THE ANATOMICAL RECORD 298:1099–1110 (2015)
A Critical Look at Mummy CT Scanning SAMANTHA L. COX1,2* Physical Anthropology Section of the Penn Museum, University of Pennsylvania, Pennsylvania, USA 2 Department of Archaeology and Anthropology, Division of Archaeology, University of Cambridge, Cambridge, UK 1
ABSTRACT Computed tomography scanning of mummies has been conducted for almost 40 years, and has become an increasingly popular method of mummy study in the 21st century. However, most CT scan analyses published today still do little more than praise the technique’s nondestructive, non-invasive properties. Despite the wealth of information contained within a modern, high definition scan, most researchers have yet to fully exploit the full potential of this technology. In addition, those that have utilized CT scanning in ancient remains continue to interpret mummified tissues as though they were living, without investigating how taphonomy and mummification could have effected the tissue images that are produced. Because of this, there is very limited information available for clear interpretation of mummy CT’s. This article presents a critical assessment of the development of mummy CT scanning and presents the results of two Egyptian mummies CT’ed at the Penn Museum as an example of the potentials and pitfalls of high-resolution scanning. Anat C 2015 Wiley Periodicals, Inc. Rec, 298:1099–1110, 2015. V
Key words: mummy; CT scan; paleopathology
For nearly four decades, computed tomography (CT) scanning has been employed in the scientific study of mummified remains. However, despite the numerous mummies that have been examined, we know almost nothing about them beyond the process of their mummification. More than that, the formula of most CT scan studies of mummies has remained almost unchanged since they first began in 1977. The vast majority of articles are case reports that come to the same conclusion: CT scans are a powerful, nondestructive, noninvasive way to study mummies. Though this is true, much time has passed and the technology continues to improve and evolve, but its potential for mummy research has never been fully exploited. While CT scan studies have become increasingly prevalent over the last 10 years, very little research has been done to explore the limitations of the technology or the differences between living and long-dead tissues in imaging. This article takes a critical look at the history of mummy CT scanning to understand the current state of paleo-imaging and address the weaknesses of the field in order utilize CT technology to its fullest potential and encourage mummy studies to continue evolving. The CT scans of two Penn Museum mummies are presented here as an example of the wealth of information available from detailed CT scans, but also as an illustration of the limiC 2015 WILEY PERIODICALS, INC. V
tations imposed on CT analysis due to the dearth of information available on tissue changes in paleo-imaging.
REVIEW OF MUMMY CT HISTORY Computed tomography scanning was first developed in 1975 for medical imaging, and it was a mere 2 years later that the technology was applied to a mummy. Lewin and Harwood-Nash (1977) performed the first CT scan on an Egyptian mummy named Nakht from the Royal Ontario Museum in 1977 to examine the brain. Though the findings of this study were not remarkable, it was a mile-stone in mummy studies for the application of the new technology. Lewin (1978) also produced the first whole-body scan of a mummy in 1977. From 1977 until 1985, CT scans were not often used for analysis in
*Correspondence to: Division of Archaeology, Department of Archaeology and Anthropology, University of Cambridge, Downing Street, Cambridge CB23DZ, E-mail: [email protected] Received 16 January 2015; Accepted 30 January 2015. DOI 10.1002/ar.23149 Published online in Wiley Online Library (wileyonlinelibrary. com).
1100
COX
their own right, but to aid in directing mummy autopsy and tissue sampling. For example, Notman and his team (1985) on the Minnesota Mummy Project, Zimmerman (1996; note the study was carried out in 1983) on Alaskan mummies from St. Lawrence Island, and Ascenzi et al. (1996; note, the study was carried out in 1983) on the mummy of a girl from Rome. From 1985 to the present at least one mummy CT publication has come out every year, and they have become increasingly numerous through time. These articles consist mostly of case reports that generally describe mummification techniques, artifacts, preservation, and pathology (e.g., Marx and D’Auria, 1986, 1987; Magid et al., 1989; Pickering et al., 1990; Watson and Meyers, 1991; Gregorczyk et al., 1993; Hughes, 1993; Baldock et al., 1994; Arriaza, 1998; Gaafar et al., 1999; Macleod et al., 2000; Thekkaniyil et al., 2000; Mininberg, 2001; Hoffman et al., 2002; Previgliano et al., 2002; Shin et al., 2003; Kim et al., 2006, 2006; Lee et al., 2006; Hawass and Saleem, 2011; Nelson et al., 2011; Wade et al., 2012), though there are some publications that specifically address pathology, facial reconstructions, and new approaches and applications. Many of the early CT papers compare traditional X-rays to CT scans to determine the strengths and weaknesses of the new technology and establish it as a reliable method for investigation (for instance, Watson and Meyers, 1993, and Hunt and Hopper, 1996). These early studies are quite constrained by the low resolutions and large file sizes produced by early CT machines. It is not surprising that most of the mummies scanned are ancient Egyptian, and the extended positions of the bodies produce large scans that take up large quantities of file space. It is for this reason that some of these early scans report slice thicknesses of up to 24 mm (Marx and D’Auria, 1986), to keep the scanners themselves from overheating and file sizes to a minimum. However, while the volume of data is easier to manage, the scans could not have contained enough detail for any kind of in-depth analysis. It was the mid-1990’s before scanning resolutions were high enough to allow for the first CT investigations of mummy dentition (as in Nikol et al., 1995; Melcher et al., 1997). Also at this time, a study of a female South American mummy is noteworthy because of her probable tuberculosis. It is one of the first studies to identify a specific pathology via CT, as well as providing some of the oldest evidence of tuberculosis (Correal-Urrego and Florez, in Cockburn, 1998). At this point in the literature, CT mummy analysis seems to be progressing along at a steady pace; some authors began to experiment with utilizing CT scans in different ways, but with limited results (e.g., Brown and Wood, 1999). However, by the beginning of the new millennium, CT studies appear to stagnate. Despite continued advances in CT technology, the focus is still primarily on describing the mummies and the mummification process on an individual basis. Friedrich et al. (2007) does this with 12 Chachapoyan mummies, Gupta et al (2008) with an Egyptian head as does Chan et al. (2008). Jackowski et al. (2008) attempt to compare mummies from Ancient Egypt to those from Peru, but the article focuses heavily on Egyptian mummification, only briefly discussing Peru and the connection between the two is unclear. Collier (2009) reports on another mummy being CT scanned in Toronto, Gerloni (2009) again considers Egyptian dentition and Lynnerup (2009) applies these techniques to the Bog Bodies housed in the British Museum.
Though this is not an exhaustive list of publications by any means, the one thing that is consistent in all these studies is the use of higher-resolution CT scans, though often a lower resolution is used for the post-cranial elements than for the head. Teeter and Vannier (2009) published the CT scan of Meresamun for exhibit at the Oriental Institute in Chicago at a reconstructed slice thickness of 0.7 mm, claiming that this is the highest resolution mummy scan published to that date. There are surprisingly few pathology studies. Though this is what the technology was designed for in a clinical setting, it has rarely been applied to anthropological investigations. Most published work of this type is rather limited in scope, and the full potential of the technology seems to have not been adequately explored (for a full review and critique see Chhem, 2006). The first pathology-specific study with a CT scan was probably € the one performed on Otzi the Iceman (Holden, 2001), since this is likely the first time that scan resolutions were high enough to allow for this type of analysis. The most famous pathology study is probably that of the mummy of King Tutankhamun to attempt to determine the cause of death of the boy king (Hawass et al., 2009). In a few instances, researchers have attempted to use CT pathology analyses in more creative ways. For example, Ceruti (2004) looked at pathology to determine cause-of-death in Peruvian mummies and compared it to historical accounts of human sacrifice. In other cases, investigators have attempted to use CT for pathology analysis, but were unsuccessful in identifying any, as was the case for Bianucci et al. (2008). It is apparent after this review that there are some large gaps in knowledge and understanding of CT scanning in ancient remains. As of yet, the vast majority of articles published are merely individual case reports. Nothing has been done to take the data presented in those articles and synthesize it into a work that has meaning applicable to the field of anthropology. There is potential here to evaluate health in mummies, as a population, given how many have been CT scanned, but focus continues to remain on mummies as individuals rather than representative members of a group. The exception to this generalization is the Horus Mummy Project team. This group has evaluated the CT scans of over 20 mummies and found persuasive evidence of heart disease that would indicate some prevalence throughout the upper class population (Allam et al., 2009). This study is the only example of a synthetic analysis that finally takes CT scans beyond the “gee, wow!” factor and begins to explore the potential of this hi-tech method to add to our knowledge of ancient Egyptian life, questioning the assumptions that have always been made regarding health and diet of ancient people. Unfortunately, only a handful of researchers have ever considered and written about any of the negative aspects of CT scanning. There is little consistency in the manner in which findings are reported, for instance the specs for the CT machines and scanning parameters are often omitted. In 2001, Weber published a detailed critique of this issue in which he proposes a scheme of information that should be standardly reported in high technology academic articles. He is of the opinion that this kind of digital media should be made freely available for all scholars to consider, so that other evaluations and comparisons can be made. It is quite apparent that
A CRITICAL LOOK AT MUMMY CT SCANNING
few publications have adopted his standard, and, to my knowledge, no one has proposed another. Few share their scans; however, this may be starting to change with the University of Pennsylvania Museum’s Online Research Scan Archive (ORSA) and the IMPACT Radiological Mummy Database, which are free for scientists that wish to use them. While there are many wonderful and useful applications of CT scans, there is also a noticeable gap in the literature concerning the discussion of its weaknesses. For instance, few consider how mummification, especially artificial mummification, may alter the appearance of tissues. Chew et al (2006) noted in their cadaver study that modern embalming changes the way that structures appear radiologically in recently dead cadavers. The image quality was greatly reduced by beam hardening artifacts, attenuation between soft tissues was reduced as compared with living people, and other less uniform changes were visible due to the embalming process. If this is the case today, it could have critical implications for the interpretations of mummified tissues, as most scans of ancient remains are read and interpreted as though they were living people. Villa and Lynnerup (2012) have begun to tackle some of these concerns, but information is still needed before mummified remains can be reliably interpreted with medical imaging techniques. To date, there is no published data, of which I am aware, that investigates embalming materials, their appearance, and their effects on tissue appearance on CT. A good case example is the famous CT scan of Tutankhamun (Hawass et al., 2009). The debate over the whether or not the femoral fracture represents an ante- or postmortem break, and thus whether or not it could have contributed to events surrounding the boy king’s death, cannot be laid to rest because the research has not been done to be say how the resin in mummies behaves through time or whether or not mummification processes would eliminate any signs of hematoma. Though many great studies have emerged with this technology, there is still much more that has to be considered before the true meaning of what has been observed can be adequately assessed and synthesized. R€ uhli et al. (2004) is one of the very few that offer caution about the interpretation of CT scans. He argues that every radiologic examination of ancient human remains should include an experienced radiologist as well as an experienced physical anthropologist to interpret the images. The radiologist can recognize pathology more readily than the anthropologist, but the anthropologist will be more adept at recognizing confounding factors that might mar the diagnosis. Resins and other artifacts of mummification are the most important factors to recognize. There is no comprehensive guide to diagnosing paleopathology in radiology; however, Chhem and Brothwell published Paeloradiology: Imaging Mummies and Fossils (2008) which dedicates about half the book (100 pages) to this topic. Though it is by no means complete, and focuses mainly on X-rays it gives pictures and characteristics of the most common pathologies in both humans and animals and is at least a good starting reference for analyzing paleoradiography.
