The invention provides a method and a
system for automatically identifying injury parts based on YOLO. The method comprises the following steps: receiving and analyzing a medical image
data set, carrying out comprehensive background support
processing, and generating a multi-dimensional data environment; on the basis of the multi-dimensional data environment, by applying a YOLOv5 target detection
algorithm, through an efficient
convolutional neural network structure, marking a
possible injury condition region, and by adopting an
image enhancement technology, performing enhancement
processing on a detected suspected injury condition region, and generating a visual quality improvement medical image
data set; based on the visual quality improvement medical image
data set, comparing the visual quality improvement medical image data set with a pre-constructed
medical knowledge graph by adopting a graph matching
algorithm, and performing semantic level analysis by adopting a semantic
analysis method to generate a fine
annotation medical image data set; and based on the finely labeled medical image data set, integrating into a safe and efficient
data management system, and generating an automatic injury part identification
data platform. According to the technical scheme provided by the invention, the accuracy and efficiency of injury part recognition are comprehensively improved.