A modified deep learning model for binary segmentation of renal pathologies

A modified deep learning model for binary segmentation of renal pathologies in computed tomography (CT) scans has been developed. The relevance of this study stems from the need to improve the accuracy and efficiency of anomaly detection during primary diagnosis, particularly given the high workload faced by medical institutions.

The aim of the study is to create a universal system for the automatic detection and localization of renal pathologies without prior classification of their morphological type.

Methods. The model is based on a modified U-Net architecture featuring five levels in both the encoder and decoder. Key modifications included replacing max pooling with strided convolutions (stride of two), implementing group normalization, and incorporating group normalization and dropout regularization. The model was trained on a clinical dataset comprising over 400 contrast-enhanced multidetector computed tomography (MDCT) scans. Preprocessing included conversion of intensities into Hounsfield units, range-free normalization, and data augmentation. A hybrid loss function combining the Dice coefficient and binary cross-entropy was used during training.

Results. A weighted average Dice coefficient of 0.85 was achieved on an independent test set. The model demonstrates high segmentation accuracy across various types of pathologies, including cysts (including those classified as complex under the Bosniak system), solid tumors, and mixed structures. The model maintains robustness in the presence of anatomical anomalies (such as horseshoe kidney, renal ectopia) when dealing with small lesions (less than 5 mm in size).

Conclusions. The proposed model offers high versatilityand applicability in routine clinical practice. Its ability to detect a wide range of renal pathologies without prior classification enhances its clinical significance and suitability for screening purposes.

Application Area. The results can be utilized in oncourology for initial CT scan analysis, surgical planning, and training of medical professionals. Integrating the model into medical image archiving and communication systems for the automated preprocessing of imaging studies holds significant promise.

1. РRossijskoe obshchestvo onkourologov i dr. Klinicheskie rekomendacii. Rak parenhimy pochki. Ministerstvo zdravoohraneniya Rossijskoj Federacii. Moskva, 2023. Available at: omnidoctor. ru/upload/iblock/3f9/ucp3yrezax5ecu0d66dfuz1pvjl597yb.pdf. Accessed 05.05.2024. (In Russ.)

2. Zlokachestvennye novoobrazovaniya v Rossii v 2023 godu (zabolevaemost' i smertnost'). A.D. Kaprin i dr., editors. M.: MNIOI im. P.A. Gercena – filial FGBU «NMIC radiologii» Minzdrava Rossii, 2024. 276 s. (In Russ.)

3. The 2023 Kidney and Kidney Tumor Segmentation Challenge (KiTS23). University of Minnesota Robotics Institute (MnRI); Helmholtz Imaging at DKFZ; Cleveland Clinic’s Urologic Cancer Program. 2023. Available at: kits-challenge.org/kits23. Accessed 15.05.2024.

4. Hatamizadeh A, Tang Y, Nath V, Yang D, et al. UNETR: Transformers for 3D Medical Image Segmentation. Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). 2022: 574-584.

5. Karpov OE, Subbotin SA, SHishkanov DV. Ispol'zovanie medicinskih dannyh dlya sozdaniya sistem podderzhki prinyatiya vrachebnyh reshenij. Sistemy podderzhki prinyatiya reshenij. 2019; 2: 16-23. (In Russ.)

6. Postanovlenie Pravitel'stva Rossijskoj Federacii ot 11 dekabrya 2012 g. №1119 «Ob utverzhdenii Trebovanij k zashchite personal'nyh dannyh pri obrabotke v informacionnyh sistemah personal'nyh dannyh». Sobranie zakonodatel'stva Rossijskoj Federacii. 2012; 50. St.7347. (In Russ.)

7. Postanovlenie Pravitel'stva Rossijskoj Federacii ot 21 marta 2012 g. №211 «O perechne mer, napravlennyh na obespechenie vypolneniya obyazannostej, predusmotrennyh Federal'nym zakonom „O personal'nyh dannyh“» (red. ot 06.09.2014). Sobranie zakonodatel'stva Rossijskoj Federacii. 2012; 12. St. 1455. (In Russ.)

8. GOST R 55036–2012. Informatizaciya zdorov'ya. Psevdonimizaciya. Metody i rekomendacii / ISO/TS 25237:2008. Vved. 2013–07–01. M.: Standartinform, 2012. 16 р. (In Russ.)

9. Federal'nyj zakon ot 27.07.2006 № 152-FZ «O personal'nyh dannyh» (s izm. i dop. ot 08.08.2024). Sobranie zakonodatel'stva Rossijskoj Federacii. 2006; 31. St.3451. (In Russ.)

10. Medlo AA, Kobcova TV, Bashkirova LV, Logovoj YUN. Metody diagnostiki i ocenki stepeni rasprostranennosti onkologicheskih zabolevanij. Prakticheskaya onkologiya. 2017; 18(1): 322-329. (In Russ.)

11. Ronneberger O, Fischer P, Brox T. U-Net: Convolutional Networks for Biomedical Image Segmentation. Proceedings of the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). 2015: P.234-241.

12. Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R. Dropout: A Simple Way to Prevent Neural Networks from Overfitting. Journal of Machine Learning Research (JMLR). 2014; 15: 1929- 1958.

13. YUrova AS. Metody avtomatizirovannoj segmentacii KT-izobrazhenij bryushnoj polosti. [dissertation] Moskva, 2018. 146 h. (In Russ.)