Image processing and image analyses applied to mastology

The new digital equipment for medical imaging (thermographic cameras, ultrasound, digital mammograms, etc.) allows the anatomical information from different sources is combined. These data (images) must be processed to enhance and extract features. In this area, techniques of image processing and pattern recognition are very important to automate certain procedures and to ease the interaction and the combination of this data.

This research intends to help in diagnostic by medical imaging, with the purpose of aiding in the early diagnosis and classification of breast diseases (benign/malignant). The project aims to integrate the main stages of image processing methods, in order to: the interpretation of the images; identification of tissue and anatomical and functional structures. The project extends from the acquisition of images and their organization in databases, through image analysis using numerical methods, until a final stage, where machine learning techniques must be used in order to be a feature extraction of breast images.




Infrared Images Databases

DMR - Database For Mastology Research

Camera used in project: FLIR SC-620,
Resolution 640 x 480: Pixel = 45 μm

Image Segmentation

Camera used in project: Flir Thermacam S45,
Resolution: 320x240

A Methodology for Breast Disease Computer-Aided Diagnosis using dynamic thermography

Camera used by the project: FLIR SC-620,
Resolution: 640 x 480 pixels