Intelligent system for supporting physician decision-making in skin neoplasm diagnostics based on dermatoscopic image analysis

The purpose of the research is to to develop an intelligent decision support system for physicians based on automated analysis of dermatoscopic images using machine learning algorithms, designed for early diagnostics and detection of malignant skin neoplasms. The development of an intelligent system for supporting physicians' decision making for both specialized specialists and general practitioners and nursing staff performing primary examination of patients with skin neoplasms is a research relevant area.

Methods. The intelligent system architecture for supporting doctors' decision-making based on the analysis of dermatoscopic images is proposed. The configuration used is a network approach based on the client-server mode. The client is a web application implementing the doctor's personal account functionality. This server hosts a cloud infrastructure that collects, stores and analyzes dermatoscopic images, and also maintains a report on the nosological group of skin lesions. In the analyzing process dermatoscopic images, machine learning methods are used based on the neural network’s usage with the virtual transformer architecture and a formed set of dermatoscopic images.

Results. The developed intelligent system for supporting physician decision-making has been practically implemented and tested in clinical conditions. It is characterized by accuracy values exceeding 93% for the Accuracy indicator and 89% – F-measure at the training stage and more than 89% (Accuracy indicator) during medical examinations. The obtained values of experimental assessments made it possible to formulate recommendations for integrating the developed intelligent system for supporting physician decision-making into the work processes of medical institutions.

Conclusion. The developed system provides automated image analysis, metadata structuring, visualization of model predictions and the possibility of expert marking and can be used not only by specialized doctors during medical examinations and studies, but also by general practitioners and mid-level medical personnel during screening examinations, mobile preventive appointments and medical examinations.

1. Vestergaard M.E., Macaskill P., Holt P.E., Menzies S.W. Dermoscopy compared with naked eye examination for the diagnosis of primary melanoma: a meta-analysis of studies performed in a clinical setting. British Journal of Dermatology. 2008;159(3):669–676. https://doi.org/10.1111/j.1365-2133.2008.08713.x

2. Jairath N., Pahalyants V., Shah R., Weed J., Carucci J.A., Criscito M.C. Artificial Intelligence in Dermatology: A Systematic Review of Its Applications in Melanoma and Keratinocyte Carcinoma Diagnosis. Dermatologic Surgery. 2024;50(9):791–798. https://doi.org/10.1097/DSS.0000000000004223

3. Brinker T.J., Hekler A., Enk A.H., Berking C., Haferkamp S., Hauschild A., Weichenthal M. Deep neural networks are superior to dermatologists in melanoma image classification. European Journal of Cancer. 2019;119:11–17. https://doi.org/10.1016/j.ejca.2019.05.023

4. Adamson A.S., Smith A. Machine Learning and Health Care Disparities in Dermatology. JAMA Dermatology. 2018;154(11):23–48. https://doi.org/10.1001/jamadermatol.2018.2348

5. Salinas M.P., Sepulveda J., Hidalgo L., Peirano D., Morel M., Uribe P. A systematic review and meta-analysis of artificial intelligence versus clinicians for skin cancer diagnosis. NPJ Digital Medecine. 2024;7(125):1–23. https://doi.org/10.1038/s41746-024-01103-x

6. Kozachok A.V., Spirin A.A., Elezkiy K.V., Kozachok E.S. A platform for collecting dermatoscopic images of patients’ neoplasms. Trudy Instituta sistemnogo programmirovaniya RAN = Proceedings of the Institute for System Programming of the Russian Academy of Sciences. 2024;36(3): 259–272. (In Russ.) https://doi.org/10.15514/ISPRAS-2024-36(3)-18

7. Suryani S., Nurdiansah N., Faizal F., Nirwana N., Metekohy A. UI/UX Design Of Mobile-Based Pharmacy Application Using Design Thinking Method. Journal of Computer Networks, Architecture and High Performance Computing. 2023;5(2):714–723. https://doi.org/10.47709/cnahpc.v5i2.2811

8. Familoni B.T., Babatunde S.O. User experience (UX) design in medical products: theoretical foundations and development best practices. Engineering Science & Technology Journal. 2024;5(3):1125–1148. https://doi.org/10.51594/estj.v5i3.975

9. Yu E.S., Cha J.M., Lee T., Kim J., Mun D. Features recognition from piping and instrumentation diagrams in image format using a deep learning network. Energies. 2019;12(23):1–19. https://doi.org/10.3390/en12234425

10. Devarakonda N., Murthy M., Reddy R., Harsha P.S. Skin Cancer Detection with Metadata Using Deep Learning Strategies. Advances in Communication and Applications. (ERCICA 2023). Singapore: Springer; 2024. P. 217–233. https://doi.org/10.1007/978-981-99-7633-1_16

11. Kozachok A.V., Spirin A.A., Samovarov O.I., Kozachok E.S. Application of machine learning models for multiclass classification of dermatoscopic images of skin neoplasms. Trudy Instituta sistemnogo programmirovaniya RAN = Proceedings of the Institute for System Programming of the Russian Academy of Sciences. 2024; 36(5):241–252. (In Russ.) https://doi.org/10.15514/ISPRAS-2024-36(5)-17

12. Khabarova R.I., Kulyova S.A. Artificial intelligence in the diagnosis of benign skin tumors in pediatric patients. Neural network integration into a mobile application. Voprosy onkologii = Oncology Issues. 2022; 68(6):820–826. (In Russ.) https://doi.org/10.37469/0507-3758-2022-68-6-820-826

13. Kamrul H., Asif A., Choon H.Y., Guang Y. A survey, review, and future trends of skin lesion segmentation and classification. Computers in Biology and Medicine. 2023;155:1– 36. https://doi.org/10.1016/j.compbiomed.2023.106624

14. Yousefi S., Najjar-Ghabel S., Danehchin R., Band S.S., Hsu C.C., Mosavi A. Automatic melanoma detection using discrete cosine transform features and metadata on dermoscopic images. Journal of King Saud University-Computer and Information Sciences. 2024;36(2):101944. https://doi.org/10.1016/j.jksuci.2024.101944