MDCT in Paleoimaging

Although the newest equipment can produce remarkable images on living patients, it is not necessarily well suited for the examination of mummified or skeletal remains. The first obstacle would be access to the equipment. Few imaging facilities can provide repeated access to the extremely expensive equipment. Therefore, specific imaging questions must be defined before an examination is undertaken. If the entire length of a 60 in. (152 cm) mummy was going to be scanned at 0.5-mm-thick slices, that would result in 3048 images! Not only would it be very time consuming to examine each image thoroughly, it would also require a tremendous amount of computer memory to store the data. In addition, it would certainly decrease the life of the x-ray tube. A better approach would be to employ conventional radiography to locate those regions of particular interest and scan only those areas at higher resolution with thinner slices.

The other, and possibly greater, problem with the newer units would be finding a technologist to operate the unit. As previously mentioned, MDCT units have specific algorithms that have been incorporated into the software based on average-size living humans. Unfortunately, the manufacturers have not included a mummified tissue algorithm. Therefore, a specific volume of the mummy may be collected in a few seconds; however, locating and applying an appropriate algorithm from the available choices may take considerable time. This is most evident when 3D reconstructions are desired. In the living patient, if an examination of the skull calls for a 3D reconstruction of the calvarium, the completed image will include all of the skeletal components with the elimination of all soft tissue elements such as skin, muscle, and brain. Since even the thinnest components of the skull are hydrated and have adjacent soft tissue structures, the reconstruction software is programmed to recognize them, and they will be included in the 3D image. Unfortunately, this isn't the case with mummified or skeletal remains. The problem is clearly demonstrated in a mummy examined in 2000. Although the bone seems "thin," the temporal regions of the skull and humeral head of the left shoulder of the remains are clearly visible on conventional radiographs (Figure 3.7). However, on the 3D CT image, the same regions of the skull appear as "holes" (Figure 3.8), and the entire surface of the humeral head appears to be "missing" (Figures 3.9A and 3.9B). When the bone algorithm was applied during the 3D reconstruction, the lower CT numbers from the "thin" areas were omitted from the reconstruction data set. Prior to the advent of the newer CT units, it would have been possible to manually input the CT numbers that would be included in the reconstruction.

Therefore, before undertaking a CT study, the specific objectives for using the advanced modality must be determined. Elimination of superimposition that prevents an interpretation on the conventional radiographs certainly is justification for a CT examination. A mummy known as Andy provides an excellent example. He was a sideshow mummy that the Ripley's Believe It or Not organization was planning to include in a new museum opening in New Orleans, Louisiana. The initial anterior-posterior (AP) and lateral radiographs of the chest suggested rib and possibly sternal fractures (Figures 3.10A and 3.10B). Axial CT images clearly demonstrated rib fractures. In addition, the soft tissue structures of the thorax, the lungs and heart, were shifted to the right. In conjunction with the rib fractures, the shift of the organs strongly suggested the individual suffered a tension pneumothorax that led to his death (Figure 3.11).

Figure 3.7 lateral radiograph of the skull and left shoulder of a mummy examined in 2000.

Another advantage of CT is the modality's ability to demonstrate soft tissue that would otherwise not be seen on conventional radiographs. This was demonstrated in a mummy known by several names including James Penn, Stoneman, and Stoneman Willie (Figure 3.15). The story surrounding the individual, who resides in a funeral home in Pennsylvania, indicates that he died in 1895. The body was embalmed with formalin and, because relatives never claimed the remains, the body still resides there. The radiographs of

Figure 3.8 areas where the skull was relatively thin appear as holes in the three-dimensional reconstruction. The entire head of the humerus was also dropped from the reconstruction.
Figure 3.9A A conventional radiograph of the left shoulder of the mummy. Note the thin cortex (arrow) of the humeral head.

his skull, chest, and abdomen reveal organs that were in an excellent state of preservation. Other than foreign bodies in his mouth, nothing irregular was noted. The foreign bodies removed under endoscopic guidance turned out to be pennies that probably were deposited as "offerings" while the body had been at the funeral home. In addition, a nail was also recovered from under the tongue, but its significance could not be determined.

