Introduction

Generally, it is believed that patients with Parkinson’ disease (PD) present with resting tremors, while those with essential tremor (ET) mainly present with kinetic tremors. Increasing evidence, however, has demonstrated the overlap in the tremor presentations of these two diseases, making it difficult to distinguish patients with PD and ET, especially during an early stage of disease. According to previous studies, the proportion of patients who were diagnosed with ET at the early disease stage but actually had PD was 15%, while 7% of patients had both tremors and comorbid PD. Therefore, a total of 22% of patients had PD, but were not diagnosed [1]. This correlation may be attributed to having ET with subsequent PD (ET-PD) [2] or to patients with PD showing late-onset characteristics of ET at the pre-diagnostic phase [3]. Therefore, how to distinguish PD or ET with tremors is still a problem waiting to be solved.

A couple of studies have employed tools, like tablets and optical capture devices, to quantify the severity of the tremor [4, 5]. Others have also analyzed whether additional parameters, including multiple differential functions, smoothness, and drawing velocity, can be used to distinguish between patients with early PD from healthy populations [6]. To clarify differences in motor functions between patients with PD and ET, Yu et al. [7] invited patients to use tablets to draw simple geometric shapes. Analyzing the velocity of these patients, they demonstrated that when drawing continuous cursive loops, the velocity of the patients with PD was significantly slower, with progressively smaller loop sizes, which was different from patients with ET. Visuomotor coordination is another phenomenon that is worth analyzing using digitized tablets, since patients with PD show degeneration of visuospatial functions, which can be compensated by visual guidance. However, as ET is partially caused by cerebellar dysfunction, these patients also have reduced movement coordination [8]. Therefore, when performing visual guidance, their tremors may sometimes become more severe. Therefore, previous studies have demonstrated that when visual guidance disappears in patients with ET, their tremors were alleviated [9]. In patients with PD, however, the errors become larger after visual guidance was lost [10]. Schwartz et al. previously employed visuomotor coordination characteristics to identify patients in the ET population with visuomotor defects that were similar to PD. They found that in these patients, the proportion of reduced striatal ligand uptake by [123I]-FP-CIT SPECT was positively correlated with the severity of visuomotor dysfunction. Schwartz et al. proposed that this form of ET with symptoms of visuomotor defects may in fact be early-stage PD [11, 12]. Previous studies have also examined differences in visual guidance between the two diseases, using indirect visual guidance where the hand movement information of participants were first transmitted to the screen, before the hand location on the screen was brought towards the guiding point. Direct visual guidance, however, would allow us to better understand differences in visual guidance effects on both diseases. In addition, previous studies have mostly used angle deviations or circular displacement, or average velocity measurements, for data analysis [12,13,14]. Given the evolution of digitized tablets, the display function of digitized tablets has now evolved. Therefore, participants can simultaneously receive visual guidance on the screen, while real-time recording of movement trajectory is being performed. It is also possible to directly analyze velocity distribution and simultaneous changes in velocity. If the ability of direct visual guidance to distinguish movement characteristics of the two diseases can be utilized, this will provide physicians with more clues to diagnose both diseases. Therefore, the aim of the current study was to compare characteristic parameters of participants with PD and ET during visually guided spiral tasks.

Materials and methods

Participants

In this study, 12 participants with PD (average age 67.3 ± 5.3 years; range 57–75 years) and 14 participants with ET (average age 70.6 ± 7.4 years; range 55–83 years) were recruited from the Department of Neurology of National Taiwan University Hospital. A neurologist was in charge of clinical evaluation. ET was evaluated using the Fahn-Tolosa-Marín Clinical Rating Scale clinical scale, while PD was evaluated using the Unified Parkinson’s Disease Rating Scale Part III for motor performance evaluation. In addition, 12 healthy participants (average age 65.7 ± 7.5 years) were included in the control group.

Table 1 outlines the clinical scale scores of the patients.

Table 1 Clinical scale score and age of the 12 participants with Parkinson’s’ disease (PD) and 14 participants with essential tremor (ET) who were recruited for the current study
Full size table

Patients with ET had to fulfill the Tremor Consensus Essential Tremor Criteria by the Movement Disorder Society [15], while patients with PD had to fulfill the UK Parkinson’s Disease Society Brain Bank Criteria [16]. Participants reduced their medication dose by half, following the instructions of the clinicians, before the experiment. Each subject fully understood the aim and procedures of the experiment and provided informed consent before the study. The study protocol was approved by the Research Ethics Committee of the National Taiwan University Hospital Hsinchu Branch.

