Deep learning skin disease diagnosis system based on optical coherence tomography

Abstract

The invention provides a deep learning skin disease diagnosis system based on optical coherence tomography, which is characterized in that an image acquisition unit adopts an OCT imaging system to carry out three-dimensional OCT image imaging on a detected skin area to realize non-invasive drawing of skin deep layer information data; secondly, an image preprocessing unit carries out noise reduction and enhancement preprocessing on an original three-dimensional OCT image, the inherent speckle noise problem of OCT imaging and the limitation problem of low sampling rate are eliminated, and the resolution of the image is improved; then, a feature extraction unit extracts features of skin anatomy and skin microvessels based on the preprocessed OCT image, and deep feature extraction and feature fusion are performed through a deep learning diagnosis model of a deep learning diagnosis unit, so that corresponding skin disease information is detected; according to the diagnosis system, the accuracy of skin disease diagnosis is improved, the dependence on diagnosis experience and level of doctors is reduced, and potential skin diseases can be found, diagnosed and treated as soon as possible under the condition that pathological biopsy is not needed.

Description

technical field [0001] The invention relates to a deep learning skin disease diagnosis system based on optical coherence tomography. Background technique [0002] Skin is the largest organ of the human body, and skin disease is a general term for diseases that occur in the skin and skin appendages. At present, the diagnosis of skin diseases still largely depends on the experience of clinicians. In particular, there are 700 million occupational groups in China, and about 200 million of them are frequently exposed to dust, chemicals, heat radiation, ultraviolet rays and other environments that are harmful to the skin. It is estimated that 20-30% of occupational diseases are skin diseases, which is one of the most common occupational diseases. In particular, the number of cases of skin cancer has risen dramatically in the past few decades. Skin cancer is the most common cancer in the world, mainly including basal cell carcinoma, squamous cell carcinoma and melanoma. Early s...

AI Technical Summary

Problems solved by technology

Most of the existing deep learning-based dermatology diagnosis only use dermoscopic image data as diagnostic input, but because dermoscopic images can only express skin surface information, but cannot reflect skin deep structure information, the intelligence of a single dermoscopic image Dermatology diagnostic systems tend to have low sensitivity and specificity

Images