Abstract
Radiology plays a key part in the investigation of non-accidental injury. Many normal variants and artefacts can simulate an abnormality associated with non-accidental injury. It is essential that radiologists reporting skeletal surveys in cases of suspected child abuse are aware of these. We present a pictorial essay to aid the reporting radiologist in the differentiation between normal variants or artefacts and true traumatic injury. We show plain film examples of potential pitfalls throughout the body.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
Non-accidental injury results from an abusive act by a parent or guardian perpetrated on a child. Approximately 1.2% of people experience severe physical violence at the hands of an adult during childhood [1]. The thorough investigation of children with suspected non-accidental injury is critical to ensure the child’s safety [2]. After soft-tissue bruising and burns, fractures are the most common presentation of non-accidental injury [3]. As a result, radiologists play a key role in the diagnosis of suspected cases. If non-accidental injury is suspected, a skeletal survey is the imaging method of choice and a specific set of radiographs is recommended (Table 1) [2].
Several normal anatomical variants can simulate an abnormality associated with non-accidental injury. These include physiological periosteal reaction, metaphyseal variants and nutrient foramina. Also, various artefacts can be confused for traumatic injury. It is important that the reporting radiologist be aware of these normal variants and commonly seen artefacts so that traumatic injury and, more important, non-accidental injury are not over-reported. These appearances are presented below in a craniocaudal approach.
Skull
Sutures
Skull sutures can be mistaken for acute traumatic injury on plain film. The parietal and occipital bones are common sites for accessory sutures because of their numerous ossification centres [4]. Accessory sutures are often bilateral and symmetrical and are more numerous in the occipital region because of increased numbers of ossification centres. Features on plain film that suggest a suture rather than a fracture include a zig-zag pattern with sclerotic borders, an absence of any diastasis, bilateral and fairly symmetrical orientation and an absence of any adjacent soft-tissue swelling [4]. Fractures, on the other hand, are typically unilateral, sharp lucencies with non-sclerotic edges. They tend to widen as they approach a suture and sometimes cross the suture. There is usually associated soft-tissue swelling. Any doubt about whether the abnormality represents a fracture or a suture can be resolved by correlating radiographs with the bone windows on the CT head examination.
Occipital synchondrosis
The occipital bone at birth is composed of a squamous segment (composed of an interparietal portion and supraoccipital portion), a basioccipital segment and paired exoccipital segments [5]. The occipital synchondrosis occurs between the exoccipital segments and supraoccipital portion of the occipital bone (Figs. 1 and 2). It begins to fuse from birth, being complete by 4 years of age. It should not be mistaken for a fracture.
Mendosal suture
The mendosal suture separates the supraoccipital portion of the squamous segment of the occipital bone from the interparietal portion [6]. It usually persists for several weeks after birth [7, 8]. It can often simulate a fracture, especially if there is a slightly oblique projection of the skull (Fig. 3).
Metopic suture
The metopic, or frontal, suture is found in the midline of the frontal bone. Fusion usually occurs by approximately 9 months although some studies have shown it to persist in children up to 8 years of age and even into adulthood [9–11]. Like the mendosal suture, it can mimic a fracture if the child is rotated (Fig. 4).
Inca bone
The interparietal portion of the occipital bone develops from three or four pairs of ossification centres [12]. Failure to fuse leads to formation of the Inca bone (also known as the interparietal bone or os interparietale) (Fig. 5) [12, 13]. This Inca bone may be subdivided by both longitudinal and transverse sutures to form a bipartite, tripartite or multipartite Inca bone.
Wormian bones
Wormian bones are small intra-sutural bones that can be found in up to 53% of children (Fig. 6) [14]. The lambdoid suture is the most common location. Wormian bones arise less commonly in relation to the coronal and sagittal sutures. They are thought to result from mechanical factors that spread sutures apart [15]. Usually a number greater than ten is considered abnormal, as is a size greater than 6 × 4 mm, and conditions such as osteogenesis imperfecta should be considered in these children [16].
