Тип публикации: доклад, тезисы доклада, статья из сборника материалов конференций
Конференция: Software Engineering: Emerging Trends and Practices in System Development; On-line; On-line
Год издания: 2026
Аннотация: Detecting pathological processes is a crucial task for medical image processing and analysisThis paper focuses on classifying MRI images of the knee joint into two classes: healthy and showing signs of osteoarthritis. Convolutional neural networks based on the architectures of EfficientNetB5, InceptionResNetV2, MobileNetV2, MobileNПоказать полностьюet, and VGG16 were used. The research dataset comprises images of healthy and osteoarthritic knees obtained from patients in the Krasnoyarsk Territory, supplemented by publicly available images from the Kaggle repository. Inadequate number of images and class imbalance were addressed by dataset augmentation using various image rotations. The neural network models were compared based on accuracy scores. The best results were achieved by the EfficientNet-B5 (95.55%) and InceptionResNetV2 (94.39%) models. Both models showed a high degree of specificity and sensitivity, which makes them suitable as screening tools. The results obtained indicate the potential of using neural networks to automate the diagnosis of osteoarthritis in knee MRI scans. #CSOC1120.
Журнал: Software Engineering: Emerging Trends and Practices in System Development
Номера страниц: 381-390
Место издания: Cham