THE PENN MUMMIES When the CT scanning of two Egyptian mummies from the University of Pennsylvania Museum (Penn
1101
Museum) was undertaken, it was with the aim of setting these two individuals within their historical contexts. It was soon clear, however, that despite all the mummies that have been CT’ed in the world, there was not sufficient information available to soundly interpret the images or understand these two people as a part of Ancient Egyptian life. They are presented here as examples of the potential and limitations faced when interpreting CT scans of ancient human remains. PUM II and Hapi-Man, from the Penn Museum were CT scanned in April 2009, at the Hospital of the University of Pennsylvania (HUP). Dr. Morrie Kricun Emeritus Professor of Radiology at HUP and the University of Pennsylvania Medical School assisted with the interpretations of the mummy images. A spiral multislice Siemens’ SOMATOM Sensation 64-Slice Computed Tomography machine scanned the mummies with a slice thickness of 1 mm (0.5 mm reconstructed). To produce the best results, both mummies were scanned in three sections, from the head to mid-thorax, mid-thorax to mid-femur, and mid-femur through the feet, producing a total of about 2,000 images for each mummy. The analysis was done using the freeware program Osirix v. 3.7 with the 64-bit upgrade on an Apple Mac Pro. Hapi-Man was excavated in Abydos in 1902 by the famous archaeologist Sir Flinders Petrie. He is thought to date from the Late Period and be about 35–45 years of age at death (Lawrence, 1980). Hapi-Man has been xrayed on two occasions during his stay at the Penn Museum and research in the Penn Museum Archives turned up a set of apparently unpublished X-rays from April 1932, by Dr. J.G. Cohen. A second envelope of completely unmarked films was also found in archives and matched to Hapi-Man based on the placement of the amulets around the neck. The date and researcher for this second set of X-rays is unknown, but it is clear that they are more recent than the first set and it is possible that they were produced in the late 1970s for the Penn Museum exhibit The Egyptian Mummy: Secrets and Science. The CT scan of Hapi-Man revealed information about his mummification and identified some possible pathology. It is clear on the scans that his abdomen was stuffed with rolled bundles of cloth that might contain the internal organs as is typical of Late Period mummification (David, 2008); however, there does not appear to be anything inside and it is likely that these are probably symbolic only (Fleming, 1980). Hapi-Man also had his brain removed through the nose and the body is completely covered in resin which lies in a pool in the cranium, thorax and abdomen. It is clear in the CT that he has an udjet eye on his forehead, no fewer than twelve amulets arranged around his neck, like a necklace, including beads, heart amulets, an udjet eye, goddess figurine, Horus falcon, scarab, and Amun feather. The left hand has an unidentifiable square amulet on the ring finger, though there does not appear to be anything attaching it to the finger (Fig. 1). On the right rib cage there are three rows of amulets, the first is three scarab beetles, the second is four djed pillars, and the third is five tyet knots with a goddess figurine, probably of Isis (Fig. 2). The placement of these kinds of amulets in the wrappings of the deceased was quite common, but it is clear from the scans that none of these objects are likely metallic, most are probably fiance (Andrews, 1994).
1102
COX
Fig. 1. Hapi-Man amulets, (top left) goddess figure, Horus falcon, and square “ring” on hand; (top right) jars, udjet eye, scaraab, beads; (bottom left) Amun feather; (bottom right) udjet eye on forehead.
The teeth are quite clear, and we can confirm the age of the individual to be between 35 and 45 years of age using cranial suture closure patterns (Walker in Buikstra and Ubelaker, 1994). There is clear evidence of uncovertebral arthropathy as well a rib fracture. The scans do show that there are numerous air spaces in the vertebral bodies that are not present in living individuals; it is unclear if this has to do with the age of the specimen or if it is pathological. There is also a uniform distribution of some substance around the periphery of the annulus of the thoracic and lumbar vertebrae which could either be attributed to the resin or pathology, likely only an autopsy would be able to tell for certain.
However, evidence of discontinuous calcification of the annulus at every disc level and clear osteoarthritis between the posterior elements of the L4 and L5 is present. The fifth lumbar vertebra is sacralized as well as fused bilaterally to the sacrum on the caudal side. This might have marginally effected the mobility of his lumbar spine but would not have been crippling. Osteoarthritis was observed on both hips, but it was apparent on the scan that there was no bone reaction to the cartilage loss, suggesting that this may be a result of specimen age or mummification process instead of disease. In the feet, there were no bone reactions and only some medial narrowing of the first metatarsophalangeal
A CRITICAL LOOK AT MUMMY CT SCANNING
joints, probably not enough to be considered arthritic, but it is a possibility. This may have only been the very beginnings of arthritic changes, just an unusually narrow joint space, or maybe an artifact of mummification. It appears on the scan that Hapi-Man’s feet became detached from his body at some point after his mummification. This is most visible in the very large tibiotarsal joint space but upon close inspection the wrappings are
Fig. 2. (left) Right rib cage, top row- three scarab beetles, middle row- four djed pillars, bottom row- five tyet knots and Isis figurine, (right) scarab beetle on left rib cage.
1103
torn through in the same plane. There are bright spots on the scan inside the joint space and on the wrappings that are hypothesized to be some kind of adhesive or restorative material to reconnect the feet to the body. It is unclear when or how this damage occurred or was repaired. Like many Egyptian mummies, PUM II (an acronym for Pennsylvania University Museum II) has a long and rather cloudy history. The area of Egypt from which he comes is unknown and he is first recorded in the early twentieth century on a ship to Philadelphia. He came to America by way of John T. Morris, a wealthy Philadelphian who then donated him to the Philadelphia Museum of Art (PMA). Though the sarcophagus is highly decorated, the name of the individual was not recorded on it nor was any clues about his life. He remained at the PMA until the end of January 1973 when he was shipped to Michigan to be autopsied on February 1 at the Wayne State University Medical School in Detroit as part of a symposium there entitled Death and Disease in Ancient Egypt. Known as PUM II because he was the second mummy to be acquired by the institution, radiographic and xerographic exams were done at the Mt. Carmel Mercy and Hutzel hospitals both before and after the autopsy (Cockburn, 1998). Following the autopsy, he was sent to be on exhibit at the Museum of Natural History in Washington DC, on loan to the Smithsonian Institution (Cockburn 1975). He is now on display and has been a part of the Egyptian Mummies: Secrets and Science exhibit since it opened in 1980 at the Penn Museum. The right toes of PUM II are flexed, pulled back towards the top of the foot. Though the X-rays and autopsy report of PUM II dismissed the unusual appearance of the toes as an effect of tight wrapping, CT scans show that it is probably a condition known as hammertoe that is resultant of a cavus deformity of the right
Fig. 3. Measured angle of cavus foot deformity, 45 is considered normal and 90 is severely cavus, here the angle on the right foot of PUM II measures about 100 (image captured using Osirix v. 3.7).