Cephalohaematoma
A cephalohaematoma is a traumatic subperiosteal haematoma of the skull [17]. Although they can occur in non-accidental injury, they usually occur post-instrumental delivery, more commonly with vacuum delivery than forceps [18]. It is important, therefore, to get appropriate clinical information about the method of delivery so not to misdiagnosis a birth-related injury. A cephalohaematoma is seen as a soft-tissue density overlying the skull and is restricted by the periosteum and sutures (Fig. 6). It therefore cannot cross the midline. Older haematomas can calcify peripherally. A cephalohaematoma normally resolves spontaneously.
Hair artefact
In an older child hair artefact can cause interpretation issues on radiographs, usually chest radiographs, where it can mimic surgical emphysema. On skull radiographs it can appear as linear high-density structures that might confuse the reporting radiologist, masking or mimicking a bony injury (Fig. 7).
Spine
Cervical spine
Cervical spine injuries in children younger than 8 years are uncommon but when they do occur, they usually occur in the upper cervical spine from the level of the occiput to C3 [19, 20]. Cervical spine injury is reported in cases of non-accidental injury [21–23]; however several normal variants can be mistaken for traumatic pathology.
Ossification centres
The normal odontoid ossification centre should not be mistaken for a traumatic injury. The basilar odontoid synchondrosis is thought to close between 3 years and 6 years of age but can persist in older children (Figs. 8 and 9) [24, 25].
Wedging of the cervical vertebrae
Anterior wedging of up to 3 mm can be seen in the cervical vertebrae and should not be mistaken for vertebral body compression fractures [19, 24, 26]. This wedging is often most prominent at the level of C3 (Fig. 9). Wedging becomes less apparent with increased age, with the vertebral bodies taking on a more rectangular appearance.
Pseudosubluxation
In almost half of children younger than 8 years there is subluxation at the C2-C3 level [19, 24, 27, 28]. However this normal physiological displacement can be differentiated from traumatic injury by appreciating the posterior cervical line [24]. This line is drawn from the anterior aspect of the C1 spinous process to the anterior aspect of the C3 spinous process. The anterior aspect of the C1, C2 and C3 spinous processes should line up within 1 mm of one another on flexion and extension views (Fig. 10).
Thoracic spine
In the newborn, the vertebral bodies are initially an oval shape [29]. Lucent notches are seen in the anterior and posterior margins of the vertebral bodies. These represent remnants of intersegmental clefts in the embryonic spine and contain nutrient canals through which arteries and veins enter the vertebral body (Fig. 11).
Lumbar spine
The neurocentral synchondroses are bilateral cartilaginous growth plates between the single anterior and the bilateral neural arch ossification centres [29]. These allow the vertebral arch to grow. On a lateral lumbar radiograph the unfused neurocentral synchondroses can be seen as lucencies between the ossified vertebral body and the ossified neural arch (Fig. 11).
Appendicular skeleton
Physiological periosteal reaction
This is a commonly seen normal variant found in the long bones of infants, most frequently at 2–3 months, although it is reported at 1–4 months [30]. The most common sites for physiological periosteal reaction are the tibia, femur, humerus, ulna and radius, and it can be unilateral or bilateral (Figs. 12 and 13) [30]. The usual appearance is a single layer of thin, smooth periosteal reaction less than 2 mm affecting one aspect of the long bones. This runs parallel to the underlying normal cortex along the diaphysis, separated from the underlying cortex by a radiolucent zone. The exact mechanism for this is uncertain, but it is thought to relate to the rapid growth of the infant and the loosely adherent periosteum [30, 31]. When physiological periosteal reaction has a thickness greater than 2 mm or occurs in infants older than 4 months, an underlying pathology such as non-accidental injury should be considered.
Normal metaphyseal variants versus classic metaphyseal lesion
The classic metaphyseal lesion, or bucket-handle fracture, is a type of fracture very specific for non-accidental injury (Fig. 14) [32, 33]. It occurs when a torsional force is applied to the immature bone adjacent to a growth plate (physis). The fracture extends transversely across the metaphysis and is thicker peripherally than in the centre. Its appearance varies according to the position of the limb to the radiograph. The corner fragments are the parts of the handle seen when the remainder of the fracture is hidden because of projectional factors (Fig. 14). Common sites include the proximal tibia, the distal tibia and the proximal humerus [34–36]. Acute fractures can be easily missed because they are not always identified radiographically. Because healing fractures are more easily visualised, these classic metaphyseal fractures are better seen on follow-up radiographs [34–36].
