Detección del Bacilo Mycobacterium Tuberculosis mediante Reconocimiento de Patrones.

An image is an excellent source of information whenever be possible to identify and extract the objects on it. The techniques of patterns recognition make it possible to isolate groups of objects for specific applications. The group of our interest is the Mycobacterium Tuberculosis bacillus (or Koch bacillus), causal of the tuberculosis infection, and it is possible to see them in the photographic images from sputum samples of patients who present signs of this disease. This article presents an algorithm based on wavelets of morphologic and statistical pattern recognition to isolate and quantify on the image this bacillus to diagnostic the severity of the infection in the patient. Finally are presented some results, conclusions and recommendations of its application in Telemedicine.