Since the organs were in such an exceptional state of preservation, it was hoped that a CT examination would contribute information regarding the health status of the individual

Figure 3.9B on the three-dimensional reconstruction, the entire humeral head (arrow) appeared to be fragmented.

Figure 3.10A A lateral Polaroid image of the upper portion of Andy's chest demonstrating an apparent fracture of the sternum (arrow).

prior to his death. Axial images of the lungs revealed adhesions that suggested an ongoing, active, infectious process at the time of the individual's death, possibly tuberculosis. Pulmonary lymph node lesions were also seen (Figure 3.12). Since the images were collected as a "volume" instead of "slice by slice," the volume was "reconstructed" into a coronal plane by a process called reformatting and demonstrated the adhesions in that plane (Figure 3.13).

Incorporated into all CT units is the ability to characterize tissue by CT numbers. A software function, termed region of interest (ROI), allows the CT number for a specific region of the image to be calculated. Using the cursor, an ROI in the shape of a circle or ellipse can be placed anywhere on the image. The computer will calculate the linear attenuation coefficient and the subsequent CT number for each pixel within the ROI. The display following completion of the data manipulation will calculate the area of the drawn ROI, the number of pixels included in that area, the average CT number, and the standard deviation of CT numbers (Figure 3.14). If there is a large standard deviation, such as that seen in ROI2, then there are probably several types of tissues with different densities, in this case bone and cartilage, included in the ROI. Therefore, for the most accurate data, the ROI deviation should not be too large. In both ROIs 1 and 2, the standard deviations were small.

The ROI may be applied to mummified remains and artifacts as well (Figure 3.15). Since few ROIs have been collected from organs or materials, such as resin (Figure 3.16) within mummified remains, at this time the exact significance of these data are not known. However, they can be compared to CT numbers from comparable living tissues to possibly

Figure 3.10B A lateral image of the lower portion of Andy's chest clearly showing the heart shadow (black arrows) and what appears to be a fractured rib (white arrow).

explain the effects of different types of desiccation and preservation processes. For example, the CT number for gray matter ranges from 32 to 44 and white matter from 24 to 36 (Webster 1988). In Sylvester, who was embalmed with arsenic, the ROI within the brain had an M of 84 with an SD of 6.5 (Figure 3.17A). However, the ROI from the brain of George/Fred, embalmed with a formaldehyde-based embalming fluid, had an M of 76 but

Figure 3.11 An axial image through the chest demonstrating a fractured rib (A) and a shift of the thoracic soft tissue structures to the right side of the chest cavity (B).
Figure 3.12 An axial image through the apical region of the thorax demonstrating a lesion (arrow) in right lung.

an SD of 70 (Figure 3.17B). Although both mean CT number values were similar to each other, they were certainly higher than that of living brain tissue. In addition, the higher standard deviation noted in George/Fred indicated that there was considerable variation within the ROI. Until more research is conducted in this area, the significance of the tissue values cannot be determined.

Figure 3.13 A coronal reconstruction of the chest showing an adhesion (arrow) between the right lung and thoracic cavity wall.

Figure 3.14 An axial section through the abdomen of a patient illustrating the calculation of three regions of interest (RoIs). The first RoI was in a pocket of gas within the stomach. The mean value, M, was -874.4 with a standard deviation, sD, of 33.9. the second RoI (arrow), a bit difficult to clearly see on the image, encompassed the costal cartilage of the rib. This region had an M of 906.9 and sD of 303.5. the third RoI was located in the spleen and had an M of 55.0 and sD of 4.1.

Figure 3.14 An axial section through the abdomen of a patient illustrating the calculation of three regions of interest (RoIs). The first RoI was in a pocket of gas within the stomach. The mean value, M, was -874.4 with a standard deviation, sD, of 33.9. the second RoI (arrow), a bit difficult to clearly see on the image, encompassed the costal cartilage of the rib. This region had an M of 906.9 and sD of 303.5. the third RoI was located in the spleen and had an M of 55.0 and sD of 4.1.

