species: apple, avocado, blueberry, cherimoya, courgette, cucumber, durian, kiwifruit, mandarin, mango, melon, nectarine, olive, onion, orange, papaya, pear, peach, pineapple, potato, tangarine, tomato, strawberry, melon, and watermelon (Hills and Clark 2003). Mealiness (woolliness in peaches) is a negative attribute of sensory texture that combines the sensation of a desegregated tissue with the sensation of lack of crispness and juiciness also known as mealiness in other fruits, has been identified fruit by fruit by instrumental means. Barreiro et al. (1998b) used magnetic resonance techniques to assess mealiness in peaches. The development of mealiness in peach is caused by different underlying physiological changes. It was confirmed the association between internal breakdown and mealiness in both types of fruits show the same histogram shape and all internal breakdown fruits were identified as mealy through confined compression test. In peaches, MRI techniques can also be used to discriminate between woolly and non woolly fruits. Not all the changes found in the histograms of woolly peaches are similar from those observed in mealy apples pointing to a different underlying physiological change in both disorders (Barreiro et al. 1999, 2000).

Internal browning in pears was studied by Hernandez-Sanchez et al. (2007) using NMR and MRI in order to develop a non-destructive procedure for on-line disorder identification. At higher magnetic field strength and for long pulse spacing, disordered tissue shows higher transverse relaxation rates compared to sound tissue, for NMR relaxometry. In their study, Hernandez-Sanchez et al. (2007) stated that at a macroscopic level, fast low angle shot MR images, effective transverse relaxation-weighted (TR 11 ms and TE 3.7 ms) and proton density-weighted (TR 7.6 ms and TE 2.5 ms), were acquired for pears at a rate of 54 mm/s. In this study they found that a minimum value of 12 % of tissue affected by breakdown can always be clearly identified.

The results of Larrigaudiere et al. (2004) show that when pears are exposed to either disorder-inducing conditions, the cell membranes lose their integrity, which results in changes in water compartmentation. Such variation, along with lack of external symptoms, suggests that non-destructive NMR and MRI are appropriate for internal inspection after storage. Moreover, studies on core breakdown development in pears, an increase in free water in ‘Bartlett’ pears as high intensity areas and the occurrence of air spaces as dark areas in the disordered region in T₁-weighted images were detected by Wang and Wang (1989). Lammertyn et al. (2003a, b) studied the spatial distribution as well as the time course of the disorder by means of MRI and X-ray computer tomography. So, MRI offered a higher sensitivity for detecting incipient browning and a better contrast between disordered and sound tissues than X-ray. Since, a constant pixel intensity threshold was set to separate and quantify disordered and sound tissue within the images. The contrast between sound and brown tissue was followed over time. This showed that the disorder does not grow spatially over time but only increases in contrast magnitude. But, these earlier MR studies on stationary fruit and with slow imaging sequences found unsuitable for rapid on-line detection of the internal disorders (Fig. 5).

Fig. 5

Examples of a seedless and b seed-containing mandarins; upper line corresponds to red green blue (RGB) images; middle line to static magnetic resonance images, and bottom line to motion-corrected magnetic resonance images (Hernandez-Sanchez et al. 2006)

To determine Gray Level Co-occurrence Matrix (GLCM) and Pixel Run-Length Matrix (PRLM) parameters of Iranian important export cultivar of pomegranate Malase-e-Torsh, Khoshroo et al. (2009) analysed images (Fig. 6) obtained by Magnetic Resonance Imaging (MRI) using texture analysis. In their study, they measured T₂ slices at 1.5 T for 4 quality classes of pomegranate semi-ripe, ripe, over-ripe and internal defects classes. According to them, mean classification accuracy for GLCM and PRLM features was 95.75 % and 91.28 % respectively. However, combination of 7 GLCM and 4 PRLM features resulted in mean accuracy of 98.33 % and the lowest type I and II errors. In their research, they showed classification accuracies 100, 98.47, 100 and 95 % for semi-ripe, ripe, over-ripe and internal defects classes. On the other hand, the use of computer vision system for quality assessment of pomegranate has also been reported by Khoshroo et al. 2006 but they are limited by lack of ability to detect internal defects or internal qualities except these succeed to a certain degree. Moreover to above, the pomegranate fruit has low respiration rate and non-climacteric respiratory pattern (Ben-Arie et al. 1984). Early harvest may impede the development of the characteristic colour, taste and aroma of pomegranates, while late-harvested fruits exhibit a reduced shelf life (Kulkarni and Aradhya 2005). So, a potential solution for detecting internal defects and maturity of pomegranate can be the use of magnetic resonance imaging (MRI) technique. Since, MRI can determine the changes in the internal texture of intact fruits during the fruit development and describes the internal distribution of affected tissues with high resolution (Clark et al. 1998). For instance, at different quality stages, Fig. 6 shows MR images of pomegranate (Khoshroo et al. 2009).