Several metaphyseal variants should not be confused with the classic metaphyseal lesion [37]. When differentiating between normal variants and pathology, AP and lateral coned views taken tangentially to the metaphyses are beneficial (Fig. 15) [2]. As mentioned above, follow-up radiographs are beneficial if there remains clinical concern. Although an isotope bone scan is less sensitive than plain film in the detection of metaphyseal fractures, it can be used to identify the more difficult to visualise acute fractures because they are positive within 7 h of a bone injury [2, 38].
Metaphyseal step-off
The area of the distal metaphysis adjacent to the physis is called the metaphyseal collar and is 1–3 mm in children as old as 7 years [39]. The metaphyseal step-off is an acute (nearly 90-degree) angle between the metaphyseal collar and the curvilinear metaphysis (Figs. 12, 16 and 17). The adjacent cortical margin might be indistinct but the trabecular pattern is maintained. This is usually seen in the long bones near the knee and wrist [37, 39].
Metaphyseal beak
A metaphyseal beak is a medial projection off the proximal humerus or proximal tibia (Figs. 12 and 18). It is well-defined and often dense, although the beak occurring at the proximal tibia is often less distinct [37].
Metaphyseal spur
This is a discrete longitudinal projection of bone continuous with the cortex that extends beyond the metaphyseal margin (Figs. 19 and 20). The most common sites are the lateral aspect of the distal femur, the lateral aspect of the distal radius, the medial aspect of the distal ulna and the metacarpals and metatarsals [37].
Distal ulna metaphyseal cupping
Cupping of the distal ulna is a known anatomical variant and the presence of this does not necessarily indicate underlying rickets if it is the only finding [40–42] (Fig. 21).
Metaphyseal fragmentation
Metaphyseal fragmentation can be seen in up to 11% of children [43]. Most commonly found in the distal femur and proximal tibia, it can occur unilaterally or bilaterally (Fig. 22). The bony fragments vary in size and shape but are usually elongated along the long axis of the adjacent bone, extending proximally along the metaphyseal margin. Metaphyseal fragmentation tends to occur in children 15 months and older, later than the age when a classic metaphyseal lesion would be suspected.
Proximal tibial cortical irregularity
This occurs at the medial aspect of the proximal tibial metadiaphysis and is seen as a focal area of irregularity in the cortex (Fig. 23) [37, 44]. There may be associated physiological periosteal reaction and it is seen bilaterally in 25% of cases [37, 44]. It should not be confused with a buckle fracture. If there is any doubt, further imaging such as follow-up radiographs, nuclear medicine bone scintigraphy or MRI should be performed.
Nutrient vessels
A long bone is supplied by a nutrient artery that enters the bone at an oblique angle through the nutrient foramen. This is directed away from the growing end of the bone. Nutrient vessels as they pass through the cortex of a long bone shaft can be mistaken for oblique fractures (Figs. 24 and 25) [45].
Skin folds
Skin folds are frequently seen on radiographs and the lucent line made by the fold can easily be dismissed as unrelated to underlying bone (Fig. 26). Occasionally a fold simulates a fracture and in the case of a skeletal survey for non-accidental injury, it causes the reporting radiologist to misinterpret it as suspicious (Fig. 26).
Intraosseous cannulae
Artefacts related to resuscitation are also seen on subsequent skeletal surveys. Intraosseous cannulae are used for paediatric patients who need rapid fluids or medications where intravenous access is limited. Unless the reporting radiologist has been told of a history of such use, the site of bone puncture might be mistaken for non-resuscitation-related injury (Fig. 27).
Conclusion
Non-accidental injury is a condition in which the radiologist plays a key role. It is essential that the reporting radiologist can differentiate between normal variants or artefacts and true traumatic injuries. Awareness of these possible pitfalls helps radiologists avoid missing a diagnosis of non-accidental injury or erroneously deeming a normal finding pathological.
