The invention discloses a cancer lesion range prediction supporting system based on deep learning and method. The system comprises a client and a server, wherein the client is used for monitoring an image acquired by the current endoscope device and transmitting the image to the server, and an analysis result fed back by the server is received and displayed; and the server is used for instantly judging the part corresponding to the image and the part features according to the image acquired by the client, and the analysis result is fed back to the client. When the endoscope device performs image acquisition and the client is triggered, the collected endoscope image is acquired and transmitted to the server; the server receives the endoscope image as a parameter, a convolutional neural network model is called for analysis on whether the endoscope image is qualified and analysis on part judgment and part feature recognition; and according to the acquired analysis result, the client callsan image representing each part and a sign representing the part feature for overlapped display. An early cancer can be diagnosed, a cancerization range is circled, the operation is simple and easy to use, and social and economic value is significant.