Endoscopic investigation of the internal organs of a 15th-century child mummy from Yangju, Korea


  • The first two authors contributed equally to this study.

Dr Dong Hoon Shin, MD, PhD, Department of Anatomy, Seoul National University College of Medicine,Yongon-Dong 28, Seoul 110-799, South Korea. T: +82 2 740 8203; F: +82 2 6230 9160, E: drdoogi@snu.ac.kr; Dr M. J. Kim, Department of Anatomy, Dankook University College of Medicine, San 29, Anseo Dong, Cheonan-si, Chungnam 330-714, South Korea. T: +82 41 550-3853; E: mjukim99@dku.edu.


Our previous reports on medieval mummies in Korea have provided information on their preservation status. Because invasive techniques cannot easily be applied when investigating such mummies, the need for non-invasive techniques incurring minimal damage has increased among researchers. Therefore, we wished to confirm whether endoscopy, which has been used in non-invasive and minimally invasive studies of mummies around the world, is an effective tool for study of Korean mummies as well. In conducting an endoscopic investigation on a 15th-century child mummy, we found that well-preserved internal organs remained within the thoracic, abdominal and cranial cavities. The internal organs – including the brain, spinal cord, lung, muscles, liver, heart, intestine, diaphragm and mesentery – were easily investigated by endoscopy. Even the stool of the mummy, which accidentally leaked into the abdominal cavity during an endoscopic biopsy, was clearly observed. In addition, unusual nodules were found on the surface of the intestines and liver. Our current study therefore showed that endoscopic observation could provide an invaluable tool for the palaeo-pathological study of Korean mummies. This technique will continue to be used in the study of medieval mummy cases in the future.


We have already published several papers detailing our studies of medieval mummies from the Chosun Dynasty of Korea (1392–1910) (Shin et al. 2003a,b; Chang et al. 2006; Kim et al. 2006). As described in these previous reports, such mummies can provide invaluable data on the physical traits of medieval Korean people. However, it should be also noted that there are many difficulties in conducting studies on mummies, in that descendants are reluctant to permit their ancenstor/mummy to be the subject of invasive scientific investigation. Therefore, most medieval mummies that have undergone preliminary investigations in Korea have been reburied in other locations without sufficient further investigation.

We have therefore been interested in recent reports on the development of minimally invasive tools for investigating mummies. In other countries, researchers have attempted to develop alternative tools for examining mummies without causing unnecessary damage given the desire for the preservation of mummies. Fibreoptic endoscopy (Manialawi et al. 1978; Gaafar et al. 1999; Spigelman & Donoghue, 2003; Hagedorn et al. 2004) or high-quality multidetector computerized tomography (MDCT) (Melcher et al. 1997; Hoffman & Hudgins, 2002; Hoffman et al. 2002; Jansen et al. 2002; Cesarani et al. 2003, 2004; Hughes et al. 2005; Winder et al. 2005) may be effective techniques for investigating mummies without causing significant damage, or indeed any damage at all.

The advantages of these two techniques for non-invasive or minimally invasive examination cannot be matched by other methods. In the case of MDCT, recent progress in CT image quality and rendering techniques has enabled scientists to conduct investigations on mummies from the operator's viewpoint without dissection. Fibreoptic endoscopy can also be a useful tool for providing additional palaeo-pathological data on mummies, as small pathological changes in intestinal organs, not easily identified by MDCT, can be observed using this technique.

In this study we set out to determine whether fibreoptic endoscopy could also be used as an effective tool for obtaining data on the internal organs of medieval Korean mummies. If proven to be useful in providing satisfactory data on the preservation status of, or palaeo-pathological clues to, the internal organs of Korean mummies, this technique should be used in all future studies of Korean mummies.

Materials and methods

Our study was performed in accordance with the Vermillion Accord on Human Remains, World Archaeological Congress, South Dakota, 1989. With regard to the child mummy investigated in this study (Fig. 1A), our previous studies (Shin et al. 2003a,b; Kim et al. 2006) offered basic data.

Figure 1.

(A) Fifteenth-century child mummy. (B) Endoscopic examination. (C) A hole (red arrow) was made in the substernal region for inserting the fibreoptic endoscope into the body cavity.

In order to view the internal organs of the mummy while incurring minimal damage to the surface of his skin, we performed fibreoptic endoscopy at Dankook University Hospital (Fig. 1B). The fibreoptic endoscope (GIF XQ230, Olympus Co., Japan) was inserted through an incision made in the epigastric area of the mummy to examine the remains of the organs within the abdominal cavity (Fig. 1C). To investigate the thoracic cavity, an incision was made in the skin, on the right chest. The organs within the cranial cavity and vertebral canal were observed through holes made in the occipital region of the skull and on the lower back, respectively.

