Automatic Tongue Crack Extraction For Real-Time Diagnosis

A survey of artificial intelligence in tongue image for disease diagnosis and syndrome differentiation

“…Li et al 56 applied this model to the tongue crack segmentation, and they improved its encoder to extract relatively more abstract high-level semantic features. Similarly, Peng et al 57 also improved the U-Net framework and designed a lightweight model P-Net with the letter “P” structure to be suitable for remote tongue image segmentation. Zhou et al 18 adding a morphological layer to U-Net aim at refining the obvious morphological errors in U-Net segmentation.…”

“…To change the model to extract specific concerns, Li et al 56 introduced a global convolution network module to extract relatively abstract high-level semantic features, while Huang et al 54 constructed a receptive field block based on the receptive field theory that the region closer to the center of retinotopic maps is more important than others in distinguishing objects, making the model deal more with the blurred edge of the tongue body. Similarly, Peng et al 57 applied an attention module to intensify the attention to the boundary and suppress useless information.…”

“…Tang et al 11 reported a solution of tooth-marked recognition based on the segmented tongue, and Li et al 56 refined the U-Net model to especially segment the cracks on the tongue, the same as the reference. 57 The feature of sublingual vein had also been noticed and extracted, 10 , 68 Qiu et al’s 10 study added the cross-channel attention module to assign more weight to the target area, which shows an improved performance in their lightweight tongue diagnosis system on mobile devices, and another paper 68 proposed a two-stage segmentation method: a fully CNN network without downsampling to reduce the loss of spatial feature information effectively and another fully CNN network with proper dilated convolution to avoid the gridding issue. For real-time diagnosis, the size of the popular models is generally too large; Li et al 64 proposed a lightweight architecture for tongue image segmentation in real-time, and they further proposed the P-Net model 57 refined from U-Net and used for real-time tongue crack extraction.…”

“… 57 The feature of sublingual vein had also been noticed and extracted, 10 , 68 Qiu et al’s 10 study added the cross-channel attention module to assign more weight to the target area, which shows an improved performance in their lightweight tongue diagnosis system on mobile devices, and another paper 68 proposed a two-stage segmentation method: a fully CNN network without downsampling to reduce the loss of spatial feature information effectively and another fully CNN network with proper dilated convolution to avoid the gridding issue. For real-time diagnosis, the size of the popular models is generally too large; Li et al 64 proposed a lightweight architecture for tongue image segmentation in real-time, and they further proposed the P-Net model 57 refined from U-Net and used for real-time tongue crack extraction. The data set to supervise the tongue segmentation model is relatively small, which may cause generalization issues; Li et al 19 applied an iterative learning method to train the network on the good samples repeatedly judging by some specific filters.…”

Tongue crack recognition using segmentation based deep learning

“…It was a weakly supervised method that added several classification branches to recognize the tooth-marked tongue and cracked tongue according to the YOLO object detection model. Peng et al 12 proposed a P-type neural network architecture based on a lightweight encoder-decoder structure which could get the detailed extraction result at pixel level. Xue et al 13 proposed to use cracked and non-cracked regions to train Alexnet to extract deep features of cracked regions.…”