Anatomy and Cell Biology

Korean Association of Anatomists (대한해부학회)
권호 표지
DOI QR코드

DOI QR Code

Convolutional neural network of age-related trends digital radiographs of medial clavicle in a Thai population: a preliminary study

  • Received : 2022.10.13
  • Accepted : 2022.12.05
  • Published : 2023.03.31
Download PDF
⟨ Previous Next ⟩

Abstract

Age at death estimation has always been a crucial yet challenging part of identification process in forensic field. The use of human skeletons have long been explored using the principle of macro and micro-architecture change in correlation with increasing age. The clavicle is recommended as the best candidate for accurate age estimation because of its accessibility, time to maturation and minimal effect from weight. Our study applies pre-trained convolutional neural network in order to achieve the most accurate and cost effective age estimation model using clavicular bone. The total of 988 clavicles of Thai population with known age and sex were radiographed using Kodak 9000 Extra-oral Imaging System. The radiographs then went through preprocessing protocol which include region of interest selection and quality assessment. Additional samples were generated using generative adversarial network. The total clavicular images used in this study were 3,999 which were then separated into training and test set, and the test set were subsequently categorized into 7 age groups. GoogLeNet was modified at two layers and fine tuned the parameters. The highest validation accuracy was 89.02% but the test set achieved only 30% accuracy. Our results show that the use of medial clavicular radiographs has a potential in the field of age at death estimation, thus, further study is recommended.

Keywords

Acknowledgement

This work was supported by the Faculty of Medicine, Chiang Mai University, grant no. 069-2565 for research funding. The authors are also gratefully thankful for the support from the Excellence Center in Osteology Research and Training Center (ORTC) with partial support from Chiang Mai University.

