The Classification And Recognition Of Chest Hydrocele Desquamate Cell With Digital Image Processing And Pattern Recognition Technique

Nowadays, cancer has been considered as the fatal illness in the world. Whether a patient who has cancer could be cured is mostly due to early diagnosis. A lot of data indicates that when a patient is diagnosed as cancer, it would be terminal cancer. That means the patient had lost the chance to be cured. For instance, we need to develop a kind of new technique to realize early diagnosis. In present, the main cancer diagnosis means include: X transillumination, CT, MRT, isotope, BF, PALB, pathology diagnosis and so on. But for the clinical diagnosis, the most credible method is pathology diagnosis, in which the cancer cell can be identified by the difference between the shape of the certain cancer cell and normal cell. Unfortunately, because of many experts check the cell only by their experience and medical rules. Inevitably, that will results in many errors. So many scientists try to developed new means of cancer cell identification, which mainly use computer image processing technique and neural net work. The purpose is to reduce the doctor's heave task and increase the precise of cancer diagnosis.In this paper, the purpose of the experience is to detect the chest hydrocele desquamate cancer cell and recognize them. According to the complicated features of the cancer cell, canny arithmetic operator and morphologic configuration element are applied to segment the cell and get the edge successfully. Then in order to recognize and classify the cancer cell, BP neural net work is used in this experience. Under a lot of training, the BP neural net work approaches the goal, so it can be used to the early diagnosis of chest hydrocele desquamate cancer.