Fig. 6

MR images of pomegranate at different quality stages: a semi-ripe, b ripe, c over-ripe, d internal defects (Khoshroo et al. 2009)

Strawberry

Fruit ripening is characterized by dramatic changes in gene expression, enzymatic activities and metabolism. Although the process of ripening has been studied extensively, we still lack valuable information on how the numerous metabolic pathways are regulated and co-ordinated. However, ripening processes can be characterized by small variations in the T₂-weighted contrast and in the relative magnitudes of T₁ between vascular and parenchymal tissue (Goodman et al. 1996).

Since, several factors such as cell size and structure, chemical composition, viscosity of the cellular contents and magnetic susceptibility varied significantly with the fruit ripening. These factors influence proton relaxation that recourse to analytical measurements of physicochemical parameters such as starch and free sugars (Callaghan et al. 1994). According to Ishida et al. (1994) to monitor susceptibility changes can be localized spectroscopy to determine the distribution of dominant chemical constituents by comparison of gradient echo and spin-echo images and using ‘H magic angle and spinning NMR spectroscopy concomitantly can be probed liquid- and solid-phase changes in cohorts (Ni and Eads 1993a,b). So, using these strategies we can interpret the changes in relaxation associated with physiological processes in fruits and vegetables to some extent.

In the study undertaken on water loss rate and decay of strawberry by Jin et al. (2007) using NMR and MRI indicated the changes of water dynamics from mature to corruption of the strawberry that had a great relationship with the water migration. Their study provided a basis for the detection of strawberry through water quality indicators. Otero and Prestamo (2009) investigated the effects of pressure processing on strawberry by nuclear magnetic resonance. They employed two different nuclear magnetic resonance techniques, namely magnetic resonance imaging and ¹H-HR-MAS NMR spectroscopy to study the extent of the damage caused by relatively low pressures (100–200 MPa) in strawberry and reported that the pressure processing causes important changes in strawberries. The observed changes included an extensive redistribution of water molecule and the concentration of air in big bubbles in the parenchyma tissue of the cortex. In their study they concluded that the parenchyma tissue of the cortex or true flesh that showed the highest T₂ and T₁ values indicates the presence of bulk of water (Fig. 7).

Fig. 7

Represents MRI maps of T₂, T₁ and ADC of control, 100 MPa and 200 MPa pressure treated strawberries (Otero and Prestamo 2009)

In addition to above findings, MRI can also used in various other application of strawberry. For instance the determination of pressure affected particular characteristics of strawberry (air content distribution in tissues, pH, and sugar profile and so on) are some of them. So, MRI techniques have been found very useful to better understand the inner quality of strawberries and now the use of MRI, gaining momentum in food industries applications. Summary of MRI/NMR applications for fruit quality inspection has been given Table 4.

Table 4 Summary of MRI/NMR applications for fruit quality inspection

Vegetables

Potato

Potatoes (Solanum tuberosum L.) are considered to be the world’s third most important source of starch as it can successfully grow in a variety of soil and climatic conditions (McComber et al. 1994; Christensen and Madsen 1996). The quality of cooked potatoes is important for many consumers and food producers alike. It is already well known that the structural and chemical composition of a potato depends on many factors including species, maturity, growing and storage conditions (Burton 1989); in particular, texture is one of the most important attributes in its acceptance by both consumers (Thybo et al. 2000) and food processors (Povlsen et al. 2003).

MRI not only can be used to determine the changes in the internal texture of intact fruits during the fruit development and describes the internal distribution of affected tissues with high resolution (Clark et al. 1998), it could also predict sensory texture attributes of cooked potatoes with nuclear magnetic resonance imaging (Thybo et al. 2004a). Since, texture of boiled potatoes is a sensory attribute of uppermost importance for the preference of potatoes, it could be predicted, to some extent, by partial least squares regression of relaxation curves obtained from raw potato samples using pulsed low-field¹H NMR. The prediction of sensory texture quality of boiled potatoes from relaxation curves of raw potatoes is also possible mainly because the NMR relaxation curves contain information on water content, which is inversely correlated to starch content. However, low-field NMR seems to offer additional information, which is not included in the chemical variables like starch content. Apparently, adhesiveness and springiness are slightly better described by low-field NMR than by chemical descriptors. This indicates that low-field NMR is able to probe certain mouth-feel variables due to its sensitivity to the state of moisture in the sample, a parameter it is not possible to measure by any other instrumental technique (Thygesen et al. 2001).