1104
COX
Fig. 4. (left) Axial slice illustrating the difference in density between the two feet (image captured with Osirix 3.7), (right) 3D reconstruction of feet colored by density, red is high density, yellow is medium, and green is low red is high density, yellow is medium, and green is low
Fig. 5. Arrow indicates small bunion on the right medial proximal first metatarsal of PUM II.
foot. Aminian and Sangeorzan (2008) define cavus foot deformity as describing “a spectrum of foot shapes that have a high arch” and is usually characterized by a high pitch angle of the calcaneus and plantar flexion of the
midfoot. Radiographically, a foot can be called a cavus foot with an increased Hibbs angle, measured through the axis of the calcaneus and metatarsal. A normal foot is around 45 while cavus can be 90 ; the angle on PUM
A CRITICAL LOOK AT MUMMY CT SCANNING
Fig. 6. Orange arrow shows large air pocket between S1 and L5 would be indication of a hernia in living people but is inconclusive here, blue arrows point to narrowing of the hip joints that could be mild arthritis or an artifact of mummification in PUM II.
Fig. 7. What appears to be DISH in PUM II may just be residual soft tissues and resin.
1105
1106
COX
Fig. 8. Mild hypertrophy of the lischka joints with Schmorl’s nodes on the caudal C3 (top) and cranial C4 mildly encroaching on the neural foramen (bottom) in PUM II.
II’s foot is about 100 (Fig. 3). The bones of the foot are formed normally and well articulated, indicating that this deformation occurred after the bones matured. Today, two-thirds of modern cases of cavus foot have an underlying neurological condition (Aminian and Sangeorzan, 2008). Other causes can be muscle spasticity leading to muscle imbalance in the foot. Based on the periosteal reaction described in the autopsy, it might be possible that a chronic infection, perhaps due to parsitic infestation as those authors suggest, could have led to muscle imbalance in that leg and caused the cavus foot. The left foot has a normal shape, but it is significantly less radiolucent than the right foot and the soft tissues
appear atrophied as compared to the other foot. The authors of the autopsy noted that the right foot appeared swollen, so perhaps what we interpreted as atrophy is actually normal and the other leg is possibly inflamed (Fig. 4). However, since the body was dehydrated with natron it is necessary to consider whether swellings in a limb would be been preserved. The coloring on the scan might also suggest that for some reason the right leg is full of resin or fat and the left is not, accounting for the difference in density. A deep subarticular cyst is visible in the center of the left and right tali as are small cysts in the MT 1 and 2 on the left foot. The medial aspect of the MT1 seems to have a small
A CRITICAL LOOK AT MUMMY CT SCANNING
1107
Fig. 9. Orange arrow indicates nasal bones that could show healed fracture, blue arrow shows where the ethmoid was broken to remove the brain of PUM II.
Fig. 10. The cloudy appearance of the right mastoid of PUM II as compared to the left could show chronic mastoiditis but could also be the air spaces filled with resin and fat.
bunion forming, perhaps as a result of the gait caused by the cavus foot (Fig. 5). The talus and tibia appear to be starting to fuse together on the lateral half to two-
thirds of the left tibiotarsal joint, perhaps be the result of some kind of trauma (Fig. 20). The knees show subchondral cysts, the beginnings of osteoarthritis,
1108
COX
Fig. 11. Green lines show the abnormal A/P measurements (ca. 25 mm at the maximum) of the sella turcica probably indicative of a tumor of the pituitary gland in PUM II.
especially in the right knee (Fig. 21). It is quite clear in Fig. 6 that the legs of PUM II are different lengths, but after careful consideration, it was decided that this is because his pelvis was not completely level when mummified, making his legs appear asymmetric; it should be noted that this kind of asymmetry is often also seen with cavus foot (Aminian and Sangeorzan, 2008). A bone island is visible in the right ilium and narrowing of the hip joints suggest mild arthritis, though it could also be an effect of mummification. The vertebral disk between the L5 and S1 is in good shape but a large air pocket in the center would indicate a hernia in a living person, and this cannot be said conclusively in the case of ancient remains (Fig. 6). Though the autopsy reported a sixth lumbar body on the X-rays, we found no indications of this; however, the vertebral bodies were removed from this area during autopsy to look for spinal cord samples, so this could be affecting the interpretation. Most of the vertebral disks are in good condition with no degenerative changes but Schmorl’s nodes are present on the caudal endplate of T11. A build-up of some substance on the thoracic vertebral bodies was originally identified as diffuse idiopathic skeletal hypertrophy (DISH) (Fig. 7). However, upon further investigation, this was not visible on the cervical spine and in Dr. Kricun’s experience, it would be incredibly unusual to suffer from DISH in the thorax without involvement of the cervical vertebrae. We finally decided that it is more likely to be resin-covered soft tissues adhering to the spine. Small osteophytes are visible on some of the vertebral bodies and those with the Schmorl’s nodes indicate a mild osteoarthritis. There appears to be a mild bilateral hypertrophy of the uncovertebral lishka joints
between the C3 and C4. An osteophyte is visible on the posterior caudal end plate of the C4 and the caudal facet of C4 also has an osteophytic cyst. There is mild encroachment on the neural foramen in this area with an osteophytic ridge and large Schmorl’s nodes on the anterior caudal aspect of C3 and anterior cranial aspect of C4 (Fig. 8). The autopsy reports did not give an estimated age-atdeath and most of his cranial sutures were removed during autopsy. However, the condition of his teeth and joints looks quite similar to Hapi-Man so we decided that it is probably reasonable to estimate that he is about the same age, 35–45 years. PUM II’s nose exhibits an unusual curve at the end that is not consistent with the general shape, perhaps the result of a healed fracture. The ethmoid has clearly been broken through for access to the cranium and removal of the brain (Fig. 9). The mastoid bone appears to be cloudy on the right side, indicative of chronic mastoiditis, but this could be also be another example of the fat and resins filling the air spaces (Fig. 10). Measurements indicate an abnormally large sella turcica, the maximum anteroposterior (A/P) dimension is about 25 mm and maximum depth is 14 mm (Fig. 11). The normal dimensions, given as a range by Jones et al. (2005) are up to 16 mm A/P and 12-mm deep. Though the depth measurement is not far from what is considered to be normal, the A/P is much larger than the given range and suggests the presence of a pituitary tumor. Pituitary tumors cause a range of different effects but the most common are hormonal imbalances, dizziness, disturbances of consciousness, and blindness (Wakai, 1981). The tumor is of significant enough size that it could have been the cause of death
A CRITICAL LOOK AT MUMMY CT SCANNING
for PUM II, but if not, it was at least a source of great discomfort and illness.
1109
(2006) concerning CT analyses on medical school cadavers serves as a warning, and we need to know more about these changes before CT scans of ancient remains can yield any meaningful results.
DISCUSSION As is always the case with studying ancient remains, there are many confounding factors at work, including preservation and taphonomic processes. There are many instances in the cases of these two mummies when a conclusion could not be reached because of insufficient data on the imaging of ancient tissues. However, the detailed scan resolution allowed for very specific analysis of minor changes, some of which can be confirmed as probable arthritis and others which are less clear as they could be artifacts of mummification. Research in the future needs to focus on sorting out the differences between ante- and postmortem changes on these kinds of medical images as a means of eventually being able to tease out the most accurate results. If there is more data available on this in the future, we could, perhaps, be able to ascertain if thin layers of resin in the joints give the appearance of arthritis when condensed on X-ray film but would then be absent on CT scans. As it stands, the amount of information on these processes is limited and means that the data cannot be properly interpreted except for large, severe pathologies. There is a distinct lack of data that has been synthesized between mummies, but discrepancies in scan reporting means that published studies cannot be compared or aggregated because of differences in scan specifications. Even if it were possible to compare all the mummies that have been scanned thus far, it is quite likely that a lack of understanding of the impacts of taphonomy and mummification on the radiologic appearance of tissues means that many published findings are likely to have been misinterpreted. It is clear from Hapi-Man and PUM II that there are some common factors, arthritis and dental pathology being the main ones, but two individuals are not enough to draw conclusions about a population. The cavus foot deformity and the enlargement of the sella turcica in PUM II are interesting pathologies to note, but they have little meaning beyond this one individual unless other examples can be found in other mummies. The smaller pathologies like arthritis that were observed numerous times in this analysis are ones that we know will translate more between individuals; however, these cannot yet be positively identified and interpreted until data is compiled to distinguish taphonomic changes from antemortem degradation. At this point, it is probable that over one hundred mummies have been CT scanned and reported on. None of this data has yet been synthesized into a work that is meaningful beyond the scope of more than the one individual scanned. Because data has been collected on a number of mummies, the next logical step would be to collate the data and see the conclusions that can be drawn about these mummies as a population rather than just as single people. However, this study shows that, despite the plethora of mummies imaged, the data to interpret scans more critically does not yet exist. This casts some doubt on the conclusions that have been drawn in these numerous investigations since we cannot reliably recognize taphonomy and mummification radiologically. The medical research published by Chew et al.
