Digital colorimetricity and deep machine learning in the diagnostics of biotissues damage
DOI:
https://doi.org/10.31649/1681-7893-2026-51-1-139-148Keywords:
forensic medicine, multispectral imaging, physics-informed neural networks, neuro-fuzzy systems, deep learning, color space, hemoglobin degradation kineticsAbstract
This paper investigates a critical challenge in forensic medical examination and clinical dermatology: the automated and objective estimation of the age of superficial biological tissue injuries based on digital and multispectral imaging. Traditional empirical approaches, grounded in the visual assessment of color dynamics in damaged skin areas, exhibit critically low precision and an excessive dependence on the expert's subjective experience, ambient lighting conditions, and the patient's physiological characteristics. The study substantiates and develops a comprehensive innovative paradigm that conceptually integrates the mathematical apparatus of digital colorimetry within the perceptually uniform CIELAB space with advanced deep machine learning architectures. A mathematical model of a neuro-fuzzy classification system is developed in detail to effectively simulate the spatio-temporal evolution of trauma, accounting for the inherent fuzziness of boundaries between the stages of blood degradation. Particular emphasis is placed on the implementation of Physics-Informed Neural Networks, which enable the integration of a system of partial differential equations – describing the biochemical kinetics of hemoglobin decay and bilirubin diffusion – directly into the multi-component loss function of the neural network. A profound analysis is conducted on the utilization of state-of-the-art semantic segmentation architectures for extracting complex spectral-spatial features. Furthermore, an optimized hardware-software pipeline is presented for the deployment and iterative training of models using Tensor Processing Units (TPUs) in a cloud environment. Based on hyperspectral analysis, it is demonstrated that the transition from heuristic approaches to end-to-end regression of injury age, considering physical constraints, significantly reduces the Mean Absolute Error and elevates the scientific evidence base of forensic medical expertise to a fundamentally new level of reliability.
References
Tirado, J., Mauricio, D., 2021. Bruise dating using deep learning. Journal of Forensic Sciences 66, 336–346. https://doi.org/10.1111/1556-4029.14578
Saini, B., Venkatesh, D., Ganesh, A., Parameswaran, A., Patil, S., Kamat, P., Choudhury, T., 2024. Colorizing Multi-Modal Medical Data: An Autoencoder-based Approach for Enhanced Anatomical Information in X-ray Images. EAI Endorsed Trans Perv Health Tech 10. https://doi.org/10.4108/eetpht.10.5540
Guler, C., Kartal, E., Ünal, S., Aybar, G., Arslan, M., Gökalp, M.A., Asirdizer, M., 2026. Investigation of the Aging of Post-Traumatic Bruises Using Traditional and Computerized Digital Color Comparison Methods. Eastern J Med 31, 53–63. https://doi.org/10.5505/ejm.2026.70094
Rasel, M.A., Kareem, S.A., Obaidellah, U., 2024. Integrating color histogram analysis and convolutional neural networks for skin lesion classification. Computers in Biology and Medicine 183, 109250. https://doi.org/10.1016/j.compbiomed.2024.109250
Keshavarz, A., Bieber, G., Wulff, D., Babian, C., Lüdtke, S., 2026. Age Prediction of Hematoma from Hyperspectral Images Using Convolutional Neural Networks. J. Imaging 12, 78. https://doi.org/10.3390/jimaging12020078
Al-Arami, S., Lüdtke, St., Dreßler, J., Babian, C., 2025. Age prediction of hematoma using hyperspectral imaging (HSI). Forensic Sci Med Pathol 21, 1699–1708. https://doi.org/10.1007/s12024-025-01076-7
Sharma, V., Rani, R., Pathak, P., Chakraborty, S., 2024. Epidermis Bruise Analysis Using Deep Learning, in: 2024 Sixth International Conference on Computational Intelligence and Communication Technologies (CCICT). Presented at the 2024 Sixth International Conference on Computational Intelligence and Communication Technologies (CCICT), IEEE, Sonepat, India, pp. 324–330. https://doi.org/10.1109/CCICT62777.2024.00060.