For better results, Martens et al. (2002) attempted for predictive modelling of sensory image description from computer-assisted image analysis and they reported that both the sensory and the computer-assisted image analyses were able to detect differences between varieties as well as storage times. The sensory image analysis gave better discrimination between varieties than the computer-assisted image analysis presently employed, and is easier to interpret. However, some sensory descriptors could be predicted from the computer-assisted image analysis but by using sensory analysis of MR-images offer new information. In another study by Thybo et al. (2004a) also predicted sensory texture quality attributes of cooked potatoes by MRI of raw potatoes in combination with different image analysis methods and reported that the MR-imaging beside giving well-known information about water distribution also gave information about anatomic structures within raw potatoes, which are of importance for the perceived textural properties of the cooked potatoes. Later on Thybo et al. (2004b) detected nonvisible internal bruise and sparing symptoms and got insight on the chemical and anatomical causes of such defects by applying magnetic resonance imaging (MRI). In their study they investigated Cultivars with internal bruise and sparing symptoms by comparison of different MR images as proton density, T₁ and T₂ weighted images and T₂ maps, and reported that the interior of the internal bruise can be characterised as being very dry (low signal in the proton-weighted image) with a small amount of highly mobile water in the shell around the bruise (high signal in T₂-weighted image and high relaxation time in T₂ map) by combining the information in the MR images. In addition, they found that the sparing spots were more diffused. However, the dry interior and highly mobile water around the sparing dots were noticed slightly similar to the appearance of internal bruise which resembled more to the appearance of human’s tumour tissues than bruise disorders in fruits. On the basis of this they concluded that the MRI can detect nonvisible internal bruise and sparing symptoms in potatoes. MRI may, therefore, be an appropriate method for detecting and for studying developmental changes of such disorders and related disorders during postharvest storage in future experiments (Thybo et al. 2004b).

In addition to above Thybo et al. (2003) differentiated between potato samples and determined dry matter content in potatoes by using low field NMR relaxation and NMR-imaging tools and reported that the MR-images showed very large visual differences between tissues, varieties and maturity. The applications of image analysis features in their study were based on the histogram analysis and were not correlated with dry matter content in whole potatoes. So, MR-imaging did not complement the correlations between NMR data and dry matter content. This means that the way the spatial distribution of water in the material was determined did not cover the full information about dry matter content. Therefore, the image analysis of MR-images requires more attention in the future.

The advancement of MRI imaging in various food applications and its success gives several new ideas as the combined use of low-field nuclear magnetic resonance (LF-NMR) and confocal laser scanning microscopy (CLSM) in the study of the effects of salting on low and high dry matter (DM) potato tissue is by Straadt et al. (2008). They illustrated the aptitude of NMR and CLSM to determine and elucidate structural changes and associated changes in water mobility in potato tissue. According to them, the simultaneous use of CLSM and NMR resulted in important information in relation to the interpretation of the origin of the NMR water populations. Since, salting caused the raw potato cells to loose weight, which in the microscopic images was observed as loss of turgor pressure still further away from the edges of the samples with increased salting time. The total water loss after salting was lowest for high dry matter potatoes. The LF-NMR analyses revealed that this could be ascribed to a faster T₂ relaxation time of the cytoplasmatic and extracellular water and thus to water being more restricted in high DM potatoes. In addition, a tendency to a faster T₂ relaxation time was observed by them for both low and high DM potatoes with increasing salting time, which reveals a salt-induced restriction of the cytoplasmatic and extracellular water population. Researchers have been suggested that MRI relates to the water distribution and some anatomic structures within the raw potatoes which are of importance for perceived textural properties of cooked potato. Investigations to date, all employing potato as the raw material, include: measurement of the oil/water interface during frying (Farkas et al. 1992); tissue uniformity following micro-waving (Tewari 1989); and monitoring freezing dynamics (Kerr et al. 1996). Recently, Khoshroo et al. (2011) reported MRI as high power on line tool for internal visualization and maturity detection of pomegranate fruit. So, the human ability to detect, interpret, assess and report visual stimuli can be employed for qualitative and quantitative image analysis.

Tomato

Tomato (Lycopersicon) is one of the most widely grown vegetable crops in the world. It is also one of the most commonly grown vegetables in home gardens. Tomatoes originate from South-America. They are members of the Solanaceae family to which potatoes and egg plants also belong. In tomato, a bruise is a type of subcutaneous tissue failure without rupture of the skin (Mohsenin 1986). A discoloration of the injured tissue indicates the damaged spot in many fruit species but this process does not prevail in tomatoes. To detect and quantify bruise damage in tomatoes, optical techniques encompass thermal imaging, blue laser light scatter imaging and ¹H-NMR imaging. According to researchers, ¹H-NMR imaging technique can be successfully used to illustrate bruise damage in immature green to red ripe fruit. Bruises appeared as strongly hydrated tissue directly beneath the skin and they could extend to the centre of the fruit for red ripe, vulnerable tomatoes. Red fruit also showed intensely hydrated cell layers under the skin, which might contribute to their greater sensitivity. However, impact-bruising over a locule could affect the entire pericarp wall bordering the chamber, and also damage of a cross wall was rather concentrated around the locular gel. The internal pressure of the locule (liquid and gas) might contribute to a higher sensitivity of the tissue directly adjacent to it (Almeida and Huber 2001).