Experimental setup and data collection

Wacom Cintiq 13HD tablets (Wacom Co., Ltd., Saitama, Japan), with a display function, were used to record the spiral images drawn by the participants (screen dimensions 13.3 in.; resolution 5080 lines/in.; capture rate 50 Hz). The computer software LabVIEW 2013 (National Instrument, Texas, USA) was used to write the experiment image design and data capture programs for two different spiral tests.

These methods were used to compare the effects of visual feedback on spiral drawing performance in patients with PD and ET. The two tests, which are shown in Fig. 1, were as follows: (1) tracing a given spiral (T1), where the participant traced a spiral on the tablet from inside to outside; (2) following a guiding point (T2), where the participant followed a guiding point with a movement velocity of 3.0 cm/s (average velocity of control group participants in the T1 test) to draw a spiral. Participants first underwent several practice trials to familiarize themselves with the process. During the drawing process, the elbow, wrist, and palm of the participants was suspended in the air; the table or tablet could not be used for support. Three valid data trials were collected from the more severe hand of every patient or the dominant hand of every healthy participant (happened to be the right hands for all participants) for data analysis. Three clinicians blindly evaluated the drawings of all the test results independently using the visual rating scale (VRS).

Fig. 1
figure 1

a The participants performed two spiral tests. The fine lines are the target trajectory, while the coarse lines are the actual trajectory drawn by the participants. The two tests are tracing a given spiral (T1) and following a guiding point (T2). b An example of T1 test results. The fine line is the target trajectory, while the coarse line is the trajectory drawn by a participant. In the current study, the mean and standard deviation of the absolute velocity of spiral drawing were calculated. The calculation method for absolute velocity is the distance between two points (d) divided by time. c The graph is an example of velocity distribution and presents the probability distribution graph of absolute velocity of spiral drawing by another participant

Full size image

Data analyses

This study investigated whether visual guidance would affect participants when drawing spirals. Therefore, the spiral drawing velocity of the participants under different visual feedback conditions was calculated to analyze whether there are significant differences between the various groups. The velocity parameters included absolute velocity (V) and standard deviation of velocity (SDv) when drawing spirals. If the drawing trajectory Pi (ri, θi) was represented and described by polar coordinates, then the absolute velocity V was calculated by taking the distance (d) between the position at the first time-point (P1) and the position at the next time-point (P2) divided by unit time (0.02 s) (Fig. 1b). In the guiding point test that followed, if the participants had better velocity control ability, their average absolute velocity was close to the velocity of the guiding point (3.0 cm/s). In addition, a distribution map was generated to represent the absolute velocity of drawing. The velocity ranged from 0 to 10 cm/s, with 0.2-cm/s intervals. The proportion of the number of times the absolute velocity appeared in each interval was calculated (Fig. 1c). Interruptions during spiral drawing resulted in a velocity of 0 cm/s. In order to decrease the effects of this factor, speeds of 0–0.2 cm/s were excluded from calculation of average velocity during spiral drawing. An independent t test was used to analyze whether there are significant intergroup differences for each parameter. Paired t tests were used to analyze whether there are significant intragroup differences under different test conditions. Finally, tracking indicators of both groups of patients and healthy control participants under visual guidance were calculated. The calculation method consisted of taking the area under the curve, within a velocity range of 2.0–4.0 cm/s, for the velocity distribution graph of the T2 test. The ratio of this area to the total area of the distribution graph was defined as a following performance. If the following performance was high, then the drawing velocity of the participant was approximately concentrated around the guiding velocity, 3.0 cm/s, which meant that the drawing status of the participant is close to the velocity of the guiding point. Since the tremor severity of patients with PD and ET were not the same, to account for the effect of tremor severity, patients with similar tremor severity (VRS scores of 4–6) in both groups were further separated and their tracking indicators were re-analyzed. Besides, the correlations between the following performance and clinical severity, FTM and UPDRS Part III scores to all ET and PD patients, respectively, were also calculated.

Results

Spiral testing was conducted using two guiding methods to compare differences between the groups in the absolute velocity and standard deviation of velocity during spiral drawing. Figure 2 shows the velocity probability distribution graph of the drawing velocity of the three groups. The velocity differences between T1 and T2 in the control group were not large, as most velocities approached 3.0 cm/s. Patients with PD were slower in T1 velocity compared with the control group, but their T2 performance approached 3.0 cm/s.

