| abstract |
An endemic skeletal fluorosis grading diagnosis system based on a multi-type image fusion neural network, relates to the technical field of image processing, and aims at the problem of low diagnostic efficiency for skeletal fluorosis in the prior art, including: a preprocessing module, a rough image segmentation module in the lesion area , a multi-type image fusion module and a disease classification diagnosis module, the multi-classification model based on the fusion of the coarse segmentation feature map and the original image makes full use of the lesion area information, and strengthens the neural network's ability to detect sensitive areas on the basis of ensuring the integrity of the information. cognitive ability. The cost function designed by the invention emphasizes the position with high lesion probability in the feature map and weakens the influence of irrelevant background, solves the problem that the lesion area accounts for a small proportion of the total image area, and improves the training and classification efficiency of the model. The invention provides an auxiliary means for the detection of fluorosis, fills the blank of intelligent diagnosis of fluorosis, and improves the diagnostic efficiency for fluorosis. |