The most valuable CT information is frequently obtained from sectional images. An example is from a sideshow mummy that can be found at Ye Old Curiosity Shop in Seattle, Washington. The mummy, known as Sylvester, has the required fantastic story that accompanies all remains that traveled on the sideshow circuit. After being caught cheating during a card game, he was supposedly shot in the abdomen. Somehow he made it out of the Arizona saloon, got on his horse, and escaped. Unfortunately, some miles into the desert he fell off the horse and died. According to the legend, mummification was accomplished by the desert sand and the dry Arizona climate. His extended body is propped up in a display cabinet with a small cloth covering his pelvic area. Above the cloth is a round opening with a red margin and a small sign indicating that the opening was the entrance of the fatal bullet.

In 2001, a radiographic survey was conducted on the remains at the curiosity shop. The initial images revealed an incredible level of preservation of the brain, and all of the thoracic contents were easily identifiable. Soft tissue structures within the abdomen were obviously present, but few could be individually identified. There was no radiographic evidence of a bullet within the abdomen, and no exit wound was detected on the body. An endoscopic examination through the entrance wound did not reveal evidence of dried blood or damage to the underlying abdominal cavity structures (Figure 3.18).

An unexpected finding on the AP and lateral projections of the skull was what appeared to be pellets from a shotgun wound to the right side of his face (Figures 3.19A and 3.19B). A visual inspection of the face did reveal a number of raised "bumps," but provided no indication of what lay beneath (Figure 3.20). Obviously, the wound had occurred quite a while prior to the individual's death since the bumps were well healed over.

Figure 3.15 An axial section through the neck of James Penn aka Stoneman Willy showing three regions of interest (ROI). The first region, within the coins lodged in the oropharynx, had a mean value of 4000, the maximum value on the CT number scale. The second region was within the trachea. The CT number for air is -1000; however, the mean value for this region was -711.54, suggesting there was material, possibly dried mucus, in the section. The third region was located within the vertebral body and had a value of -76.68. The low value was due to the dehydration of tissue.

Figure 3.15 An axial section through the neck of James Penn aka Stoneman Willy showing three regions of interest (ROI). The first region, within the coins lodged in the oropharynx, had a mean value of 4000, the maximum value on the CT number scale. The second region was within the trachea. The CT number for air is -1000; however, the mean value for this region was -711.54, suggesting there was material, possibly dried mucus, in the section. The third region was located within the vertebral body and had a value of -76.68. The low value was due to the dehydration of tissue.

Therefore, two specific questions regarding Sylvester could possibly be answered using CT. First, was there evidence of a gunshot wound to the abdomen? And what are the precise placements of the pellets in the face? In 2007, a CT study was undertaken to address these questions. A GE MDCT was employed to collect the volume of the abdomen that included the gunshot wound. From that volume, axial, sagittal, and coronal sections through the wound were reconstructed (Figures 3.21A and 3.21B). No evidence of bleeding or damage to the underlying abdominal structures was demonstrated. The measurement function was also utilized to determine the size of the opening in the abdomen. The dimension was obtained from each of the three planes, and revealed that it was approximately V4 in. (0.64 cm) in diameter for each (Figures 3.22A, 3.22B, and 3.22C). If the wound had been premortem, there is little chance that it would be perfectly round. The dehydration process that was evident over the entire surface of the body would have altered the shape of the entrance wound. It appeared the opening was created postmortem to substantiate the legend.

The pellets beneath the face proved a more interesting problem. The axial sections provided precise information as to the location of each pellet, none of which appeared to have penetrated the bone (Figure 3.23). Several pellets were located in the oropharynx (Figure 3.24). Since none of the pellets seemed to have sufficient force to penetrate very deeply, the sections suggest that his mouth must have been open at the time of the

Figure 3.16 An axial section through the brain of the Egyptian mummy "Palb" at the level of the temporal mandibular joint showing two ROIs in the resin. The layering of the resin indicated that there had been two pourings. The difference in the ROI mean values revealed that the composition of each layer was not identical. In both ROIs, the large SD value suggested that the materials in each layer were not homogeneous.