References

  1. Morgan OW, Sribanditmongkol P, Perera C, Sulasmi Y, Van Alphen D, Sondorp E. Mass fatality management following the South Asian tsunami disaster: case studies in Thailand, Indonesia, and Sri Lanka. PLoS Med 2006;3:e195.
  2. Srinak N. The study of unclaimed and unidentified bodies management in Thailand and other countries. J Thai Justice Syst 2020;13:139-52.
  3. Evert L. Unidentified bodies in forensic pathology practice in South Africa: demographic and medico-legal perspectives [thesis]. Hatfield: University of Pretoria; 2011.
  4. National Missing and Unidentified Persons System. The nation's silent mass disaster [Internet]. Washington, D.C.: National Institute of Justice [cited 2022 Oct 29]. Available from: https://namus.nij.ojp.gov/.
  5. Manzoor Mughal A, Hassan N, Ahmed A. Bone age assessment methods: a critical review. Pak J Med Sci 2014;30:211-5.
  6. Hermetet C, Saint-Martin P, Gambier A, Ribier L, Sautenet B, Rerolle C. Forensic age estimation using computed tomography of the medial clavicular epiphysis: a systematic review. Int J Legal Med 2018;132:1415-25. https://doi.org/10.1007/s00414-018-1847-z
  7. Shirley NR. Age and sex estimation from the human clavicle: an investigation of traditional and novel methods [PhD dissertation]. Knoxville: University of Tennessee; 2009.
  8. Schulz R, Muhler M, Mutze S, Schmidt S, Reisinger W, Schmeling A. Studies on the time frame for ossification of the medial epiphysis of the clavicle as revealed by CT scans. Int J Legal Med 2005;119:142-5. https://doi.org/10.1007/s00414-005-0529-9
  9. Marera DO, Satyapal KS. Fusion of the medial clavicular epiphysis in the South African and Kenyan populations. Int J Morphol 2018;36:1101-7. https://doi.org/10.4067/S0717-95022018000301101
  10. Milenkovic P, Djukic K, Djonic D, Milovanovic P, Djuric M. Skeletal age estimation based on medial clavicle--a test of the method reliability. Int J Legal Med 2013;127:667-76. https://doi.org/10.1007/s00414-012-0791-6
  11. Botha D, Lynnerup N, Steyn M. Age estimation using bone mineral density in South Africans. Forensic Sci Int 2019;297:307-14. https://doi.org/10.1016/j.forsciint.2019.02.020
  12. Kranioti EF, Bonicelli A, Garcia-Donas JG. Bone-mineral density: clinical significance, methods of quantification and forensic applications. Res Rep Forensic Med Sci 2019;9:9-21. https://doi.org/10.2147/RRFMS.S164933
  13. Rowe P, Koller A, Sharma S. Physiology, bone remodeling. StatPearls. Treasure Island: StatPearls Publishing; 2022.
  14. Seeman E, Delmas PD. Bone quality--the material and structural basis of bone strength and fragility. N Engl J Med 2006;354:2250-61. https://doi.org/10.1056/NEJMra053077
  15. Li Y, Huang Z, Dong X, Liang W, Xue H, Zhang L, Zhang Y, Deng Z. Forensic age estimation for pelvic X-ray images using deep learning. Eur Radiol 2019;29:2322-9. https://doi.org/10.1007/s00330-018-5791-6
  16. LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015;521:436-44. https://doi.org/10.1038/nature14539
  17. Szegedy C, Liu W, Jia Y, Sermanet P, Reed S, Anguelov D, Erhan D, Vanhoucke V, Rabinovich A. Going deeper with convolutions. Paper presented at: 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2015 Jun 7-12; Boston, USA. p. 1-9.
  18. Benito M, Sanchez JA, Codinha S. Age-at-death estimation based on radiological and image analysis methods in clavicle in a current Spanish population. Int J Legal Med 2014;128:523-33. https://doi.org/10.1007/s00414-014-0989-x
  19. Chantharawetchakun T, Vachirawongsakorn V. Age estimation in the Thai male population using epiphyseal union of the medial clavicle. Chiang Mai Med J 2021;60:149-55. https://doi.org/10.12982/CMUMEDJ.2021.13
  20. Schulz R, Muhler M, Reisinger W, Schmidt S, Schmeling A. Radiographic staging of ossification of the medial clavicular epiphysis. Int J Legal Med 2008;122:55-8. https://doi.org/10.1007/s00414-007-0210-6
  21. Schmeling A, Grundmann C, Fuhrmann A, Kaatsch HJ, Knell B, Ramsthaler F, Reisinger W, Riepert T, Ritz-Timme S, Rosing FW, Rotzscher K, Geserick G. Criteria for age estimation in living individuals. Int J Legal Med 2008;122:457-60. https://doi.org/10.1007/s00414-008-0254-2
  22. Subramanian S, Viswanathan VK. Bone age. StatPearls. Treasure Island: StatPearls Publishing; 2022.
  23. Fourcade A, Khonsari RH. Deep learning in medical image analysis: a third eye for doctors. J Stomatol Oral Maxillofac Surg 2019;120:279-88. https://doi.org/10.1016/j.jormas.2019.06.002
  24. Ott SM. Cortical or trabecular bone: what's the difference? Am J Nephrol 2018;47:373-5. https://doi.org/10.1159/000489672
  25. Navega D, Coelho JD, Cunha E, Curate F. DXAGE: a new method for age at death estimation based on femoral bone mineral density and artificial neural networks. J Forensic Sci 2018;63:497-503. https://doi.org/10.1111/1556-4029.13582
  26. Thurzo A, Kosnacova HS, Kurilova V, Kosmel S, Benus R, Moravansky N, Kovac P, Kuracinova KM, Palkovic M, Varga I. Use of advanced artificial intelligence in forensic medicine, forensic anthropology and clinical anatomy. Healthcare (Basel) 2021;9:1545.
  27. Guo YC, Han M, Chi Y, Long H, Zhang D, Yang J, Yang Y, Chen T, Du S. Accurate age classification using manual method and deep convolutional neural network based on orthopantomogram images. Int J Legal Med 2021;135:1589-97. https://doi.org/10.1007/s00414-021-02542-x
  28. Mutasa S, Sun S, Ha R. Understanding artificial intelligence based radiology studies: What is overfitting? Clin Imaging 2020;65:96-9. https://doi.org/10.1016/j.clinimag.2020.04.025
  29. Chartrand G, Cheng PM, Vorontsov E, Drozdzal M, Turcotte S, Pal CJ, Kadoury S, Tang A. Deep learning: a primer for radiologists. Radiographics 2017;37:2113-31. https://doi.org/10.1148/rg.2017170077
  30. Zhang P, Zhong Y, Li X. ACCL: adversarial constrained-CNN loss for weakly supervised medical image segmentation. arXiv. 2005.00328 [Preprint]. 2020 [cited 2022 Dec 6]. Available from: https://doi.org/10.48550/arXiv.2005.00328.
  31. Guan S, Loew M. Breast cancer detection using synthetic mammograms from generative adversarial networks in convolutional neural networks. J Med Imaging (Bellingham) 2019;6:031411.
  32. Wang Y, Zhou L, Wang M, Shao C, Shi L, Yang S, Zhang Z, Feng M, Shan F, Liu L. Combination of generative adversarial network and convolutional neural network for automatic subcentimeter pulmonary adenocarcinoma classification. Quant Imaging Med Surg 2020;10:1249-64. https://doi.org/10.21037/qims-19-982
  33. Dodge S, Karam L. Understanding how image quality affects deep neural networks. Paper presented at: 2016 Eighth International Conference on Quality of Multimedia Experience (Qo-MEX); 2016 Jun 6-8; Lisbon, Portugal. p. 1-6.
  34. Aggarwal A, Mittal M, Battineni G. Generative adversarial network: an overview of theory and applications. Int J Inf Manag Data Insights 2021;1:100004.
  35. Kazeminia S, Baur C, Kuijper A, van Ginneken B, Navab N, Albarqouni S, Mukhopadhyay A. GANs for medical image analysis. Artif Intell Med 2020;109:101938.
  36. Buda M, Maki A, Mazurowski MA. A systematic study of the class imbalance problem in convolutional neural networks. Neural Netw 2018;106:249-59. https://doi.org/10.1016/j.neunet.2018.07.011