Saltveit (1991) determined tomato fruit maturity with nondestructive in vivo nuclear magnetic resonance imaging. In their study, they imaged ripening tomato and reported differences in water content associated with developmental changes in seed, locular, pericarp and other tissue types. The measurements were also made of the stage of ripeness, rate of carbon dioxide and ethylene production, lycopene and chlorophyll content, density of the pericarp wall, and consistency of the locular tissue. Increased signal was coincidental with liquefaction of the locular tissue, and likewise, signal was higher for the pericarp of mature red than immature green tomato (Ishida et al. 1989). Ultimately, they concluded that the increased intensity was associated with liquefaction of the placental tissue during ripening, while graininess of the pericarp wall was associated with decrease in wall density. So, images of mature-green fruits obtained by NMR imaging technique can be used to differentiate mature and immature fruits. The cost of operation and the long time required to produces an image currently precludes the use of NMR imaging technique to sort the mature-green tomato fruits.

Quantitative studies are now attempting to unravel the factors contributing to time-course changes in image intensity by examining responses in r, and T₂ relaxation, or parameters such as the self-diffusion coefficient of water (Callaghan et al. 1994; Ishida et al. 1995). The patterns are species and perhaps cultivar-dependent. In tomato, for example, T, relaxation increased in all tissues as the fruit ripened (Ishida et al. 1994), demonstrated that MRI could distinguish red tomato from green, but that discrimination of immature-green from mature-green tomatoes was poor (Chen et al. 1989).

Sequi et al. (2007) demonstrated the possibility of assessing unambiguously the place of origin of fresh cherry tomatoes by means of MRI experiments. Since, this method determined informative physical and morphological parameters, transverse relaxation times and thicknesses. They approached to innovate the use of MRI based on a set of four empirical equations which have taken into account the pericarp thickness, the width of the inner and outer spherical crown composing the pericarp itself, and their transverse relaxation times T₂ to distinguish PGI (Protected Geographical Indication) cherry tomatoes from non-PGI ones, as well as cv. Naomi from cv. Shiren samples. According to them, this approach had successfully recognized the cultivated variety of cherry tomatoes 90 % of the analyzed samples and the geographical origin of cherry tomatoes 80 % of the investigated cherry tomatoes. Musse et al. (2009a) monitored the postharvest ripening of tomato fruit using quantitative MRI and NMR relaxometry and identified air spaces close to seeds and their shrinkage during the ripening period estimation from the spin echo images. The development of the bubbles in the outer pericarp during ripening was estimated from the ratio of the long- and short-echo time gradient echo MRI images and supported by the macrovision imaging. Variations in the transverse (T₂) and longitudinal (T₁) relaxation times were determined from quantitative MRI images. They provided information about all major sub-cellular compartments and showed there was water redistribution among compartments during ripening. In addition to the relaxometry measurements, water content, weight loss and concentration of neutral sugars and acids were measured on some of the tomato fruit. Cell size and organization were investigated by macrovision experiments. Although the overall dependence of the relaxation time on tissue type was to some extent explained by chemical composition and cell dimension, no relationships between trends in MR data and tissue properties were established. An investigation of the structural aspects of the tomato fruit by means of quantitative nuclear magnetic resonance imaging was also done by Musse et al. (2009b). They applied MRI to study the structural aspects of the tomato fruit using a 0.2-T electromagnet scanner. For this, they acquired spin–echo images (Fig. 8) to visualize the tomato macrostructure and evaluate the air bubble content in tissues by exploiting susceptibility effects using multiple gradient echo images and by measuring spin–spin (T₂) and spin–lattice (T₁) relaxation time distributions and studied the microstructure. In order to emphasize the MRI results, NMR relaxometry, macro vision imaging and chemical analysis were used as complementary and independent experimental methods by Musse et al. (2009b). MRI images showed that the air bubble content varied between tissues. They also attested the presence of gas by macro vision images and found that the quantitative imaging by MRI reflected the structural differences between tomato tissues and made it possible to distinguish between them. Consequently, they also concluded that the cell size and chemical composition contributed to the relaxation mechanism. Since, the specific character of a bruise tomato can enable the imaging of the affected area by means of MRI can be used as a powerful technique to visualize bruise damage in mature green to red ripe tomatoes. Applications of MRI/NMR in quality inspection of vegetables have been summarized in Table 5.