Figure 3.16 An axial section through the brain of the Egyptian mummy "Palb" at the level of the temporal mandibular joint showing two ROIs in the resin. The layering of the resin indicated that there had been two pourings. The difference in the ROI mean values revealed that the composition of each layer was not identical. In both ROIs, the large SD value suggested that the materials in each layer were not homogeneous.

incident. For aesthetic reasons, a 3D reconstruction of the head was requested. The resulting images, obtained using a computed tomographic angiography (CTA) algorithm, were dramatic especially since the entire head could be rotated or tilted. However, the CTA algorithm automatically performed a function called smoothing when encountering the pellets. Each appeared to be more of a skin lesion rather than the location of a foreign body (Figure 3.25).

Another advantage of the volumetric acquisition of CT data is the possibility of slicing that acquired volume in any desired plane. An excellent example was demonstrated in another mummy by what appeared to be a gunshot wound in the distal right femur just above the knee. The mummy, known as George/Fred, was a sideshow mummy that belonged to Ripley's Believe It or Not organization. On the conventional radiographs, a defect was noted on the AP and lateral projections (Figures 3.26A and 3.26B), but it couldn't be well visualized. The initial CT examination collected axial sections of the region and clearly demonstrated the wound as it progressed through 12 axial slices (Figure 3.27). An obvious entrance was noted on the medial aspect, and similarly an exit was seen on the lateral aspect. Evidence of healing was also observed at the entrance, suggesting some time had passed since the wound was acquired. Unfortunately, because the wound was diagonal, it was not seen in a single axial section over its entire path. The volume was reformatted to create a section in the axial, sagittal, and coronal planes so that each would be entirely

Figure 3.17A An axial section through the brain of Sylvester at the level above the ventricles. The mummy had been preserved with an arsenic-based embalming fluid. The ROI had a mean value of 84 and a standard deviation (SD) of 6.5. The small SD indicated that the materials within the oval have similar values.

Figure 3.17A An axial section through the brain of Sylvester at the level above the ventricles. The mummy had been preserved with an arsenic-based embalming fluid. The ROI had a mean value of 84 and a standard deviation (SD) of 6.5. The small SD indicated that the materials within the oval have similar values.

Figure 3.17B An axial section through the brain of George/Fred at the level above the lateral ventricles. The mummy had been preserved with a formaldehyde-based embalming fluid. The first ROI, oval 1, had a mean CT number of 76, but an SD of 70; the tissue values were more varied in this mummy. The material posterior to the cerebral hemispheres was probably dried embalming fluid and had a mean value of -502.

Figure 3.18 Endoscopic view through the bullet hole, showing no interior anatomical damage.
Skull Shotgun Hank Iii
Figure 3.19A The lateral projection of sylvester's skull, demonstrating some of the shotgun pellets (arrows) embedded in the mummy's face.
Figure 3.19B The AP projection of sylvester's skull film was intentionally underexposed to more clearly visualize the relationship between the shotgun pellets (arrows) and the skull.

through the wound (Figures 3.28A and 3.28B). The off-axis reformatting used in this case further demonstrates the application of CT scanning with specific objectives in mind.

As previously mentioned, CT data can be complied in a manner so as to achieve a 3D image. But when is a 3D study necessary? An Egyptian mummy on display at the Barnum Museum in Bridgeport, Connecticut, provides an excellent example. A conventional radiographic examination was conducted at the museum. Once it was determined that the mummy was stable, it was transported by ambulance to a free-standing imaging center approximately 5 mi from the museum. Thin (0.6-mm-width) sections were collected of the head and pelvis. From the axial, sagittal, and coronal images, it was clear that the brain had not been removed by the traditional route, that is, through the cribriform plate as seen in some Egyptian mummies. Instead, the medial wall of the left orbit had been destroyed in order to access the cranial contents (Figure 3.29). A 3D image was desired to more dramatically demonstrate the defect. The technologist spent more than an hour applying various 3D algorithms until she achieved the desired results with the CTA algorithm (Figure 3.30).

