Background: We directly compared the mammography image processing results obtained with the help of the YOLOv4 convolutional neural network (CNN) model versus those obtained with the help of the NCA-based nested contours algorithm model. Method: We used 1080 images to train the YOLOv4, plus 100 images with proven breast cancer (BC) and 100 images with proven absence of BC to test both models. Results: the rates of true-positive, false-positive and false-negative outcomes were 60, 10 and 40, respectively, for YOLOv4, and 93, 63 and 7, respectively, for NCA. The sensitivities for the YOLOv4 and the NCA were comparable to each other for star-like lesions, masses with unclear borders, round- or oval-shaped masses with clear borders and partly visualized masses. On the contrary, the NCA was superior to the YOLOv4 in the case of asymmetric density and of changes invisible on the dense parenchyma background. Radiologists changed their earlier decisions in six cases per 100 for NCA. YOLOv4 outputs did not influence the radiologists’ decisions. Conclusions: in our set, NCA clinically significantly surpasses YOLOv4.

No abstract

No abstract

No abstract

Objective. Atypical breast cysts are often quite a serious problem in noninvasive ultrasound differential diagnosis. To develop a system for automated analysis of grayscale ultrasound images, which on the principles of mathematical processing would make it possible to increase the specificity of diagnosis in this situation.Material and methods. The authors developed the CystChecker 1.0 software package. To test this system, they used a set of 217 ultrasound images: 107 cystic (including 53 atypical lesions that were hardly differentially diagnosed by standard methods) and 110 solid (both benign and malignant) breast masses. All the masses were verified by cytology and/or histology. Visual assessment was carried out analyzing grayscale ultrasound, color/power Doppler, and elastography images.Results. Using the system developed by the authors could correctly identify all (n = 107 (100%)) typical cysts, 107 (97.3%) of 110 solid masses, and 50 (94.3%) of 53 atypical cysts. On the contrary, the standard visual assessment provided a possibility of correctly identifying all (n = 107 (100%)) typical cysts, 96 (87.3%) of 110 solid masses, and 32 (60.4%) of 53 atypical cysts (p < 0.05). The corresponding values of the overall specificity of automated and visual assessments were 98 and 87%, respectively.Conclusion. Using the system developed by the authors for automated analysis provides a higher specificity than the visual assessment of an ultrasound image, which is carried out by a qualified specialist.