Fig. 8

Spin-echo MRI image of the median equatorial plane of tomato with tissue annotation (Musse et al. 2009a)

Table 5 Summary of MRI/NMR applications for vegetable quality inspection

Other fruits and vegetables

Clark et al. (1997) published a review of MRI application to fruits and vegetables. Krutz et al. (1993) reported a correlation between the NMR result and the Magness-Taylor force of only (r = 0.48), since the estimation of firmness by NMR is usually achieved by quantification of the maturity state or water mobility. Chen et al. (1993, 1996) used NMR with avocados (both statically and on-line) and obtained T₂ and T₁-enriched spectra and images to quantify the water state. Chen et al. (1993) were able to use MR to measure the oil/water ratio in intact avocado as an index of fruit maturity. A similar methodology (used with static fruit) was found useful for detecting internal softening in heated papayas (Suzuki et al. 1994), internal decay in melons (Zion 1994). Mealiness (or wooliness), a combination of softening with a reduction in water mobility, can be detected in nectarines (Sonego et al. 1995) by NMR. However, the use of NMR for the on-line, non-destructive estimation of firmness is restricted by the high cost of MRI devices and the need to resolve problems associated with capturing signals from moving samples. In addition to this, Andaur et al. (2004) used MRI to measure wine grapes, developed reconstruction techniques for 3-D representation of berry clusters. Since, size distribution of berries is important to predict extractability in grapes and the extractability is a function of the skin-to-flesh ratio and it is inversely proportional to the volume of the berries.

Cereals

Modern food industries intended to develop high quality food products with superior taste, texture, freshness and shelf life that are ready to eat or quick and easy to prepare. For this food industries all over world are continuously innovating in every food segments. One of them, cereals and cereal products are important segment of food innovation. Cereal products such as bread, biscuits, snacks and staple foods (preprocessed) comprises important features (crispness and crunchiness etc.) in terms of consumer appreciation of dry porous cereal materials. Water/moisture absorption, distribution and mobility in these products during storage and preparation are responsible to the loss of crispness and crunchiness. So, on line technology is needed to prevent and reduce water migration efficiently from the moist filling to the dry porous cereal material in real time. For this, the knowledge of the micro-structure and its relation to water mobility is necessary to develop staple products. MRI, the novel technique allows the real time visualization of the effected hydration in terms of transport/diffusion rate and distribution of structural changes (Mohoric et al. 2004; Ramos-Cabrer et al. 2005; Van Dalen et al. 2005).

With the advancement of application of MRI technology, Horigane et al. (1999) analysed water distribution and structural changes in cooked rice grains with NMR micro imaging of protons (¹H) nondestructively. They observed internal hollows in all examined cooked rice grains, and proposed a mechanism to explain their formation. The origin of the hollows was hypothesized to be cracks or fissures, and hollows resulted from sealing of such lacerations by gelatinized starch in the peripheral layer in combination with expansion of the grain during cooking. Consequently they concluded that the NMR micro imaging has a considerable potential for whole grain analysis. However, in the view of results of Takeuchi et al. (1997b), hollow formation needs to be confirmed for other cultivars and evaluated for use as a control for the texture of cooked rice.

For better results, Mohoric et al. (2009) investigated the combined use of X-ray microtomography and MRI to explore the effect of rice kernel microstructure on cooking behavior. They examined the mesostructure of the rice kernels by DSC (Differential scanning calorimetry) and XRD (X-ray diffraction), the microstructure by μ-CT, and hydration behaviour by MRI in a real-time manner during cooking and reported that a novel MRI method allowed the observation of the ingress of water in rice kernels during cooking in real-time mode, and with good spatial resolution in 3D.

Furthermore, Takeuchi et al. (1997a) observed the change of moisture distribution in a rice grain during boiling in real-time by NMR T₂ imaging and the rise in moisture profile was unexpected. They analyzed fully the error in the NMR imaging method caused by the effect of diffusion on the apparent T₂ by using the pictorial phase diagram method and then result showed that the cell walls have little effect on resisting moisture migration. Similarly, to examine the water distribution and migration in single rice kernels during the tempering process, Hwang et al. (2009) applied MRI technique and performed imaging experiments in a Bruker 9.4 T MRI system and revealed that the MR imaging technique is useful in investigating the dynamic moisture migration within individual rice kernel. According to them, it can also be used as a reference or validation for the simulation model of moisture migration, and as an efficient tool to examine the mechanisms of moisture migration within cereal grains. In another study conducted by Ishida et al. (2004) the loss of moisture from harvested rice seeds during drying process was visualized with the single point mapping imaging (SPI) technique by MRI at various temperatures, and compared the results obtained with those of the oven-drying method. Consequently, they concluded that the most of the water was present in the embryo and endosperm and the water reduction rate was larger on the outside than in the central position of the rice seeds at 50 °C. Although this discrepancy was not obvious at 40 °C but water reduction was brought about with time according to the kinetics of the multiple components, for both MR imaging (Fig. 9) and the ventilated-oven method. But they noted that the reduction rate of water from rice kernels increased rapidly with temperature (up) to near 60 °C and then rose slowly above 60 °C when images were measured continuously (10 min per image for 100 min).