During endoscopic investigation, photographs and videos were taken to record our findings. The videos, recorded from analog signals, were reformatted to a digital multimedia file.


In the course of our endoscopic investigation, we observed dark-brown internal organs within the abdominal, thoracic, cranial and spinal cavities. Because all of the investigated organs were dehydrated or shrunken, we could not easily differentiate one from the other. When we inserted the endoscope into the abdominal cavity, the intestines could be identified (Fig. 2A–C). During a biopsy of the intestines, possible stool material escaped from the intestinal lumen (Fig. 2D).

Figure 2.

Intestine observed by fibreoptic endoscopy. (A–C) When the endoscope was inserted into the abdominal cavity, the intestine could be identified. Because the intestine was severely dehydrated or shrunken, subsections of intestine could not be differentiated. (D) When we performed a biopsy on the intestine, stool-like material (red arrow) spilled out from the intestinal lumen.

When the endoscope was turned upwards, the liver and diaphragm were observed. The multimedia file (see supplementary Fig. S1) shows our endoscopic findings on the intestine, liver and diaphragm. Unusual nodules were noted on the surface of the liver (Fig. 3A,B), but not on the adjacent diaphragm (Fig. 3C) and mesenteries (Fig. 3D). Similar nodules were also observed on the serosal surface of the intestinal wall (Fig. 3E,F).

Figure 3.

(A,B) Liver (Lv) and diaphragm (Di) observed by fibreoptic endoscopy. Unusual nodules were spread on the surface of the liver. The diaphragm (C) and mesenteries (D) did not exhibit such surface nodules. (E,F) Similar nodules were also observed on the serosal surface of the intestinal wall.

In the thoracic cavity, the remains of the lungs, ribs and intercostal muscles were clearly observed (Fig. 4A,B). Even a heart remnant was found within the thoracic cavity, although it was notably shrunken (Fig. 4C,D).

Figure 4.

Thoracic cavity observed by fibreoptic endoscopy. (A,B) Remants of the lung (Lu), ribs (asterisks) and intercostal muscles (ICM). (C,D) The heart remnant (Ht) was also found within the thoracic cavity, though the organ was markedly shrunken.

Finally, we investigated the brain and spinal cord within the skull and vertebral canal, respectively. In the cranial cavity, the contracted brain remnant was observed with heavy calcification-like substances spread over its surface (Fig. 5A). On inserting the endoscope into the hole made via a laminectomy of vertebra, the spinal cord was identified within the vertebral canal (Fig. 5B).

Figure 5.

Brain and spinal cord observed by fibreoptic endoscopy. (A) Within the skull, a shrunken brain remnant (Br) was observed with heavy calcification-like substances (asterisk) spread over its surface. (B) When the endoscope was inserted into the hole made by laminectomy of the vertebra, the spinal cord (Sp) was identified within the vertebral canal.


Medieval Koreans did not believe in resurrection or immortality of the soul. Instead, they thought that misfortune might accompany the incomplete disintegration of the corpses of their ancestors. Therefore, they chose graveyards in which their ancestors’ corpses and the accompanying burial materials would decay completely. In addition, medieval documents such as The National Code for the Five Rituals (Gukjooryeiui in Korean) or The Family Rituals of Chu-si (Jujagarye in Korean), by which the construction of tombs during the Chosun Dynasty was strictly regulated, do not contain any mention of intentional or non-intentional treatments for artificial mummification, and therefore mummification in Korea seems not to have involved artificial mummification processes. Given that the climate of Korea is not appropriate for natural mummification, the Korean mummification process has thus remained unresolved. Considering that Korean mummies have only been found in tombs where the lime–soil mixture barrier (Fig. 6) was not broken until they were found, we speculate that the separation of the inner space of the coffins from the outer space by the lime–soil mixture barrier may play a crucial role in mummification in these cases. However, the exact cause of mummification in Korea will be elucidated in forthcoming studies.

Figure 6.

Structure of a medieval tomb with lime–soil mixture, from the Chosun Dynasty, found in Hadong, Korea, in April 2006. The burial conditions associated with this case are representative of other medieval tombs, including those in which mummies have been found. (A) When the medieval tomb with lime–soil mixture (LSMB) was found, the coffin was found to be completely encapsulted by LSMB. The thickness of the LSMB is generally about 10 cm or more. (B) When the LSMB was broken open, the inner coffin (IC) and outer coffin (white circle) could be seen. (C) After the upper lid of the IC was removed, the buried mummy (asterisk), with various cultural remains including clothes or documents, was visible. (D) When the mummy and the cultural remains were moved, the basal plate (BP), on which holes (dotted yellow circles) representing the Great Dipper were punched, were exposed. (E) After the BP was removed, a charcoal layer (Ch) was visible. (F) Archaeological drawings (c) of the tomb seen from above. The red and blue dotted lines in (c) are virtual slice planes. The other drawings represent the coffin in (c) sectioned along the blue dotted line and seen horizontally (a); and sectioned along the red dotted line and seen horizontally (b). To show the structure of the basal side of the coffin more clearly, the part surrounded by the red rectangle in (a) is magnified (d). OC, outer coffin.