Nesheli, S.J., Sabet, M., Koozari, A., Mirzaghavami, P., Eftekhar, A., Elhaie, M., Rouhi, S., Lariche, N.J., Abidi, M., Rezaeijo, S.M., 2025. Transformer-based and CNN-based models for clinically effective 2D and 3D pelvic bone segmentation in CT imaging. BMC Musculoskelet Disord 27, 68. https://doi.org/10.1186/s12891-025-09444-8
Semantic Segmentation with PyTorch: U-NET from scratch. Available at: https://medium.com/@alessandromondin/semantic-segmentation-with-pytorch-u-net-from-scratch-502d6565910a (accessed 05.03.2026).
Mehdar, K.M., Soomro, T.A., Ali, A., Bin Ubaid, F., Irfan, M., Halawani, H.T., Mashraqi, A.M., Mohammed Elshafie, S.E., Asiri, A.A., Abdelkafi Magzoub, M., 2026. Deep neural network-based robust framework for automated skin lesion segmentation and analysis. Digital Health 12, 20552076261427501. https://doi.org/10.1177/20552076261427501.
Shad, R., Cunningham, J.P., Ashley, E.A., Langlotz, C.P., Hiesinger, W., 2021. Designing clinically translatable artificial intelligence systems for high-dimensional medical imaging. Nat Mach Intell 3, 929–935. https://doi.org/10.1038/s42256-021-00399-8
Hou, J., Che, Y., Fang, Y., Bai, H., Sun, L., 2024. Early Bruise Detection in Apple Based on an Improved Faster RCNN Model. Horticulturae 10, 100. https://doi.org/10.3390/horticulturae10010100
Le, P.T., Pham, B.-T., Chang, C.-C., Hsu, Y.-C., Tai, T.-C., Li, Y.-H., Wang, J.-C., 2023. Anti-Aliasing Attention U-net Model for Skin Lesion Segmentation. Diagnostics 13, 1460. https://doi.org/10.3390/diagnostics13081460
Carini, L., Jensen, M., Nürnberg, R., 2023. Deep learning for gradient flows using the Brezis–Ekeland principle. Arch. Math. (Brno) 249–261. https://doi.org/10.5817/AM2023-3-249
Zhao, A., Fattahi, D., Hu, X., 2025. Physics-informed neural networks for physiological signal processing and modeling: a narrative review. Physiol. Meas. 46, 07TR02. https://doi.org/10.1088/1361-6579/adf1d3.
Kucuk, A., Finlayson, G.D., Mantiuk, R., Ashraf, M., 2023. Performance Comparison of Classical Methods and Neural Networks for Colour Correction. J. Imaging 9, 214. https://doi.org/10.3390/jimaging9100214
Skin Lesion Segmentation using Unet – Kaggle. Available at: https://www.kaggle.com/code/hashbanger/skin-lesion-segmentation-using-unet (accessed 05.03.2026).
Downloads
-
pdf (Українська)
Downloads: 0
Published
How to Cite
Issue
Section
License
Автори, які публікуються у цьому журналі, погоджуються з наступними умовами:- Автори залишають за собою право на авторство своєї роботи та передають журналу право першої публікації цієї роботи на умовах ліцензії Creative Commons Attribution License, котра дозволяє іншим особам вільно розповсюджувати опубліковану роботу з обов'язковим посиланням на авторів оригінальної роботи та першу публікацію роботи у цьому журналі.
- Автори мають право укладати самостійні додаткові угоди щодо неексклюзивного розповсюдження роботи у тому вигляді, в якому вона була опублікована цим журналом (наприклад, розміщувати роботу в електронному сховищі установи або публікувати у складі монографії), за умови збереження посилання на першу публікацію роботи у цьому журналі.
- Політика журналу дозволяє і заохочує розміщення авторами в мережі Інтернет (наприклад, у сховищах установ або на особистих веб-сайтах) рукопису роботи, як до подання цього рукопису до редакції, так і під час його редакційного опрацювання, оскільки це сприяє виникненню продуктивної наукової дискусії та позитивно позначається на оперативності та динаміці цитування опублікованої роботи (див. The Effect of Open Access).