Fig. 2
figure 2

Mean velocity distribution graph of the control, Parkinson’ disease (PD), and essential tremor (ET) groups for the T1 and T2 tests

Full size image

According to the intergroup statistical results (Figs. 3a and 4b), the average absolute T1 velocity of patients with PD was significantly lower than that of the control group. In T2, the average absolute velocity of the patients with ET was significantly higher than the control group (p < 0.05). According to the results of the intragroup paired t test, the average absolute velocity and standard deviation of velocity in the PD group showed significant differences in both T1 and T2 tests (p < 0.05). In addition, from the tracking performance of the groups based on the following performance (Fig. 3c), the tracking performance of the control group was the best and significantly higher than the PD and ET groups. This was followed by the performance of the PD group, which was significantly higher than the ET group. When both groups of patients with a VRS severity of 4–6 were independently analyzed (Fig. 3d), there are significant differences in the following performance between PD and ET (p < 0.05). Figure 4 shows the relationships between the tremor severity in ET or motor symptoms in PD and their following performance. The Pearson correlation coefficients between the tremor severity and the following performance for all ET patients and between motor symptoms and the following performance were both below 0.5 (p > 0.05).

Fig. 3
figure 3

a Mean absolute velocity and b standard deviation of velocity for the T1 and T2 tests of the control, Parkinson’s disease (PD), and essential tremor (ET) groups. An independent t test was used to analyze whether there were significant intergroup differences between the tremor groups and the control group in the two tests (*p < 0.05); a paired t test was used to analyze whether there were significant intragroup differences in different tests (#p < 0.05). c The tracking indicators of the three groups in the T2 test. Tracking was the highest in the control group, at 0.69; tracking of the PD group was significantly higher than the ET group. d When tracking indicators of both groups of patients with the same visual rating scale (VRS) severity range were analyzed, they were significantly higher in the PD group than in the ET group. (*p < 0.05)

Full size image
Fig. 4
figure 4

Correlation between disease severity and following performance in PD and ET patients in T2. Pearson correlations show in thin lines. There was no significant correlation between UPDRS Part III scores in PD and following performance (R = − 0.49; p = 0.11). All following performance ratios of PD patients were more than 0.4. There was no significant correlation between FTM scores in ET patients and their following performance (R = − 0.25; p = 0.38). All performance ratios of ET patients were less than 0.5. In our study, approximately 0.47 (red dotted line) of following performance in T2 can separate ET from PD patients. UPDRS, Unified Parkinson’s Disease Rating Scale; FTM, Fahn-Tolosa-Marín Clinical Rating Scale; PD, Parkinson’s disease; ET, Essential tremor

Full size image

Discussion

The current study demonstrated that during static tracing (T1), the velocity of patients with PD was significantly slower than normal participants. During dynamic tracing, i.e., following a guiding point (T2), however, patients with PD could achieve the velocity of normal participants. The phenomenon of slower tracing velocity of patients with PD cannot, however, be explained by rigidity or bradykinesia, because during dynamic visual guidance (T2), the drawing velocity of these patients increased, even achieving the required execution velocity (3.0 cm/s).

We postulated that, the cause of slower velocity could be abnormalities in cortical sensorimotor integration related to vision in patients with PD [17], rather than bradykinesia. There are studies where patients with PD were tasked to draw houses, but the windows and house were not connected. This indicated problems in precise positioning and control in planar spaces [18]. In another study, before patients took their medicine, the area of the drawn circles were small, with a significantly worse angle and prediction rate than normal participants [19]. Therefore, when patients were required to draw along the template as closely as possible, they might take more time to complete the task, thus, decreasing the drawing velocity [7].

When dynamic guidance is provided and patients with PD were only requested to follow the velocity of the dot, they could still achieve the required velocity. This means that the velocity control based on external visual guiding was still functional in patients with PD. However, this does not mean that dynamic guidance can cause all drawings to be normal, since the variation in T2 speed in patients with PD was significantly higher than in normal participants. This demonstrates that their stability in tracing the path has become worse and the dexterity of their hands is still abnormal compared with normal subjects. In comparison, there were no significant changes in velocity parameters in ET patients, regardless of whether they were tracing or following a guiding point. In T2, the absolute velocity of patients with ET was significantly higher than the control group (Fig. 3a). This means that both control and PD groups can approach the guiding point velocity of 3.0 cm/s based on dynamic visual feedback. The ability of patients with ET to follow the dot, however, was poorer, resulting in significant differences between the groups.

