“…Author ( Rohmah & Bustamam, 2020 ) combined self-supervised learning and supervised segmentation to accomplish hybrid led learning for breast ultrasound images using an enhanced U-Net algorithm. To increase image segmentation accuracy, the author used the Resnet module with convolutional attention in the encoding part of the U-Net network to organize the output features and thus improve blood vessel segmentation accuracy ( Pratiwi et al, 2021 ). Accuracy; developed a fusion neural network for ultrasound foetal head edge identification using the U-Net network and the last layer of U-Net++ features.…”