I.I. Gulinskaya1, N.Yu. Ilyasova2, K.E. Delov3, N.S. Demin4, E.N. Alekhin5

1–4 Samara National Research University (Samara, Russia)
2, 4 Image Processing Systems Institute, NRC «Kurchatov Institute» (Moscow, Russia)
5 Tyumen State Medical University (Tyumen, Russia)
1 Gira.2004@mail.ru, 2 ilyasova.nata@gmail.com, 3 delov_konstantin@mail.ru, 4 volfgunus@gmail.com, 5 a-eduard-n@yandex.ru

The development of modern medical imaging techniques and the introduction of artificial intelligence into clinical practice create the prerequisites for improving the diagnosis of pathological conditions, including tumor and metastatic lesions. One of the key tasks of radionuclide diagnostics is the precise localization of foci of pathological accumulation of radiopharmaceuticals, which requires a high level of expertise and may be associated with subjective errors of interpretation. This determines the need to develop auxiliary tools based on neural networks that can improve the accuracy and objectivity of diagnostics.

The purpose of the work – development of an algorithm for automated detection of foci of pathological accumulation of a radiopharmaceutical based on neural networks for use as a second opinion in the interpretation of radionuclide studies.

A computational experiment was conducted using trained models based on test data, during which the accuracy of detecting pathological foci was achieved at the level of 89%. The obtained results demonstrate the prospects of using neural networks to automate the analysis of radionuclide images.

The proposed algorithm can be implemented in the clinical practice of nuclear medicine departments as an auxiliary tool to improve diagnostic accuracy, reduce the burden on radiologists and improve the quality of interpretation of research.

Gulinskaya I.I., Ilyasova N.Yu., Delov K.E., Demin N.S., Alekhin E.N. Technology for detecting foci of pathological accumulation of a radiopharmaceutical drug based on neural networks. Biomedicine Radioengineering. 2025. V. 28. № 5. P. 145–148. DOI: https:// doi.org/10.18127/j15604136-202505-29 (In Russian)

1. Nathan M., Gnanasegaran G., Adamson K., Fogelman I. Bone Scintigraphy: Patterns, Variants, Limitations and Artefacts. Radionuclide and Hybrid Bone Imaging. 2013. V. 1. P. 377–408.
2. Hajianfar G., Sabouri M., Salimi Y., Amini M., Bagheri S., Jenabi E., Hekmat S., Maghsudi M., Mansouri Z., Khateri M., Jamshidi M.H., Jafari E., Rajabi A.B., Assadi M., Oveisi M., Shiri I., Zaidi H. Artificial intelligence-based analysis of whole-body bone scintigraphy: The quest for the optimal deep learning algorithm and comparison with human observer performance. Zeitschrift fur medizinische Physik. 2024. V. 34(2), P. 242–257.
3. Zhao Z., Pi Y., Jiang L, Xiang Y., Wei J., Yang P., Zhang W., Zhong X., Zhou K., Li Y., Li L., Yi Z., Cai H. Deep neural network based artificial intelligence assisted diagnosis of bone scintigraphy for cancer bone metastasis. Scientific Reports. 2020. V. 10(1). P. 51–65.
4. Hsieh T.-C., Liao C.-W., Lai Y.-C., Law K.-M., Chan P.-K., Kao C.-H. Detection of Bone Metastases on Bone Scans through Image Classification with Contrastive Learning. Journal of Personalized Medicine. 2021. V. 11(12). P. 2–13.
5. Aslantaş A., Dandıl E., Çakıroğlu M. CADBOSS: A computer-aided diagnosis system for whole-body bone scintigraphy scans. Journal of Cancer Research and Therapeutics. 2016. V. 12. № 2. P. 787–792.
6. Nasef M. M., Eid F. T., Amin M., Mausad A. An efficient segmentation technique for skeletal scintigraphy image based on sharpness index and salp swarm algorithm. Biomedical Signal Processing and Control. 2023. V. 79. № 1.
7. Jeens P., Freyer R. Synergetic relaxation labelling algorithm for segmentation of SPECT images using a connective model. Proceedings of 17th International Conference of the Engineering in Medicine and Biology Society (Montreal, QC, Canada). 1995.
8. Sajn L., Kukar M., Kononenko I., Milcinski M. Computerized segmentation of whole-body bone scintigrams and its use in automated diagnostics. Computer Methods and Programs in Biomedicine. 2005. V. 80. № 1. P. 47–55.
9. Šajn L., Kononenko I., Milčinski M. Computerized segmentation and diagnostics of whole-body bone scintigrams. Computerized Medical Imaging and Graphics. 2007. V. 31. № 7. P. 531–541.
10. Calin M.A., Elfarra F.-G., Parasca S.V. Object-oriented classification approach for bone metastasis mapping from whole-body bone scintigraphy. Physica Medica. 2021. V. 84. № 9. P. 141–148.