Application of 3D reconstruction was also employed to confirm the sex of the Barnum mummy. The hieroglyphs on the coffin indicated that the mummy was a male priest known as PaIb. The lateral radiograph of the skull suggested a female (Figure 3.31), and

Figure 3.20 A photograph of the right side of Sylvester's face, showing some of the raised bumps (arrows).

the 3D surface renderings of the genital area confirmed that the remains were not male (Figure 3.32).

In addition to addressing specific questions, 3D reconstruction can be valuable for the aesthetic, nondiagnostic images that can be produced. The same technologist who spent so much time manipulating the Palb images expended even more time trying to find the correct algorithm for the 3D reconstruction of an extended ibis mummy (Figure 3.33). The mummified bird and 3D images were incorporated into an Egyptian exhibit at the Ripley's Believe It or Not Museum in New York City.

Because of the sectional presentation of the images and the ability to accurately measure structures, CT-guided biopsies are commonly performed medical procedures. In general, the value of biopsy of tissues or organs from mummified human remains is well established. Tissue biopsies can be used to answer a variety of research questions designed to add scientific knowledge or paleopathological data to a case (Cockburn et al. 1998). Tissue biopsies may be used to determine if organs are present. Taphonomic impact may result in morphological changes, making organs indistinguishable from surrounding tissues. Biopsy can help determine if the tissue collected is, in fact, from a specific organ. Biopsy has been instrumental in detecting evidence of any paleopathological processes that are present in the remains. Biopsy tissue samples have been rehydrated for histological examination as well as both qualitative and quantitative elemental analysis

Figure 3.21B the sagittal section through the track of the hole (arrow) in the abdomen.
Figure 3.22A The axial section demonstrating an identical measurement, 0.656 cm (0.258 in.), of the external and internal hole in the abdomen.

(Aufderheide 2003). A variety of tissues have proved useful in answering anthropological and paleopathological questions regarding mummified human remains. Hair, skin, muscle, various organs, and associated artifacts have all yielded substantial data for bio-anthropological studies.

There are several methods currently employed to attain biopsied material for analysis. All of these methods are destructive in that they remove material from the mummified remains, thereby altering the body or the anatomical context permanently. Among these, anatomical dissection (mummy autopsy) is the most reliable method of obtaining biopsies

Figure 3.22B The sagittal section with a measurement, 0.586 cm (0.231 in.), of the hole in the abdomen.

xMon 420 IWnrnl-IIOISnVtl t 4

Figure 3.22C a measurement, 0.586 cm (0.231 in.), of the hole from one of the coronal sections.

that are from the correct anatomical target location and are of a volume that can yield the most data (Aufderheide 2003). Full anatomical dissection "disassembles" the remains and alters the original morphological context. It is important to note that the material resulting from such dissections does not disappear; rather, the material is appropriately

150CC Optl 300

C -245

iter

A

INLAMD PACIFIC 1 MAG IMG

Sensation 16

/1&50 M 55 Y

'791 OB

^ H-SP-CR

Figure 3.23 An axial section through the skull demonstrating the precise position of one of the pellets (A) just beneath the skin. The excellent state of preservation is evident from the clearly discernable cerebellum (B).

Figure 3.24 An axial section through the skull with a pellet (A) resting against the ramus of the right mandible and another (B) within the left oropharynx.
Sylvester The Mummy
Figure 3.25 A 3D reconstruction demonstrating the position of the shotgun pellets. Some of the pellets have been identified with arrows.
Figure 3.26A The AP projection of George/Fred's left knee with diagonal radiopaque line (arrows) that may be the margin of the hole through the distal femur.