LITERATURE CITED Allam, AH, Thompson RC, Wann LS, Miyamoto MI, Thomas GS, 2009. Computed tomographic assessment of atherosclerosis in ancient Egyptian mummies. JAMA 302(19):2091–2094. Aminian A, Sangeorzan BJ. 2008. The anatomy of cavus foot deformity. Foot Ankle Clin N Am 13:191–198. Andrews C. 1994. Amulets of Ancient Egypt. United Kingdom: British Museum Press. pp. 190–236. Arriaza BT. 1998. South American mummies: culture and disease. In: Cockburn A, Cockburn E, Reyman TA, editors. Mummies, disease, and ancient cultures. 2nd ed. Cambridge, UK: Cambridge University Press. pp. 190–236. Ascenzi A, Bianco P, Nicoletti R, Ceccarini G, Fornaseri M, Graziani G, Guiliani MR, Rosicarello R, Ciuffarella L, GrangerTaylor H. 1996. The Roman mummy of Grottarossa. In: Spindler K, Wilfing H, Rastbichler-Zissernig E, zur Nedden D, Nothduffer H, editors. Human mummies. Austria: Springer-Verlag/Wien. pp. 205–218. Baldock C, Hughes SW, Whittaker DK, Taylor J, Davis R, Spender AJ, Tonge K, Sofat A. 1994. 3D reconstruction of an ancient Egyptian mummy using X-ray computer tomography. JRSM 87:806–808. Bianucci R, Mattutino G, Lallo R, Charlier P, Jouin-Spriet H, Peluso A, Highan T, Torre C, Massa ER. 2008. Immunilogical evidence of plasmodium falciparum infection in an Egyptian child mummy from the early dynastic period. J Archaeol Sci 35: 1880–1885. Brown KR, Wood H. 1999. The utility of minimal CT scanning in the study of two Egyptian mummy heads. Int J Osteoarchaeol 9: 199–204. Buikstra JE, Ubelaker DH. 1994. Standards for data collection from human skeletal remains. Fayetteville, AK: Akransas Archaeological Survey Report Number 44. Ceruti C. 2004. Human bodies as objects of dedication at inca mountain shrines (north-western Argentina). World Archaeol 36: 103–122. Chan SS, Elias JP, Hysell ME, Hallowell MJ. 2008. CT of a ptolemaic period mummy from the ancient Egyptian city of Akhmim. RadioGraphics 28:2023–2032. Chew FS, Relyea-Chew A, Ochoa ER. 2006. Postmortem computed tomography of cadavers embalmed for use in teaching gross anatomy. J Comput Assist Tomo 30:949–954. Chhem RK, Brothwell DR. 2008. Paleoradiology: imaging mummies and fossils. New York: Springer. Collier R. 2009. Radiologists virtually unwrap mummy’s secrets. CMAJ 180:E23–E24. Cockburn A, Barraco RA, Reyman TA, Peck WH. 1975. Autopsy of an Egyptian mummy. Science 187. pp. 115–1160. Cockburn A, Barraco RA, Rayman TA, Peck WH. 1998. A Classic MummyL PUM II. In: Cockburn A, Cockburn E, Reyman TA, editors. Mummies, disease, and ancient cultures. 2nd ed. Cambridge, UK: Cambridge University Press. David R. 2008. Mummies and modern science. Cambridge: Cambridge University Press. Fleming S, Fishman B. 1980. The Egyptian mummy: secrets and science. Philadelphia: University Museum. Friedrich KM, Nemec S, Czerny C, Fischer H, Plischke S, Gahleitner A, Viola TB, Imhof H, Seidler H. 2007. The story of 12 chachapoyan mummies through multidetector computed tomography. Eur J Radiol. doi:10.1016/j.ejrad.2009.07.009 Gaafar H, Abdel-Monem MH, Elsheikh S. 1999. Nasal edoscopy and CT study of pharonic and roman mummies. Acta Otolaryngol 119: 257–260. Gregorczyk A, Hydzik A, Moczulska K, Niwinski A. 1994. CT investigation of Egyptian mummy form the XXI dynasty period. Przegl Lek 51:57–61.
1110
COX
Hawass Z, Saleem SN. 2011. Mummified daughters of king tutankhamun: archaeologic and CT studies. AJR Am J Roentgenol 197: W829–W836. Hawass Z, Shafik M, R€ uhli FJ, Selim A, El Sheikh I, Abdel Fatah S, Amer H, Gabella F, Gamal Eldin A, Egarter-Vigl E, et al. 2009. Computer tomographic evaluation of pharaoh tuutankhamun ca. 1300 BC. Ann Serv Antiq Egypt 81:159–174. Hunt DR, Hopper LM. 1996. Non-invasive investigations of human mummified remains by radiographic techniques. In: Spindler K, Wilfing H, Rastbichler-Zissernig E, zur Nedden D, Nothduffer H, editors. Human mummies. Austria: Springer-Verlag/Wien. pp.15– 32. Jackowski C, Bolliger S, Thali MJ. 2008. Common and unexpected findings in mummies from ancient Egypt and South America as revealed by CT. RadioGraphics 28:1477–1492. Jones RM, Faqir A, Millett TD, Moos KF, McHugh S. 2005. Bridging and dimensions of the sella turcica in subjects treated by Surgical-orthodontic means or orthodontics only. Angle Orthod 75:714–718. Kim MJ, Park SS, Bok GD, Coi YH, Lee IS, Shin kj Han GR, Youn M, Han SH, Kang IW, Chang BS, Cho YJ, Chung YH, Shin DH. 2006. Medieval mummy from yangju. Archaeol Ethn Anthr Euras 4:122–129. Kim SB, Shin JE, Park SS, Bok GD, Chang YP, Kim J, Chung YH, Yi YS, Shin MH, Chang BS, Dong HS, Kim MJ. 2006. Endoscopic investigation of the internal organs of a 15th-century child mummy from yangju, korea. J Anat 209:681–688. Lawrence SV. 1980. Unraveling the mysteries of the mummies. Science 118:362–364. Lee IS, Kin MJ, Yoo DS, Lee YS, Park SS, Bok GD, Han SH, Chung YH, Chang BS, Yi YS, Oh CS, Shin DH. 2006. Threedimensional reconstruction of medieval child mummy in Yangju, Korea, using multi-detector computed tomography. Ann Anat 189: 558–568. Lewin PK. 1978. Whole body CT scan of an Egyptian mummy (abstract). Paleopath Assn T7–T8. Lewin PK, Harwood-Nash DC. 1977. Computerized axial tomography in medical archaeology. Paleopath Newsl 17:8–9. Lynnerup N. 2009. Medical imaging of mummies and bog bodies. Gerontology 56(5):441–448. MacLeod RI, Wrights AR, McDonald J, Eremin K. 2000. Historical review: mummy 1911-210-1. RCSE 45:85–92. Magid D, Bryan BM, Drebin RA, Ney D, Fishman EK. 1989. Three-dimensional imaging of an Egyptian mummy. Clin Imag 13:239–240. Marx M, D’Auria SH. 1986. CT examination of eleven Egyptian mummies. RadioGraphics 6:321–330.
Marx M, D’Auria SH. 1987. Three-dimensional CT reconstructions of an ancient human Egyptian mummy. AJR 150:147–149. Melcher AH, Holowka S, Pharaoh M, Lewin PK. 1997. Non-invasive computed tomography and three-dimensional reconstruction of the dentition of a 2800-year-old Egyptian mummy exhibiting extensive dental disease. Am J Phys Anthr 103:329–340. Mininberg DT. 2001. The museum’s mummies: an inside view. Neurosurgery 49:192–199. Nelson AJ, Chhem R, Cunningham IA, Friendman SN, Garvin G, Gibson G, Granton PV, Holdworth DW, Holowka S, Longstaffe F, Lywood V, Nguyen N, Shaw R, Trumpour M, Wade AD, White CD. 2011. The ROM/UWO mummy project: a microcosm of progress in mummy research. Yearbook of Mummy Studies. Vol. 1. Pfeil, F. M€ uchen, Germany: Verlag. Nickol T, Germer R, Lieberenz S, Schmidt F, Wilke W. 1995. An examination of the dental state of an Egyptian mummy by means of computer tomography: a contribution to “dentistry in ancient egypt.” J Histol Dent 43:105–112. Notman DNH, Tashjian J, Aufderheide AC, Cass OW, Shane OC, Berquist TH, Gray JE, Gedgaudas E. 1985. Modern imaging and endoscopic biopsy techniques in Egyptian mummies. AJR 146:93–96. Pickering RB, Conces DJ, Braunstein EM, Yurco F. 1990. Threedimensional computed tomography of the mummy wenuhotep. Am J Phys Anthr 83:49–55. Previgliano CH, Ceruti C, Reinhard J, Araoz FA, Diez JG. 2002. Radiological evaluation of the llullaillaco mummies. AJR 181: 1473–1479. Villa C, Lynnerup N. 2012. Hounsfield units ranges in CT-scans of bog bodies and mummies. Anthropol Anz 69:127–145. Wade AD, Garvin GJ, Hurnanen JH, Williams LL, Lawson B, Nelson AJ, Tampieri D. 2012. Scenes from the past: multidetector CT of Egyptian mummies of the redpath museum. Radiographics 32:1235–1250. Wakai S, Fukushima T, Teramoto A, Sano K. 1981. Pituitary apoplexy: its incidence and clinical significance. J Neurol 55. pp. 187–193. Watson EJ, Myers M. 1993. The mummy of Baket-en-her-nakhy in teh hancock museum: a radiological update. J Egypt Archaeol 79: 179–187. Weber GW. 2001. Virtual anthropology (VA): a call for glasnost in paleoanthropology. Anat Rec 265:193–201. Zimmerman MR. 1996. Mummies of the Arctic regions. In: Spindler K, Wilfing H, Rastbichler-Zissernig E, zur Nedden D, Nothduffer H, editors. Human mummies. Austria: Springer-Verlag/Wien. pp. 83–92.
