Enhancing placental pathology detection with GAMatrix-YOLOv8 model.

Purpose: Given the advancements in machine vision, deep learning, convolutional neural networks, and other technologies, various artificial intelligence techniques have been used in placental pathology examinations. However, there are limitations in achieving real-time recognition and precise localization. This study aims to enhance and validate the YOLOv8 model while investigating its significance in the pathological detection of placental tissues. Methods: Embedding GAM attention into the YOLOv8 backbone network and incorporating critical preprocessing steps like image enhancement and normalization enable the model's middle layer to focus more effectively on essential image features, improving computational efficiency and accuracy. Results: This study developed a new model,GAMatrix-YOLOv8,which enhances object detection accuracy and efficiency while offering superior generalization capabilities. Both the training and validation accuracy are close to 100 %. Compared with GoogleNet, ResNet18,YOLOv8 and other models,GAMatrix-YOLOv8 had an Accuracy of 0.997,Precision of 0.975,Recall of 0.970,and F1-Score of 0.972,which were significantly higher than those of other models. Based on the Gamatrix-YOLOV8 model, we developed an intuitive and easy-to-use graphical user interface, which enables users to easily upload pictures, view the prediction results of the model in real time, analyze and verify the detection results. Conclusions: GAMatrix-YOLOv8 achieves unprecedented prediction accuracy in detecting delayed villous maturation in placental pathological tissues through algorithmic innovation. The graphical user interface allows users to upload images, view model prediction results in real time, and analyze and verify test outcomes. This provides a theoretical and technical foundation for further research into artificial intelligence-based pathological auxiliary diagnostic systems.