stored for future scientific analysis. Short of a full autopsy, a target organ or tissue resection may be conducted as well, thus preserving the remainder of the mummified remains for future research. Another method of tissue biopsy is needle biopsy. This procedure may be conducted "blind" by using anatomical landmarks, or by using direct visualization with endoscopic techniques. Another great potential for site-specific tissue biopsy is seen in the computed tomography-guided percutaneous needle biopsy (CTPNB) method. This technique was reported in the case of an arsenic-embalmed sideshow mummy known as "Marie O'Day" (Conlogue et al. 2005; Figures 3.34A and 3.34B). The research employed a nongravity coaxial needle biopsy method without aspiration. Although the technique was well demonstrated, not enough tissue was extracted for analysis. Another study (Ruhli et al. 2002) reported the applicability of the CTPNB method to extract tissue adjacent to the vertebral spine. Using this method requires moving the mummified human remains to the center where the CT scanner is located, risking reorientation of internal structures during transit. The report also described a gravity-dependent approach to tissue collection through a bone aspiration biopsy needle. The gravity-dependent positioning is well suited for an imaging center setting, where the disruption of internal body context has already been altered. However, the method is not suited for field biopsy as a CT scanner is difficult

Figure 3.26B The lateral projection of the knee showing the radiolucent entry (A) and exit (B) wounds.
Figure 3.27 An axial section demonstrating the entry wound on the medial aspect of the knee. The "C"-shaped structure within the wound (arrow) is an indication that healing had begun prior to death.
Figure 3.28A an off-axis axial section through the entire track of the wound.

and costly to transport to the field. The fact that many (not all) mummified remains in situ are interred in a horizontal orientation resting on the floor of an earthen tomb often within a burial coffin or in a seated position makes gravity-dependent biopsy impractical.

We undertook a study to determine if tissue biopsy could be conducted in a non-gravity-dependent position. The research question was addressed by replicating the described CTPNB method, but substituting the gravity-dependent biopsy procedure with a non-gravity-dependent needle scrape/aspiration technique modeled after the transthoracic needle aspiration biopsy employed in clinical medicine (Beckett et al. 2008).

The subject of this study was George/Fred, a late 19th to early 20th century mummified male (described in Chapter 2). The mummy's external and internal preservation was very good, making the subject an excellent choice for organ tissue biopsy. For this study, the liver was targeted for biopsy as its radiographic density indicated that the hepatic tissue was easily recognizable and accessible for biopsy.

Figure 3.28B an off-axis coronal section through the entire track of the wound.

Figure 3.29 Two coronal sections showing the destruction of the medial and superior wall of the left orbit (arrows).
Figure 3.30 A 3D reconstruction from the interior perspective of the skull showing the route of brain extraction (A) and the mass of resin (B) in the occipital region.
Figure 3.31 A sagittal section through Palb's skull showing the lack of a distinctive browridge (arrow).
Figure 3.32 A 3D reconstruction of the pelvic region revealing what appears to be labia (arrows) and a lack of male genitalia.
Figure 3.33 (A) A 3D reconstruction of an Egyptian ibis mummy in an extended position. Due to the reconstruction algorithm, the mummy's wrappings have a reddish appearance. (B) A closer view of the body cavities with some of the wrapping removed by image manipulation.

A standard CT-guided percutaneous needle biopsy method was employed to determine if a needle scrape/aspiration technique would yield a useable sample. The subject was in a supine position. To assess the potential of a non-gravity-dependent tissue sampling method, biopsies were attempted from an anterior-lateral approach. A 14-gauge coaxial needle was introduced intercostally on the left side at approximately the anterior axillary line using the CT axial images to guide the needle introduction (Figure 3.35). Due to the durable nature of the mummified liver, the standard coaxial biopsy was unable to attain a sample. In fact, the tip of the coaxial needle was bent and rendered useless for the procedure (Figure 3.36). An 8-gauge bone marrow biopsy aspiration needle was introduced, again under CT guidance, to the intrahepatic region. The cutting inner can-nula was removed leaving the outer cannula in place as a guide. A new coaxial needle

SEQ 9 PRE BIOPSY

Was this article helpful?

+1 0

Post a comment