Research on Improved U-Net-based Ultrasound Prostate Segmentation Algo-rithm
Prostate ultrasound image segmentation plays a crucial role in clinical diagnosis and pathological research.However,it is often challenging to accurately locate the prostate due to complex scenarios such as protrusions and irregularities in the image edges.To address this issue,this paper proposes an improved network model based on the U-Net architecture.By incorpora-ting the Efficient Layer Aggregation Network(ELAN)module,the encoder is able to capture richer feature information,thus enhancing the model's perception and discrimination abilities.Experimental results demonstrate that the proposed algorithm achieves significant improvements in terms of mean absolute error(MPA),mean intersection over union(MIoU),and Dice coeffi-cient compared to the baseline model,with increases of 3.28％,2.97％,and 3.04％respectively.
NETL
NSTL
万方数据