A Critical Look at Mummy CT Scanning SAMANTHA L. COX1,2* Physical Anthropology Section of the Penn Museum, University of Pennsylvania, Pennsylvania, USA 2 Department of Archaeology and Anthropology, Division of Archaeology, University of Cambridge, Cambridge, UK 1
ABSTRACT Computed tomography scanning of mummies has been conducted for almost 40 years, and has become an increasingly popular method of mummy study in the 21st century. However, most CT scan analyses published today still do little more than praise the technique’s nondestructive, non-invasive properties. Despite the wealth of information contained within a modern, high definition scan, most researchers have yet to fully exploit the full potential of this technology. In addition, those that have utilized CT scanning in ancient remains continue to interpret mummified tissues as though they were living, without investigating how taphonomy and mummification could have effected the tissue images that are produced. Because of this, there is very limited information available for clear interpretation of mummy CT’s. This article presents a critical assessment of the development of mummy CT scanning and presents the results of two Egyptian mummies CT’ed at the Penn Museum as an example of the potentials and pitfalls of high-resolution scanning. Anat C 2015 Wiley Periodicals, Inc. Rec, 298:1099–1110, 2015. V
Key words: mummy; CT scan; paleopathology
For nearly four decades, computed tomography (CT) scanning has been employed in the scientific study of mummified remains. However, despite the numerous mummies that have been examined, we know almost nothing about them beyond the process of their mummification. More than that, the formula of most CT scan studies of mummies has remained almost unchanged since they first began in 1977. The vast majority of articles are case reports that come to the same conclusion: CT scans are a powerful, nondestructive, noninvasive way to study mummies. Though this is true, much time has passed and the technology continues to improve and evolve, but its potential for mummy research has never been fully exploited. While CT scan studies have become increasingly prevalent over the last 10 years, very little research has been done to explore the limitations of the technology or the differences between living and long-dead tissues in imaging. This article takes a critical look at the history of mummy CT scanning to understand the current state of paleo-imaging and address the weaknesses of the field in order utilize CT technology to its fullest potential and encourage mummy studies to continue evolving. The CT scans of two Penn Museum mummies are presented here as an example of the wealth of information available from detailed CT scans, but also as an illustration of the limiC 2015 WILEY PERIODICALS, INC. V
tations imposed on CT analysis due to the dearth of information available on tissue changes in paleo-imaging.
REVIEW OF MUMMY CT HISTORY Computed tomography scanning was first developed in 1975 for medical imaging, and it was a mere 2 years later that the technology was applied to a mummy. Lewin and Harwood-Nash (1977) performed the first CT scan on an Egyptian mummy named Nakht from the Royal Ontario Museum in 1977 to examine the brain. Though the findings of this study were not remarkable, it was a mile-stone in mummy studies for the application of the new technology. Lewin (1978) also produced the first whole-body scan of a mummy in 1977. From 1977 until 1985, CT scans were not often used for analysis in
*Correspondence to: Division of Archaeology, Department of Archaeology and Anthropology, University of Cambridge, Downing Street, Cambridge CB23DZ, E-mail: [email protected] Received 16 January 2015; Accepted 30 January 2015. DOI 10.1002/ar.23149 Published online in Wiley Online Library (wileyonlinelibrary. com).
1100
COX
their own right, but to aid in directing mummy autopsy and tissue sampling. For example, Notman and his team (1985) on the Minnesota Mummy Project, Zimmerman (1996; note the study was carried out in 1983) on Alaskan mummies from St. Lawrence Island, and Ascenzi et al. (1996; note, the study was carried out in 1983) on the mummy of a girl from Rome. From 1985 to the present at least one mummy CT publication has come out every year, and they have become increasingly numerous through time. These articles consist mostly of case reports that generally describe mummification techniques, artifacts, preservation, and pathology (e.g., Marx and D’Auria, 1986, 1987; Magid et al., 1989; Pickering et al., 1990; Watson and Meyers, 1991; Gregorczyk et al., 1993; Hughes, 1993; Baldock et al., 1994; Arriaza, 1998; Gaafar et al., 1999; Macleod et al., 2000; Thekkaniyil et al., 2000; Mininberg, 2001; Hoffman et al., 2002; Previgliano et al., 2002; Shin et al., 2003; Kim et al., 2006, 2006; Lee et al., 2006; Hawass and Saleem, 2011; Nelson et al., 2011; Wade et al., 2012), though there are some publications that specifically address pathology, facial reconstructions, and new approaches and applications. Many of the early CT papers compare traditional X-rays to CT scans to determine the strengths and weaknesses of the new technology and establish it as a reliable method for investigation (for instance, Watson and Meyers, 1993, and Hunt and Hopper, 1996). These early studies are quite constrained by the low resolutions and large file sizes produced by early CT machines. It is not surprising that most of the mummies scanned are ancient Egyptian, and the extended positions of the bodies produce large scans that take up large quantities of file space. It is for this reason that some of these early scans report slice thicknesses of up to 24 mm (Marx and D’Auria, 1986), to keep the scanners themselves from overheating and file sizes to a minimum. However, while the volume of data is easier to manage, the scans could not have contained enough detail for any kind of in-depth analysis. It was the mid-1990’s before scanning resolutions were high enough to allow for the first CT investigations of mummy dentition (as in Nikol et al., 1995; Melcher et al., 1997). Also at this time, a study of a female South American mummy is noteworthy because of her probable tuberculosis. It is one of the first studies to identify a specific pathology via CT, as well as providing some of the oldest evidence of tuberculosis (Correal-Urrego and Florez, in Cockburn, 1998). At this point in the literature, CT mummy analysis seems to be progressing along at a steady pace; some authors began to experiment with utilizing CT scans in different ways, but with limited results (e.g., Brown and Wood, 1999). However, by the beginning of the new millennium, CT studies appear to stagnate. Despite continued advances in CT technology, the focus is still primarily on describing the mummies and the mummification process on an individual basis. Friedrich et al. (2007) does this with 12 Chachapoyan mummies, Gupta et al (2008) with an Egyptian head as does Chan et al. (2008). Jackowski et al. (2008) attempt to compare mummies from Ancient Egypt to those from Peru, but the article focuses heavily on Egyptian mummification, only briefly discussing Peru and the connection between the two is unclear. Collier (2009) reports on another mummy being CT scanned in Toronto, Gerloni (2009) again considers Egyptian dentition and Lynnerup (2009) applies these techniques to the Bog Bodies housed in the British Museum.
Though this is not an exhaustive list of publications by any means, the one thing that is consistent in all these studies is the use of higher-resolution CT scans, though often a lower resolution is used for the post-cranial elements than for the head. Teeter and Vannier (2009) published the CT scan of Meresamun for exhibit at the Oriental Institute in Chicago at a reconstructed slice thickness of 0.7 mm, claiming that this is the highest resolution mummy scan published to that date. There are surprisingly few pathology studies. Though this is what the technology was designed for in a clinical setting, it has rarely been applied to anthropological investigations. Most published work of this type is rather limited in scope, and the full potential of the technology seems to have not been adequately explored (for a full review and critique see Chhem, 2006). The first pathology-specific study with a CT scan was probably € the one performed on Otzi the Iceman (Holden, 2001), since this is likely the first time that scan resolutions were high enough to allow for this type of analysis. The most famous pathology study is probably that of the mummy of King Tutankhamun to attempt to determine the cause of death of the boy king (Hawass et al., 2009). In a few instances, researchers have attempted to use CT pathology analyses in more creative ways. For example, Ceruti (2004) looked at pathology to determine cause-of-death in Peruvian mummies and compared it to historical accounts of human sacrifice. In other cases, investigators have attempted to use CT for pathology analysis, but were unsuccessful in identifying any, as was the case for Bianucci et al. (2008). It is apparent after this review that there are some large gaps in knowledge and understanding of CT scanning in ancient remains. As of yet, the vast majority of articles published are merely individual case reports. Nothing has been done to take the data presented in those articles and synthesize it into a work that has meaning applicable to the field of anthropology. There is potential here to evaluate health in mummies, as a population, given how many have been CT scanned, but focus continues to remain on mummies as individuals rather than representative members of a group. The exception to this generalization is the Horus Mummy Project team. This group has evaluated the CT scans of over 20 mummies and found persuasive evidence of heart disease that would indicate some prevalence throughout the upper class population (Allam et al., 2009). This study is the only example of a synthetic analysis that finally takes CT scans beyond the “gee, wow!” factor and begins to explore the potential of this hi-tech method to add to our knowledge of ancient Egyptian life, questioning the assumptions that have always been made regarding health and diet of ancient people. Unfortunately, only a handful of researchers have ever considered and written about any of the negative aspects of CT scanning. There is little consistency in the manner in which findings are reported, for instance the specs for the CT machines and scanning parameters are often omitted. In 2001, Weber published a detailed critique of this issue in which he proposes a scheme of information that should be standardly reported in high technology academic articles. He is of the opinion that this kind of digital media should be made freely available for all scholars to consider, so that other evaluations and comparisons can be made. It is quite apparent that
A CRITICAL LOOK AT MUMMY CT SCANNING
few publications have adopted his standard, and, to my knowledge, no one has proposed another. Few share their scans; however, this may be starting to change with the University of Pennsylvania Museum’s Online Research Scan Archive (ORSA) and the IMPACT Radiological Mummy Database, which are free for scientists that wish to use them. While there are many wonderful and useful applications of CT scans, there is also a noticeable gap in the literature concerning the discussion of its weaknesses. For instance, few consider how mummification, especially artificial mummification, may alter the appearance of tissues. Chew et al (2006) noted in their cadaver study that modern embalming changes the way that structures appear radiologically in recently dead cadavers. The image quality was greatly reduced by beam hardening artifacts, attenuation between soft tissues was reduced as compared with living people, and other less uniform changes were visible due to the embalming process. If this is the case today, it could have critical implications for the interpretations of mummified tissues, as most scans of ancient remains are read and interpreted as though they were living people. Villa and Lynnerup (2012) have begun to tackle some of these concerns, but information is still needed before mummified remains can be reliably interpreted with medical imaging techniques. To date, there is no published data, of which I am aware, that investigates embalming materials, their appearance, and their effects on tissue appearance on CT. A good case example is the famous CT scan of Tutankhamun (Hawass et al., 2009). The debate over the whether or not the femoral fracture represents an ante- or postmortem break, and thus whether or not it could have contributed to events surrounding the boy king’s death, cannot be laid to rest because the research has not been done to be say how the resin in mummies behaves through time or whether or not mummification processes would eliminate any signs of hematoma. Though many great studies have emerged with this technology, there is still much more that has to be considered before the true meaning of what has been observed can be adequately assessed and synthesized. R€ uhli et al. (2004) is one of the very few that offer caution about the interpretation of CT scans. He argues that every radiologic examination of ancient human remains should include an experienced radiologist as well as an experienced physical anthropologist to interpret the images. The radiologist can recognize pathology more readily than the anthropologist, but the anthropologist will be more adept at recognizing confounding factors that might mar the diagnosis. Resins and other artifacts of mummification are the most important factors to recognize. There is no comprehensive guide to diagnosing paleopathology in radiology; however, Chhem and Brothwell published Paeloradiology: Imaging Mummies and Fossils (2008) which dedicates about half the book (100 pages) to this topic. Though it is by no means complete, and focuses mainly on X-rays it gives pictures and characteristics of the most common pathologies in both humans and animals and is at least a good starting reference for analyzing paleoradiography.