Fig. 9

Changes of the MR image of rice seeds during drying at various temperatures. Arrows indicate cracks in rice grain (Ishida et al. 2004)

Difference in water content and mobility in various components of corn kernels during the steeping were measured by Ruan and Litchfield (1992) using MRI. Water mobility and distribution inside the corn kernel during the room temperature steeping was found by them over the narrow range studied. So, this technique can be used to study corn steeping process, such as the SO₂ and lattice acid effects on the structure change and water absorption of maize kernel. However, Borompichaichartkul et al. (2005) captured MR-images within corn kernels during the freezing process for characterization of the physical state of water in frozen corn (Fig. 10 ) and they stated that the brighter areas show locations where proton mobility is high, and thus water exists as a liquid. In this example, MRI provides useful information for characterizing the physical state of water in frozen corn. However, changes in tissue morphology and the distribution of water occur when samples are frozen, then thawed. This effect has important implications for the frozen-food industry. But in the case of fresh fruits and vegetables, crops exposed to frosts may develop undesirable quality characteristics during storage if external signs of freezing damage are not present which enable spoiled samples to be identified and culled at harvest.

Fig. 10

MRI images showing the distribution of water and its freezing behavior in different areas within the corn kernels: a images captured before freezing at different moisture contents; b and c images acquired at specified temperatures and moisture content levels (Borompichaichartkul et al. 2005)

NMR imaging can also be used to study the boiling and steaming of intact wheat grains by mapping the distribution of water on the central cross section of grains at various cooking times (Stapley et al. 1997). The results showed different distributions of water occurring with boiling and steaming. Boiled grains showed moisture in front of which gradually moved towards the centre of the grain, although some water was able to diffuse ahead of the front. A 3D magnetic resonance imaging (MRI) technique was adapted by Song et al. (1998) for the probing of single wheat grain at storage moistures (c. 12 % wet basis). Since, to improve the milling performance and to increase yield, a better knowledge of moisture distribution and migration in individual wheat grains during tempering is essential. So, a three-dimensional microscopic MRI with 3DPR technique was developed by them and variations in endosperm moisture concentration (estimated range 7·3–16·4 % wb) were revealed. Later on, the diffusion weighted MRI was used to obtain diffusion constants for water in the embryo and endosperm of wheat by Gruwel et al. (2008). In their experiments, they found a significant difference between diffusion constant for two components. Water diffusion in both the endosperm and embryo deviates from the typically observed Gaussian behavior in bulk fluid. They also concluded that the diffusion weighted MRI can be used to obtain diffusion constant of water inside seeds.

So, it can be concluded on the basis of above researchers experimental results that MR imaging could be used as an efficient tool to examine the mechanisms of moisture migration within cereal grains nondestructively. For instance, Table 6 summarized some applications of MRI/NMR for grain quality inspection.

Table 6 Summary of MRI/NMR applications for grain quality inspection

Fish and meat products

The application of MRI to inspect fish and fish products is not new. Since, MRI is a technique that offers a unique opportunity to produce high quality cross-section images of intact organisms and thus to obtain basic insight into a number of issues related to anatomical studies, composition and structure of tissues, distribution maps of fat, water and salt as well as temperature mapping. In fish processing, MRI can be used as a tool for the optimization of various unit operations such as salting, freezing and thawing, some of which are described by Hills (1998). Howell et al. (1996) demonstrated frozen and unfrozen zones in partially frozen fish and Nott and Hall (1999) studied the influence of the freezing methods such as repeated freeze-thawing cycles and storage time of fresh and frozen-thawed trout on the MR measurable parameters. In their another study of freezing and thawing of cod and mackerel had showed that the relaxation time T₁ decreased in both lean and fatty fish samples after the freezing-thawing cycle, while the magnetization transfer rate and T₁ increased (Nott and Hall 1999). According to Collewet et al. (2004), strong longitudinal relaxation time (T₁) weighing of ¹H images can be used to produce good contrast between muscle and fatty tissue for visualization purposes. Later on several fresh and frozen-thawed trout samples (from 1 to 41 days frozen storage time) were simultaneously investigated by Foucat et al. (2001) using several MRI parameters such as magnetization transfer ratio, T₂ relaxation time and diffusion constants parallel and perpendicular to the muscle fiber orientation and concluded that the freezing-thawing cycle significantly affects only mean T₂ and diffusion perpendicular values. However, for a long frozen storage period (41 days) only the diffusion perpendicular value was affected, compared to that in fresh fish. So, the low field (LF) NMR has been proven to be a versatile analytical method for studying various topics such as effect of processing (Steen and Lambelet 1997; Erikson et al. 2004), muscle fat and water content (Toussaint et al. 2001; Sorland et al. 2004).

Furthermore, for the better results, Bock et al. (2002) demonstrated the feasibility of using MRI to image live unaesthetized fish in a flow tube through and MRI scanner. They used flow-weighted ¹H MRI to study the effect of temperature on circulatory performance of Baltic Sea eelpout. Similarly, Foucat et al. (2004) carried out ¹H MRI studies of water dynamics in white rainbow trout muscle with “soft flesh” which is a known phenomenon occurring apparently at random in farmed fish. Belly bursting in herring is a phenomenon responsible for the severe quality deterioration of the catch. But when Veliyulin et al. (2007) studied this process in real time they reported that neither the stomach itself nor zooplankton in the stomach are sources of enzymes responsible for belly bursting in herring. They conclude that at least two anatomical structures are the origin of belly bursting in herring: the ventral muscle and the part of the intestine where the pyloric caeca empty their enzymes.