References
Radford L, Corral S, Bradley C et al (2011) Child abuse and neglect in the UK today. http://www.nspcc.org.uk/Inform/research/findings/child_abuse_neglect_research_PDF_wdf84181.pdf (Accessed 18 May 2013)
Royal College of Radiologists (2008) Standards for radiological investigations of suspected non-accidental injury. http://www.rcr.ac.uk/docs/radiology/pdf/RCPCH_RCR_final.pdf (Accessed 18 May 2013)
Offiah A, van Rijn RR, Perez-Rosello JM et al (2009) Skeletal imaging of child abuse (non-accidental injury). Pediatr Radiol 39:461–470
Sanchez T, Stewart D, Walvick M et al (2010) Skull fracture vs. accessory sutures: how can we tell the difference? Emerg Radiol 17:413–418
Nakahara K, Utsuki S, Shimizu S et al (2006) Age dependence of fusion of primary occipital sutures: a radiographic study. Childs Nerv Syst 22:1457–1459
Miller AJ, Kim U, Carrasco E (2010) Differentiating a mendosal suture from a skull fracture. J Pediatr 157:691
Caffey J (1978) Pediatric x-ray diagnosis, vol 1, 7th edn. Year Book Medical Publishing, Chicago
Nayak SR, Krishnamurthy A, Madhan Kumar SJ et al (2007) The mendosal suture of the occipital bone: occurrence in Indian population, embryology and clinical significance. Surg Radiol Anat 29:329–332
Vu HL, Panchal J, Parker E et al (2001) The timing of physiologic closure of the metopic suture: a review of 159 patients using reconstructed 3D CT scans of the craniofacial region. J Craniofac Surg 12:527–532
Mathijissen IM, Vaadrager JM, can der Meulen JC et al (1996) The role of bone centers in the pathogenesis of craniosynostosis: an embryologic approach using CT measurements in an isolated craniosynostosis and Apert and Crouzon syndromes. Plast Reconstr Surg 98:17–26
Bademci G, Kendi T, Agalar F (2007) Persistent metopic suture can mimic the skull fractures in the emergency setting? Neurocirugia 18:238–240
Matsumura G, Uchiumi T, Kida K et al (1993) Developmental studies on the interparietal part of the human occipital squama. J Anat 182:197–204
Hanihara T, Ishida H (2001) Os incae: variation in frequency in major human population groups. J Anat 198:137–152
Marti B, Sirinelli D, Maurin L et al (2013) Wormian bones in a general paediatric population. Diagn Interv Imaging 94:428–432
Sanchez-Lara PA, Graham JM Jr, Hing AV et al (2007) The morphogenesis of wormian bones: a study of craniosynostosis and purposeful cranial deformation. Am J Med Genet A 143:3243–3251
Cremin B, Goodman H, Spranger J et al (1982) Wormian bones in osteogenesis imperfecta and other disorders. Skelet Radiol 8:35–38
Glass RB, Fernbach SK, Norton KI et al (2004) The infant skull: a vault of information. Radiographics 24:507–522
Johanson RB, Menon BK (2000) Vacuum extraction versus forceps for assisted vaginal delivery. Cochrane Database Syst Rev 2, CD000224
Khanna G, El-Khoury GY (2007) Imaging of cervical spine injuries of childhood. Skelet Radiol 36:477–494
Hill SA, Miller CA, Kosnik EJ et al (1984) Pediatric neck injuries. A clinical study. J Neurosurg 60:700–706
Swischuk LE (1969) Spine and spinal cord trauma in the battered child syndrome. Radiology 92:733–738
Rooks VJ, Sisler C, Burton B (1998) Cervical spine injury in child abuse: report of two cases. Pediatr Radiol 28:193–195
Brown RL, Brunn MA, Garcia VF (2001) Cervical spine injuries in children: a review of 103 patients treated consecutively at a level 1 pediatric trauma center. J Pediatr Surg 36:1107–1114
Lustrin ES, Karakas SP, Ortiz AO et al (2003) Pediatric cervical spine: normal anatomy, variants, and trauma. Radiographics 23:539–560
Ogden JA (1984) Radiology of postnatal skeletal development. XII. The second cervical vertebra. Skelet Radiol 12:169–177