THE PENN MUMMIES When the CT scanning of two Egyptian mummies from the University of Pennsylvania Museum (Penn
1101
Museum) was undertaken, it was with the aim of setting these two individuals within their historical contexts. It was soon clear, however, that despite all the mummies that have been CT’ed in the world, there was not sufficient information available to soundly interpret the images or understand these two people as a part of Ancient Egyptian life. They are presented here as examples of the potential and limitations faced when interpreting CT scans of ancient human remains. PUM II and Hapi-Man, from the Penn Museum were CT scanned in April 2009, at the Hospital of the University of Pennsylvania (HUP). Dr. Morrie Kricun Emeritus Professor of Radiology at HUP and the University of Pennsylvania Medical School assisted with the interpretations of the mummy images. A spiral multislice Siemens’ SOMATOM Sensation 64-Slice Computed Tomography machine scanned the mummies with a slice thickness of 1 mm (0.5 mm reconstructed). To produce the best results, both mummies were scanned in three sections, from the head to mid-thorax, mid-thorax to mid-femur, and mid-femur through the feet, producing a total of about 2,000 images for each mummy. The analysis was done using the freeware program Osirix v. 3.7 with the 64-bit upgrade on an Apple Mac Pro. Hapi-Man was excavated in Abydos in 1902 by the famous archaeologist Sir Flinders Petrie. He is thought to date from the Late Period and be about 35–45 years of age at death (Lawrence, 1980). Hapi-Man has been xrayed on two occasions during his stay at the Penn Museum and research in the Penn Museum Archives turned up a set of apparently unpublished X-rays from April 1932, by Dr. J.G. Cohen. A second envelope of completely unmarked films was also found in archives and matched to Hapi-Man based on the placement of the amulets around the neck. The date and researcher for this second set of X-rays is unknown, but it is clear that they are more recent than the first set and it is possible that they were produced in the late 1970s for the Penn Museum exhibit The Egyptian Mummy: Secrets and Science. The CT scan of Hapi-Man revealed information about his mummification and identified some possible pathology. It is clear on the scans that his abdomen was stuffed with rolled bundles of cloth that might contain the internal organs as is typical of Late Period mummification (David, 2008); however, there does not appear to be anything inside and it is likely that these are probably symbolic only (Fleming, 1980). Hapi-Man also had his brain removed through the nose and the body is completely covered in resin which lies in a pool in the cranium, thorax and abdomen. It is clear in the CT that he has an udjet eye on his forehead, no fewer than twelve amulets arranged around his neck, like a necklace, including beads, heart amulets, an udjet eye, goddess figurine, Horus falcon, scarab, and Amun feather. The left hand has an unidentifiable square amulet on the ring finger, though there does not appear to be anything attaching it to the finger (Fig. 1). On the right rib cage there are three rows of amulets, the first is three scarab beetles, the second is four djed pillars, and the third is five tyet knots with a goddess figurine, probably of Isis (Fig. 2). The placement of these kinds of amulets in the wrappings of the deceased was quite common, but it is clear from the scans that none of these objects are likely metallic, most are probably fiance (Andrews, 1994).
1102
COX
Fig. 1. Hapi-Man amulets, (top left) goddess figure, Horus falcon, and square “ring” on hand; (top right) jars, udjet eye, scaraab, beads; (bottom left) Amun feather; (bottom right) udjet eye on forehead.
The teeth are quite clear, and we can confirm the age of the individual to be between 35 and 45 years of age using cranial suture closure patterns (Walker in Buikstra and Ubelaker, 1994). There is clear evidence of uncovertebral arthropathy as well a rib fracture. The scans do show that there are numerous air spaces in the vertebral bodies that are not present in living individuals; it is unclear if this has to do with the age of the specimen or if it is pathological. There is also a uniform distribution of some substance around the periphery of the annulus of the thoracic and lumbar vertebrae which could either be attributed to the resin or pathology, likely only an autopsy would be able to tell for certain.
However, evidence of discontinuous calcification of the annulus at every disc level and clear osteoarthritis between the posterior elements of the L4 and L5 is present. The fifth lumbar vertebra is sacralized as well as fused bilaterally to the sacrum on the caudal side. This might have marginally effected the mobility of his lumbar spine but would not have been crippling. Osteoarthritis was observed on both hips, but it was apparent on the scan that there was no bone reaction to the cartilage loss, suggesting that this may be a result of specimen age or mummification process instead of disease. In the feet, there were no bone reactions and only some medial narrowing of the first metatarsophalangeal
A CRITICAL LOOK AT MUMMY CT SCANNING
joints, probably not enough to be considered arthritic, but it is a possibility. This may have only been the very beginnings of arthritic changes, just an unusually narrow joint space, or maybe an artifact of mummification. It appears on the scan that Hapi-Man’s feet became detached from his body at some point after his mummification. This is most visible in the very large tibiotarsal joint space but upon close inspection the wrappings are
Fig. 2. (left) Right rib cage, top row- three scarab beetles, middle row- four djed pillars, bottom row- five tyet knots and Isis figurine, (right) scarab beetle on left rib cage.
1103
torn through in the same plane. There are bright spots on the scan inside the joint space and on the wrappings that are hypothesized to be some kind of adhesive or restorative material to reconnect the feet to the body. It is unclear when or how this damage occurred or was repaired. Like many Egyptian mummies, PUM II (an acronym for Pennsylvania University Museum II) has a long and rather cloudy history. The area of Egypt from which he comes is unknown and he is first recorded in the early twentieth century on a ship to Philadelphia. He came to America by way of John T. Morris, a wealthy Philadelphian who then donated him to the Philadelphia Museum of Art (PMA). Though the sarcophagus is highly decorated, the name of the individual was not recorded on it nor was any clues about his life. He remained at the PMA until the end of January 1973 when he was shipped to Michigan to be autopsied on February 1 at the Wayne State University Medical School in Detroit as part of a symposium there entitled Death and Disease in Ancient Egypt. Known as PUM II because he was the second mummy to be acquired by the institution, radiographic and xerographic exams were done at the Mt. Carmel Mercy and Hutzel hospitals both before and after the autopsy (Cockburn, 1998). Following the autopsy, he was sent to be on exhibit at the Museum of Natural History in Washington DC, on loan to the Smithsonian Institution (Cockburn 1975). He is now on display and has been a part of the Egyptian Mummies: Secrets and Science exhibit since it opened in 1980 at the Penn Museum. The right toes of PUM II are flexed, pulled back towards the top of the foot. Though the X-rays and autopsy report of PUM II dismissed the unusual appearance of the toes as an effect of tight wrapping, CT scans show that it is probably a condition known as hammertoe that is resultant of a cavus deformity of the right
Fig. 3. Measured angle of cavus foot deformity, 45 is considered normal and 90 is severely cavus, here the angle on the right foot of PUM II measures about 100 (image captured using Osirix v. 3.7).