Moreover to above, several other studies such as quantification of salt and effect of energy diet on fat content were also done by researchers. Toussaint et al. (2005) made study on the effect of a low- and a high-energy diet on fat distribution in brown trout by using ¹H MRI and Veliyulin et al. (2006) have published an MRI study of freezing of cod fillet pieces where T₁ and T₂ relaxation times as well as the fraction of unfrozen water in the fish sample were calculated as a function of temperature. In recent, Veliyulin et al. (2009) have demonstrated a possibility of quantitative sodium imaging in model food samples by using the dedicated MRI protocol (SPRITE). Gallart-Jornet et al. (2007) studied salting of cod and salmon fillets, with skin on, at different brine concentrations by using 23Na MRI. However, in another study, Aursand et al. (2010) performed a 23Na/¹H MRI study on Atlantic salmon to see the effect of ante-mortem handling stress and rigor mortis on muscle water properties and subsequent post-rigor salting. According to them, 23Na MRI revealed differences between the two treatments in fillet salt distribution, where the salt penetration in exhausted fillets was more pronounced. Summary of MRI/NMR applications in fish and meat quality inspection has been given in Table 7.

Table 7 Summary of MRI/NMR applications in fish and meat quality inspection

Processed food products

Food can be non-invasively and in real time tested in transparent to radiofrequency waves containers and if it is free of ferromagnetic components. Magnetic resonance imaging (MRI) applies external linear magnetic field gradients for the association of each spatial region with characteristic magnetic resonance frequency of nucleus. MRI can be applied in many fields of food processing and storage. Since, MRI can be applied in many fields of food processing and storage.

Colour, taste, aroma, texture, the concentration of factors such as vitamins, sugars and lipids, and the extent of denaturation and gelatinization of biopolymers are some of the factors that determine the quality of a food. According to Hills (1995), these quality factors are created and altered during processing and storage by enzymic and non-enzymic chemical reactions. By providing real-time images of time depended moisture content distribution, temperature distribution and some quality factors (lipid and solute concentrations), MRI can greatly facilitate time consuming traditional operations. Lipid and solute concentrations (by chemical shift imaging) were reported more amenable to MRI analysis by researchers. Furthermore, Duce and Hall (1995) visualized the hydration of butter bean, rigatoni pasta and a snack food non-invasively by NMR imaging. They observed the rate at which water is absorbed into the samples and changes results, and investigated the relationship between the intensity of the image signal and water content. In their study, they reported that the movement of the water front into the butter bean can be monitored. In the case of the Pot Noodle, they studied the hydration continuously inside the magnet and demonstrated a direct relationship between the intensity of the image signal of the pasta and water content. Thus by carefully designing the imaging experiment, it is possible to determine the moisture content of pasta during hydration.

For the better results, researchers have now been used MRI micro imaging and localized spectroscopy to image the production and spatial distribution of maltose and other metabolites in germinating barley seeds which affect the quality of beers and spirits. McCarthy and Kauten (1990) have also been monitored the transport of fat into bread from peanut butter with MRI. Since, in food only water and lipids can produce adequate signals for imaging and the signal intensity can alter due to different concentration of water molecules, relaxation time (T₁ and T₂-weighted images), diffusion coefficient, or flow velocity (Olendorf and Olendorf 1991; Hennel and Kryst-Widzgowska 1995). Water relaxation times depend on quality factors such as protein and polysaccharide aggregation, gelatinization, pH and solid to liquid ratio (Hills 1995; Kerr et al. 1998). The examples of other processes that are amenable to study by MRI are blanching, frying, micro-waving, extrusion, curing, mixing, draining, freezing and freeze-drying (Hills 1995).

So, MRI imaging is very useful in monitoring of the formation of ice during food freezing. Loss of the proton mobility as a result of phase transition is detected by the decrease of signal strength (Tang and Hills 2000). In some instances, MRI research was conducted to measure moisture and its migration in food systems; other researches demonstrated the potential of using MRI to investigate physical or biological properties of food products; still some other work sought to estimate the yield (Du and Sun 2004).

Moreover the above applications of MRI, NMR relaxation time measurements provide a possibility to view the food microstructure. Many experiments are possible using a low-resolution NMR spectrometer system. According to Guthausen et al. (2002), for the simultaneous determination of oil and moisture content, Hahn-spin-echo-sequence can be used with the upper limit of water of about 15 % but in fresh food, the determination of fat or oil content, with high water content is possible when pulsed-field-gradient spin-echo-method is used. However, the field gradient NMR is used to suppress unwanted signals in studies of water droplet size distribution in oil (margarine, butter) (Fourel et al. 1995) as well as in measurements of oil droplet size distribution on oil-in-water emulsion (mayonnaise, dressings) (Goudappel et al. 2001). Hinrichs et al. (2003) have been concluded that the T₂ measurements and diffusion experiments are very useful for characterizing gels (e.g. yogurts).