The paired t test used to analyze results based on different test conditions for individual participants (Fig. 3a) demonstrated no significant difference in velocity between T1 and T2 tests for patients with ET. This represents that differences in static and dynamic visual feedback did not result in differences in drawing velocity for patients with ET; however, this was obviously different for patients with PD. Intragroup paired t test of normal participants showed that there was no significant difference between T1 and T2 in velocity and its variation. This implied that the basal ganglia and cerebellum still functioned like normal. However, normal participants differed from patients in ET, during T2. The proportion of their overall performance tending to the specified velocity was higher, while the variation was lower. This indicates that they were able to obtain more stable movement speed under dynamic guidance.

In the current study, patients with ET had an average VRS of 5.2, which was more severe than the average VRS of patients with PD (3.2). Therefore, we could not exclude that the poorer tracking performance in the ET group was not due to kinetic tremors. Therefore, we grouped patients with PD who had kinetic tremors with those with ET who had a similar VRS severity (VRS = 4–6) for a subgroup analysis. The patients with ET still could not, however, maintain stable tracking velocity when compared with patients with PD. This subgroup analysis demonstrated that the stability of tracking velocity was not related to the tremors itself. Hence, the possible reason for poor tracking status could be deficits in eye-hand coordination in patients with ET, since the velocity distribution graph (Fig. 2) shows two peaks in the distribution of ET velocity, which appeared when the velocity approached zero. We hypothesized that this phenomenon could be attributed to deficit in eye-hand coordination or intentional tremors during the tracking process of patients with ET. In order to complete the spiral tests, patients with ET had to slow down or stop their movements before re-aiming at the target. Previous studies have highlighted that in addition to kinetic tremors, the symptoms of ET also include functional changes in eye-hand coordination [14, 20] and motor timing disorders [21, 22]. Currently, it is widely believed that these symptoms are due to cerebellar dysfunction. In visuomotor arm tracking experiments in primates, the Purkinje cells in the cerebellum played a role in the modulation of arm direction and speed [23]. In addition, fMRI revealed that, in humans, the superomedial and anterior cerebellum have corresponding control function in the speed of execution movements [24]. Therefore, it is hypothesized that patients with ET have poorer adaptation in their tracking speed than patients with PD due to dysfunction of their cerebellum. Since patients with PD do not have cerebellar degeneration, they can rely on external visual guidance for velocity compensation. Besides, it was noteworthy that the velocity control performance was not correlated well to tremor severity in ET patients, either correlated to motor symptoms severity in PD patients. Therefore, it presumed that velocity tracking performance could be essential characteristics represented to ET patients only, and independent to the tremor pathophysiological circuit or movement generating pathway. Therefore, it presumed that dysfunction in velocity tracking performance could be essential characteristics represented more to ET patients, but little in PD patients. Seemly, this characteristic is an independent factor to ET tremor generation, because it would not aggravate when tremor severity worsens. Although FTM scale for ET patients and UPDRS Part III scores for PD patients were divergent to each other, they covered major aspects of the disease conditions. Future studies could establish correlations between the following performance and disease severity with more comprehensive clinical scales. Previous visuomotor feedback studies have mostly examined the error rates in path tracing or changes in tremors to explain differences between PD and ET. However, we used T1 and T2 to examine changes in tracking velocity, velocity distribution performance, and differences in tracking velocity indicators, and found that velocity control itself differs between the two diseases. Therefore, this may be used as an action parameter to distinguish between ET and PD, particularly at the early disease stage.

Conclusions

Guiding point tests were used to compare the effects of visual feedback on spiral drawing in patients with PD and ET. Patients with PD had slower static drawing velocity due to sensorimotor integration abnormalities that can be compensated using dynamic visual feedback, which increased the velocity. The performance of patients with ET, however, was different from those with PD. This could be attributed to poorer eye-hand coordination and an inability to follow the guiding point. This phenomenon did not have a direct correlation with tremors. The method of the current study can be applied to analyze temporal tremor parameters and compensate for limitations in previous studies in observing spatial changes in tremors. This, in turn, could assist clinicians in distinguishing between PD and ET, which have similar clinical tremor presentations.