Deep convolutional neural networks for automatic detection of uterine fibroids in ultrasound images

Uterine fibroids are a frequent harmless tumour that affects women who are fertile. If uterine fibroids (UF) are detected and diagnosed early, treatment can be successful. In this study, we assessed our novel dual-path deep CNN framework's efficacy in UF identification against the cutting-edge DL architectures VGG16, ResNet50, and InceptionV3. Several metrics are utilised to assess the model performance once the photos are used for training and validating the models based on deep learning. Our proposed DPCNN model attained 99.8% accuracy which is maximum as compared to current DL models. Our results demonstrate the efficacy of using DL-based techniques to enable automatic UF identification from medical photos. The best results were obtained with our novel DPCNN architecture, however optimised versions of models that had been trained such as ResNet50 and InceptionV3 also produced impressive outcomes. This study establishes the groundwork for subsequent research and may improve the accuracy and applicability of UF detection.