1104
COX
Fig. 4. (left) Axial slice illustrating the difference in density between the two feet (image captured with Osirix 3.7), (right) 3D reconstruction of feet colored by density, red is high density, yellow is medium, and green is low red is high density, yellow is medium, and green is low
Fig. 5. Arrow indicates small bunion on the right medial proximal first metatarsal of PUM II.
foot. Aminian and Sangeorzan (2008) define cavus foot deformity as describing “a spectrum of foot shapes that have a high arch” and is usually characterized by a high pitch angle of the calcaneus and plantar flexion of the
midfoot. Radiographically, a foot can be called a cavus foot with an increased Hibbs angle, measured through the axis of the calcaneus and metatarsal. A normal foot is around 45 while cavus can be 90 ; the angle on PUM
A CRITICAL LOOK AT MUMMY CT SCANNING
Fig. 6. Orange arrow shows large air pocket between S1 and L5 would be indication of a hernia in living people but is inconclusive here, blue arrows point to narrowing of the hip joints that could be mild arthritis or an artifact of mummification in PUM II.
Fig. 7. What appears to be DISH in PUM II may just be residual soft tissues and resin.
1105
1106
COX
Fig. 8. Mild hypertrophy of the lischka joints with Schmorl’s nodes on the caudal C3 (top) and cranial C4 mildly encroaching on the neural foramen (bottom) in PUM II.
II’s foot is about 100 (Fig. 3). The bones of the foot are formed normally and well articulated, indicating that this deformation occurred after the bones matured. Today, two-thirds of modern cases of cavus foot have an underlying neurological condition (Aminian and Sangeorzan, 2008). Other causes can be muscle spasticity leading to muscle imbalance in the foot. Based on the periosteal reaction described in the autopsy, it might be possible that a chronic infection, perhaps due to parsitic infestation as those authors suggest, could have led to muscle imbalance in that leg and caused the cavus foot. The left foot has a normal shape, but it is significantly less radiolucent than the right foot and the soft tissues
appear atrophied as compared to the other foot. The authors of the autopsy noted that the right foot appeared swollen, so perhaps what we interpreted as atrophy is actually normal and the other leg is possibly inflamed (Fig. 4). However, since the body was dehydrated with natron it is necessary to consider whether swellings in a limb would be been preserved. The coloring on the scan might also suggest that for some reason the right leg is full of resin or fat and the left is not, accounting for the difference in density. A deep subarticular cyst is visible in the center of the left and right tali as are small cysts in the MT 1 and 2 on the left foot. The medial aspect of the MT1 seems to have a small
A CRITICAL LOOK AT MUMMY CT SCANNING
1107
Fig. 9. Orange arrow indicates nasal bones that could show healed fracture, blue arrow shows where the ethmoid was broken to remove the brain of PUM II.
Fig. 10. The cloudy appearance of the right mastoid of PUM II as compared to the left could show chronic mastoiditis but could also be the air spaces filled with resin and fat.
bunion forming, perhaps as a result of the gait caused by the cavus foot (Fig. 5). The talus and tibia appear to be starting to fuse together on the lateral half to two-
thirds of the left tibiotarsal joint, perhaps be the result of some kind of trauma (Fig. 20). The knees show subchondral cysts, the beginnings of osteoarthritis,
1108
COX
Fig. 11. Green lines show the abnormal A/P measurements (ca. 25 mm at the maximum) of the sella turcica probably indicative of a tumor of the pituitary gland in PUM II.
especially in the right knee (Fig. 21). It is quite clear in Fig. 6 that the legs of PUM II are different lengths, but after careful consideration, it was decided that this is because his pelvis was not completely level when mummified, making his legs appear asymmetric; it should be noted that this kind of asymmetry is often also seen with cavus foot (Aminian and Sangeorzan, 2008). A bone island is visible in the right ilium and narrowing of the hip joints suggest mild arthritis, though it could also be an effect of mummification. The vertebral disk between the L5 and S1 is in good shape but a large air pocket in the center would indicate a hernia in a living person, and this cannot be said conclusively in the case of ancient remains (Fig. 6). Though the autopsy reported a sixth lumbar body on the X-rays, we found no indications of this; however, the vertebral bodies were removed from this area during autopsy to look for spinal cord samples, so this could be affecting the interpretation. Most of the vertebral disks are in good condition with no degenerative changes but Schmorl’s nodes are present on the caudal endplate of T11. A build-up of some substance on the thoracic vertebral bodies was originally identified as diffuse idiopathic skeletal hypertrophy (DISH) (Fig. 7). However, upon further investigation, this was not visible on the cervical spine and in Dr. Kricun’s experience, it would be incredibly unusual to suffer from DISH in the thorax without involvement of the cervical vertebrae. We finally decided that it is more likely to be resin-covered soft tissues adhering to the spine. Small osteophytes are visible on some of the vertebral bodies and those with the Schmorl’s nodes indicate a mild osteoarthritis. There appears to be a mild bilateral hypertrophy of the uncovertebral lishka joints
between the C3 and C4. An osteophyte is visible on the posterior caudal end plate of the C4 and the caudal facet of C4 also has an osteophytic cyst. There is mild encroachment on the neural foramen in this area with an osteophytic ridge and large Schmorl’s nodes on the anterior caudal aspect of C3 and anterior cranial aspect of C4 (Fig. 8). The autopsy reports did not give an estimated age-atdeath and most of his cranial sutures were removed during autopsy. However, the condition of his teeth and joints looks quite similar to Hapi-Man so we decided that it is probably reasonable to estimate that he is about the same age, 35–45 years. PUM II’s nose exhibits an unusual curve at the end that is not consistent with the general shape, perhaps the result of a healed fracture. The ethmoid has clearly been broken through for access to the cranium and removal of the brain (Fig. 9). The mastoid bone appears to be cloudy on the right side, indicative of chronic mastoiditis, but this could be also be another example of the fat and resins filling the air spaces (Fig. 10). Measurements indicate an abnormally large sella turcica, the maximum anteroposterior (A/P) dimension is about 25 mm and maximum depth is 14 mm (Fig. 11). The normal dimensions, given as a range by Jones et al. (2005) are up to 16 mm A/P and 12-mm deep. Though the depth measurement is not far from what is considered to be normal, the A/P is much larger than the given range and suggests the presence of a pituitary tumor. Pituitary tumors cause a range of different effects but the most common are hormonal imbalances, dizziness, disturbances of consciousness, and blindness (Wakai, 1981). The tumor is of significant enough size that it could have been the cause of death
A CRITICAL LOOK AT MUMMY CT SCANNING
for PUM II, but if not, it was at least a source of great discomfort and illness.
1109
(2006) concerning CT analyses on medical school cadavers serves as a warning, and we need to know more about these changes before CT scans of ancient remains can yield any meaningful results.
DISCUSSION As is always the case with studying ancient remains, there are many confounding factors at work, including preservation and taphonomic processes. There are many instances in the cases of these two mummies when a conclusion could not be reached because of insufficient data on the imaging of ancient tissues. However, the detailed scan resolution allowed for very specific analysis of minor changes, some of which can be confirmed as probable arthritis and others which are less clear as they could be artifacts of mummification. Research in the future needs to focus on sorting out the differences between ante- and postmortem changes on these kinds of medical images as a means of eventually being able to tease out the most accurate results. If there is more data available on this in the future, we could, perhaps, be able to ascertain if thin layers of resin in the joints give the appearance of arthritis when condensed on X-ray film but would then be absent on CT scans. As it stands, the amount of information on these processes is limited and means that the data cannot be properly interpreted except for large, severe pathologies. There is a distinct lack of data that has been synthesized between mummies, but discrepancies in scan reporting means that published studies cannot be compared or aggregated because of differences in scan specifications. Even if it were possible to compare all the mummies that have been scanned thus far, it is quite likely that a lack of understanding of the impacts of taphonomy and mummification on the radiologic appearance of tissues means that many published findings are likely to have been misinterpreted. It is clear from Hapi-Man and PUM II that there are some common factors, arthritis and dental pathology being the main ones, but two individuals are not enough to draw conclusions about a population. The cavus foot deformity and the enlargement of the sella turcica in PUM II are interesting pathologies to note, but they have little meaning beyond this one individual unless other examples can be found in other mummies. The smaller pathologies like arthritis that were observed numerous times in this analysis are ones that we know will translate more between individuals; however, these cannot yet be positively identified and interpreted until data is compiled to distinguish taphonomic changes from antemortem degradation. At this point, it is probable that over one hundred mummies have been CT scanned and reported on. None of this data has yet been synthesized into a work that is meaningful beyond the scope of more than the one individual scanned. Because data has been collected on a number of mummies, the next logical step would be to collate the data and see the conclusions that can be drawn about these mummies as a population rather than just as single people. However, this study shows that, despite the plethora of mummies imaged, the data to interpret scans more critically does not yet exist. This casts some doubt on the conclusions that have been drawn in these numerous investigations since we cannot reliably recognize taphonomy and mummification radiologically. The medical research published by Chew et al.
