Intelligent System for the Automated Diagnosis of Histological Images

The problem of diagnosing histological/cytological images is the main objective of this work. A system for the analysis of this kind of digital images is proposed. This system aims to systematize the image processing/machine vision/classification procedures needed to reproduce the physician interpretive abilities and collect them in standardized protocols both reproducible and of quantifiable performance. Flexible soft– computing techniques will be needed to reach an adequate performance at the level of image interpretation, while a knowledge-centric, object-oriented implementation approach will be preferable to hide the details of these procedures from the specialist. A case study of pathology diagnosis is proposed as illustration of the difficulties and the techniques that will be used in the system. Key-Words: Image Interpretation, Histological Images, Medical Protocols, Texture Analysis, Structural Image Analysis, Follicular Lymphoma