Swischuk LE, Swischuk PN, John SD (1993) Wedging of C-3 in infants and children: usually a normal finding and not a fracture. Radiology 188:523–526
Cattell HS, Filtzer DL (1965) Pseudosubluxation and other normal variations in the cervical spine in children. A study of 160 children. J Bone Joint Surg Am 47:1295–1309
Swischuk LE (1977) Anterior displacement of C2 in children: physiologic or pathologic. Radiology 122:759–763
Freyschmidt J, Brossmann J, Wiens J et al (2003) Borderlands of normal and early pathological findings in skeletal radiography, 5th edn. Thieme, Stuttgart
Kwon DS, Spevak MR, Fletcher K et al (2002) Physiologic subperiosteal new bone formation: prevalence, distribution, and thickness in neonates and infants. AJR Am J Roentgenol 179:985–988
Rana RS, Wu JS, Eisenberg RL (2009) Periosteal reaction. AJR Am J Roentgenol 193:W259–W272
Kleinman PK, Marks SC Jr, Richmond JM et al (1995) Inflicted skeletal injury: a postmortem radiologic-histopathologic study in 31 infants. AJR Am J Roentgenol 165:647–650
Kleinman PK (2008) Problems in the diagnosis of metaphyseal fractures. Pediatr Radiol 38:S388–S394
Kleinman PK, Marks SC Jr (1996) A regional approach to the classic metaphyseal lesion in abused infants: the proximal tibia. AJR Am J Roentgenol 166:421–426
Kleinman PK, Marks SC Jr (1996) A regional approach to classic metaphyseal lesion in abused infants: the distal tibia. AJR Am J Roentgenol 166:1207–1212
Kleinman PK, Marks SC Jr (1996) A regional approach to the classic metaphyseal lesion in abused infants: the proximal humerus. AJR Am J Roentgenol 167:1399–1403
Kleinman PK, Belanger PL, Karellas A et al (1991) Normal metaphyseal radiologic variants not to be confused with findings of infant abuse. AJR Am J Roentgenol 156:781–783
Sty JR, Starshak RJ (1983) The role of bone scintigraphy in the evaluation of the suspected abused child. Radiology 146:369–375
Oestreich AE, Ahmad BS (1992) The periphysis and its effect on the metaphysis: I. Definition and normal radiographic pattern. Skelet Radiol 21:283–286
Glaser K (1949) Double contour, cupping and spurring in roentgenograms of long bones in infants. AJR Am J Roentgenol Radium Ther 61:482–492
Dwek JR (2011) The radiographic approach to child abuse. Clin Orthop Relat Res 469:776–789
Slovis T, Chapman S (2008) Evaluating the data concerning vitamin D insufficiency/deficiency and child abuse. Pediatr Radiol 38:1221–1224
Kleinman PK, Sarwar ZU, Newton AW et al (2009) Metaphyseal fragmentation with physiologic bowing: a finding not to be confused with the classic metaphyseal lesion. AJR Am J Roentgenol 192:1266–1268
Williams H (2008) Normal anatomical variants and other mimics of skeletal trauma. In: Johnson KJ, Bache E (eds) Imaging in pediatric skeletal trauma. Springer, Berlin, pp 91–118
Keats TE, Anderson MW (2012) Atlas of normal roentgen variants that may simulate disease, 9th edn. Elsevier Saunders, Philadelphia
Conflicts of interest
None.
Author information
Authors and Affiliations
Corresponding author
Additional information
CME activity
This article has been selected as the CME activity for the current month. Please visit the SPR Web site at www.pedrad.org on the Education page and follow the instructions to complete this CME activity.
Rights and permissions
About this article
Cite this article
Quigley, A.J., Stafrace, S. Skeletal survey normal variants, artefacts and commonly misinterpreted findings not to be confused with non-accidental injury. Pediatr Radiol 44, 82–93 (2014). https://doi.org/10.1007/s00247-013-2802-2
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00247-013-2802-2