LITERATURE CITED Allam, AH, Thompson RC, Wann LS, Miyamoto MI, Thomas GS, 2009. Computed tomographic assessment of atherosclerosis in ancient Egyptian mummies. JAMA 302(19):2091–2094. Aminian A, Sangeorzan BJ. 2008. The anatomy of cavus foot deformity. Foot Ankle Clin N Am 13:191–198. Andrews C. 1994. Amulets of Ancient Egypt. United Kingdom: British Museum Press. pp. 190–236. Arriaza BT. 1998. South American mummies: culture and disease. In: Cockburn A, Cockburn E, Reyman TA, editors. Mummies, disease, and ancient cultures. 2nd ed. Cambridge, UK: Cambridge University Press. pp. 190–236. Ascenzi A, Bianco P, Nicoletti R, Ceccarini G, Fornaseri M, Graziani G, Guiliani MR, Rosicarello R, Ciuffarella L, GrangerTaylor H. 1996. The Roman mummy of Grottarossa. In: Spindler K, Wilfing H, Rastbichler-Zissernig E, zur Nedden D, Nothduffer H, editors. Human mummies. Austria: Springer-Verlag/Wien. pp. 205–218. Baldock C, Hughes SW, Whittaker DK, Taylor J, Davis R, Spender AJ, Tonge K, Sofat A. 1994. 3D reconstruction of an ancient Egyptian mummy using X-ray computer tomography. JRSM 87:806–808. Bianucci R, Mattutino G, Lallo R, Charlier P, Jouin-Spriet H, Peluso A, Highan T, Torre C, Massa ER. 2008. Immunilogical evidence of plasmodium falciparum infection in an Egyptian child mummy from the early dynastic period. J Archaeol Sci 35: 1880–1885. Brown KR, Wood H. 1999. The utility of minimal CT scanning in the study of two Egyptian mummy heads. Int J Osteoarchaeol 9: 199–204. Buikstra JE, Ubelaker DH. 1994. Standards for data collection from human skeletal remains. Fayetteville, AK: Akransas Archaeological Survey Report Number 44. Ceruti C. 2004. Human bodies as objects of dedication at inca mountain shrines (north-western Argentina). World Archaeol 36: 103–122. Chan SS, Elias JP, Hysell ME, Hallowell MJ. 2008. CT of a ptolemaic period mummy from the ancient Egyptian city of Akhmim. RadioGraphics 28:2023–2032. Chew FS, Relyea-Chew A, Ochoa ER. 2006. Postmortem computed tomography of cadavers embalmed for use in teaching gross anatomy. J Comput Assist Tomo 30:949–954. Chhem RK, Brothwell DR. 2008. Paleoradiology: imaging mummies and fossils. New York: Springer. Collier R. 2009. Radiologists virtually unwrap mummy’s secrets. CMAJ 180:E23–E24. Cockburn A, Barraco RA, Reyman TA, Peck WH. 1975. Autopsy of an Egyptian mummy. Science 187. pp. 115–1160. Cockburn A, Barraco RA, Rayman TA, Peck WH. 1998. A Classic MummyL PUM II. In: Cockburn A, Cockburn E, Reyman TA, editors. Mummies, disease, and ancient cultures. 2nd ed. Cambridge, UK: Cambridge University Press. David R. 2008. Mummies and modern science. Cambridge: Cambridge University Press. Fleming S, Fishman B. 1980. The Egyptian mummy: secrets and science. Philadelphia: University Museum. Friedrich KM, Nemec S, Czerny C, Fischer H, Plischke S, Gahleitner A, Viola TB, Imhof H, Seidler H. 2007. The story of 12 chachapoyan mummies through multidetector computed tomography. Eur J Radiol. doi:10.1016/j.ejrad.2009.07.009 Gaafar H, Abdel-Monem MH, Elsheikh S. 1999. Nasal edoscopy and CT study of pharonic and roman mummies. Acta Otolaryngol 119: 257–260. Gregorczyk A, Hydzik A, Moczulska K, Niwinski A. 1994. CT investigation of Egyptian mummy form the XXI dynasty period. Przegl Lek 51:57–61.
1110
COX
Hawass Z, Saleem SN. 2011. Mummified daughters of king tutankhamun: archaeologic and CT studies. AJR Am J Roentgenol 197: W829–W836. Hawass Z, Shafik M, R€ uhli FJ, Selim A, El Sheikh I, Abdel Fatah S, Amer H, Gabella F, Gamal Eldin A, Egarter-Vigl E, et al. 2009. Computer tomographic evaluation of pharaoh tuutankhamun ca. 1300 BC. Ann Serv Antiq Egypt 81:159–174. Hunt DR, Hopper LM. 1996. Non-invasive investigations of human mummified remains by radiographic techniques. In: Spindler K, Wilfing H, Rastbichler-Zissernig E, zur Nedden D, Nothduffer H, editors. Human mummies. Austria: Springer-Verlag/Wien. pp.15– 32. Jackowski C, Bolliger S, Thali MJ. 2008. Common and unexpected findings in mummies from ancient Egypt and South America as revealed by CT. RadioGraphics 28:1477–1492. Jones RM, Faqir A, Millett TD, Moos KF, McHugh S. 2005. Bridging and dimensions of the sella turcica in subjects treated by Surgical-orthodontic means or orthodontics only. Angle Orthod 75:714–718. Kim MJ, Park SS, Bok GD, Coi YH, Lee IS, Shin kj Han GR, Youn M, Han SH, Kang IW, Chang BS, Cho YJ, Chung YH, Shin DH. 2006. Medieval mummy from yangju. Archaeol Ethn Anthr Euras 4:122–129. Kim SB, Shin JE, Park SS, Bok GD, Chang YP, Kim J, Chung YH, Yi YS, Shin MH, Chang BS, Dong HS, Kim MJ. 2006. Endoscopic investigation of the internal organs of a 15th-century child mummy from yangju, korea. J Anat 209:681–688. Lawrence SV. 1980. Unraveling the mysteries of the mummies. Science 118:362–364. Lee IS, Kin MJ, Yoo DS, Lee YS, Park SS, Bok GD, Han SH, Chung YH, Chang BS, Yi YS, Oh CS, Shin DH. 2006. Threedimensional reconstruction of medieval child mummy in Yangju, Korea, using multi-detector computed tomography. Ann Anat 189: 558–568. Lewin PK. 1978. Whole body CT scan of an Egyptian mummy (abstract). Paleopath Assn T7–T8. Lewin PK, Harwood-Nash DC. 1977. Computerized axial tomography in medical archaeology. Paleopath Newsl 17:8–9. Lynnerup N. 2009. Medical imaging of mummies and bog bodies. Gerontology 56(5):441–448. MacLeod RI, Wrights AR, McDonald J, Eremin K. 2000. Historical review: mummy 1911-210-1. RCSE 45:85–92. Magid D, Bryan BM, Drebin RA, Ney D, Fishman EK. 1989. Three-dimensional imaging of an Egyptian mummy. Clin Imag 13:239–240. Marx M, D’Auria SH. 1986. CT examination of eleven Egyptian mummies. RadioGraphics 6:321–330.
Marx M, D’Auria SH. 1987. Three-dimensional CT reconstructions of an ancient human Egyptian mummy. AJR 150:147–149. Melcher AH, Holowka S, Pharaoh M, Lewin PK. 1997. Non-invasive computed tomography and three-dimensional reconstruction of the dentition of a 2800-year-old Egyptian mummy exhibiting extensive dental disease. Am J Phys Anthr 103:329–340. Mininberg DT. 2001. The museum’s mummies: an inside view. Neurosurgery 49:192–199. Nelson AJ, Chhem R, Cunningham IA, Friendman SN, Garvin G, Gibson G, Granton PV, Holdworth DW, Holowka S, Longstaffe F, Lywood V, Nguyen N, Shaw R, Trumpour M, Wade AD, White CD. 2011. The ROM/UWO mummy project: a microcosm of progress in mummy research. Yearbook of Mummy Studies. Vol. 1. Pfeil, F. M€ uchen, Germany: Verlag. Nickol T, Germer R, Lieberenz S, Schmidt F, Wilke W. 1995. An examination of the dental state of an Egyptian mummy by means of computer tomography: a contribution to “dentistry in ancient egypt.” J Histol Dent 43:105–112. Notman DNH, Tashjian J, Aufderheide AC, Cass OW, Shane OC, Berquist TH, Gray JE, Gedgaudas E. 1985. Modern imaging and endoscopic biopsy techniques in Egyptian mummies. AJR 146:93–96. Pickering RB, Conces DJ, Braunstein EM, Yurco F. 1990. Threedimensional computed tomography of the mummy wenuhotep. Am J Phys Anthr 83:49–55. Previgliano CH, Ceruti C, Reinhard J, Araoz FA, Diez JG. 2002. Radiological evaluation of the llullaillaco mummies. AJR 181: 1473–1479. Villa C, Lynnerup N. 2012. Hounsfield units ranges in CT-scans of bog bodies and mummies. Anthropol Anz 69:127–145. Wade AD, Garvin GJ, Hurnanen JH, Williams LL, Lawson B, Nelson AJ, Tampieri D. 2012. Scenes from the past: multidetector CT of Egyptian mummies of the redpath museum. Radiographics 32:1235–1250. Wakai S, Fukushima T, Teramoto A, Sano K. 1981. Pituitary apoplexy: its incidence and clinical significance. J Neurol 55. pp. 187–193. Watson EJ, Myers M. 1993. The mummy of Baket-en-her-nakhy in teh hancock museum: a radiological update. J Egypt Archaeol 79: 179–187. Weber GW. 2001. Virtual anthropology (VA): a call for glasnost in paleoanthropology. Anat Rec 265:193–201. Zimmerman MR. 1996. Mummies of the Arctic regions. In: Spindler K, Wilfing H, Rastbichler-Zissernig E, zur Nedden D, Nothduffer H, editors. Human mummies. Austria: Springer-Verlag/Wien. pp. 83–92.