The redistribution of water in prototype food systems, comprising phases with contrasting water activity (Aw), was investigated by Weglarz et al. (2008). In their research, they adapted MRI techniques to study the migration rate and RARE and SPI measuring methods were used to monitor water redistribution in crunchy inclusions in water and biscuit shells with apple filling during storage, respectively. In the first case, fast migration, which typically last some tens of minutes, was monitored with a temporal resolution 3.5 min or better, while in the second case of slow migration it was monitored during 2 months with 10 days temporal resolution. 3D MR images with sub-millimeter resolution visualize the spatial redistribution of moisture and allow a quantification of its rate and extent of matrix swelling. Correlation with high resolution X-ray (XRT) image allows to identify structural elements responsible for unwanted fast hydration. The results obtained demonstrate the potential of a combined MRI/XRT approach for dynamical monitoring of the migration of moisture into porous cereal material in a broad range of Aw contrast and hydration levels. The measured migration rates through samples of different internal structures can be used for validating models for prediction of water redistribution in multiple phase systems in a quantitative manner.

Other applications of MRI/NMR techniques

The NMR was also used to identify swill oil, waste oil and edible oil. In the research, the value of solid fat content (SFC) was an indicator which was assumably 0 in edible oil, while very high in swill oil and waste oil. This method could identify the 1 % incorporation of the waste oil in fresh oil (Wang et al. 2008). The study demonstrated by Shaarani et al. (2010) on monitoring development and ripeness of oil palm fruit (Elaeisguneensis) by MRI and Bulk NMR showed that both MRI and bulk NMR effective techniques to measure the ripening of an intact oil palm fruit. According to them, the mono-exponential and biexponential curve fit to T₂ data obtained from MRI provided reliable information about the oil and moisture content, which was separately measured, because of the difference in their spin–spin relaxation time. The increase in T₂ values for both mesocarp and kernel with WAA (weeks after anthesis) indicated the increase in the oil content in both components of the fruit. The multiexponential fitting of the bulk NMR T₂ relaxation data for the intact fruit demonstrated three distinct components; T₂₁ (intermediate component, 23–30 ms) represents the water and T₂₂ (long component; 64–144 ms) assigned to oil, while T₂₃ (short component; 3–6 ms) was related to the shell.

The research conducted by Chen et al. (2006) proved that MRI technology was an effective method for the identification of different types and origins of tea. In the study they analyzed 20 chemicals of 35 different types of tea collected from Fujian, Yunnan, Guangzhou, Jiangxi and other places with ¹H NMR fingerprint technologies, and they found the chemical composition that could distinguish between different types and origin of tea by principal component analysis (PCA) and cluster analysis. Zhou et al. (2008) developed a proton nuclear magnetic resonance (¹HNMR) spectra fingerprinting to identify beer origin. They analyzed 19 kinds of commercial canned beer, and successfully identified the origin place of all kinds of beer by pattern recognition analysis. Some important industrial applications of MRI techniques have been listed in Table 8.

Table 8 List of some important industrial applications of MRI techniques

Drawbacks, limitations and advantages

The main drawbacks and limitations of the MRI technology when it comes to the use as an on-line scanning technique in food detection are:

• The expensive equipment limits its application in the food industry

• The complex analysis of testing data requires specialized software and algorithms and it is difficult for ordinary inspectors to master it.

• It needs long imaging time

• Often too small magnet bore opening creates some difficulty

• Shear size and infrastructure requirements

• High costs and requirement of trained personnel

The main advantages of MRI/NMR techniques are:

• Non-invasive technique permits repeated analysis at a later date

• An image slice can be taken in any plane through an object

• Chemical shift may be encoded so that only protons from water, sugar or oil appear in the image

• Contrast may be manipulated by changing image sensitivity to T₁ or T₂ relaxation, or to display dynamic processes such as flow or self-diffusion.

Conclusion

From the above work it has been found that MRI is an effective technique for internal quality assessment in a wide variety of fruit species: apple, avocado, blueberry, cherimoya, courgette, cucumber, durian, kiwifruit, mandarin, mango, melon, nectarine, olive, onion, orange, papaya, pear, peach, pineapple, potato, tangarine, tomato, strawberry, melon, and watermelon. In this work, we have discussed fundamental principle, components of MRI system and techniques applied recently in postharvest processing, sorting and quality evaluation of fruits, vegetable, cereals and processed foods.

Magnetic resonance imaging (MRI) is routinely used for diagnostic purposes in medicine, and its non-invasive and non-destructive features, together with the capability of providing highly resolved spatial information about proton density and their relaxation processes in soft tissues, make MRI a powerful and attractive analytical technique for physiological investigation of agricultural produce.