Whatever the principal cause of mummification in Korea, the development of tools for investigating mummies with minimal damage is crucial for researchers in Korea because descendants will not readily consent to invasive scientific investigations. The development of non-invasive or minimally invasive techniques for investigation of mummies has been actively pursued by researchers around the world. Two of the most successful techniques have been fibreoptic endoscopy and high-quality MDCT scanning. In the case of MDCT scans, virtual reconstructions of hard-to-access internal organs or heavily wrapped faces of Egyptian mummies have been conducted successfully (Melcher et al. 1997; Hoffman & Hudgins, 2002; Hoffman et al. 2002; Jansen et al. 2002; Cesarani et al. 2003, 2004; Hughes et al. 2005; Winder et al. 2005).

Although MDCT scans can provide invaluable data on the internal organs of mummies without any damage to the surface of the mummy, fibreoptic endoscopy has an advantage that is not easily attainable by MDCT: direct visualization of internal organs and their small pathological changes, which is crucial for deducing the pathophysiolocial states of the mummies. For example, the endoscopic study by Hagedorn et al. (2004) successfully showed dentogenic sinusitis and chronic middle-ear infections with intracranial perforation in mummies from Upper Egypt. Other nasal endoscopic trials by Gaafar et al. (1999) also showed evidence of brain removal during artifical mummification in ancient Egypt. Because they found an enlarged conduit between the cranial and nasal cavities in all of the mummies investigated, they concluded that brain removal in Egypt was likely to have been performed by well-trained personnel with specially designed instruments for endonasal craniotomy. Therefore, it was shown that endoscopic examinations could provide meaningful data on mummies that are not as easily acquired using other techniques (Manialawi et al. 1978; Spigelman & Donoghue, 2003).

Most descendants of Koreans whose bodies have been mummified are opposed to invasive examinations of their mummified ancestors, but using fibreoptic endoscopy to examine the internal organs of Korean mummies may turn out to be more acceptable, if endoscopy can be proved to be a useful non-invasive or minimally invasive technique. In this regard, our current endoscopic study on the internal organs of a previously examined child mummy could be the basis for future investigations into medieval Korean mummies.

We have already published a brief profile of the 15th-century child mummy that is the subject of the present study (Shin et al. 2003a,b; Kim et al. 2006). To summarize, the child mummy was found in a wooden coffin with the outermost lime–soil mixture barrier intact; the child had been alive at some time during the period from 1411 to 1442 ad; the child had died at about 4.5–6.6 years of age, possibly due to asphyxia by massive respiratory bleeding; and the histological conservation pattern was similar to those in cases reported from other countries.

Among the previously acquired data on the child mummy, the finding that is most relevant here is a radiographic one (Shin et al. 2003a). That is, plain X-ray and CT scans showed that the brain, spinal cord, liver, back muscles, and the residues of the spleen, heart and lung seemed to be well preserved within the body cavities (Shin et al. 2003a). However, because direct visualization of the preserved organs was not attempted in our previous studies, we subsequently performed an endoscopic study to reveal the state of preservation of the internal organs. In our current endoscopic study, we showed that the liver, heart, brain, spinal cord, diaphragm and mesentery were well preserved, despite considerable shrinkage of the organs. In the course of our investigation, clues to the palaeo-pathological changes in some of the mummified organs could be observed, and we confirmed the presence of unusual nodules on the surface of the liver and intestine. Given that there were no such nodules on any of the other organs, these nodules on the liver or intestinal surface might have been caused by unknown pathological changes resulting from, for example, secondary tumour metastasis, tuberculosis or hydatid disease. However, we are cautious in suggesting the possibility of pathological changes as the exact cause of the changes could not be clearly elucidated, and because the nodules might have been formed as a result of the mummification process. As fibreoptic endoscopy evidently is a useful tool for collecting palaeo-pathological clues in Korean mummies, endoscopic investigation will be performed in all future studies, if descendants allow minimal incisions to the skin to be made.


All of the investigations into and examinations of the mummy found in Yangju were superintended principally by the Seok Ju Seon Memorial Museum and the Institute for Oriental Studies, Dankook University. We thank, in particular, Ki Rok Choi, a producer for the Korean Broadcast System (KBS), for his devotion to the preservation of medieval Korean mummies. Figure 1(A) is included with the generous consent of the staff of KBS, who took the photograph. This study was supported by Korea Research Foundation Grant (KRF-2003-I01485-E00006).