Optimized Deep Learning for Medical Image Retrieval and Classification: A Framework

Computer Aided Diagnosis (CAD) has been a key research area for a decade. CAD systems can help to detect and classify diseases by automating/semi-automating medical image analysis. Computer Aided Diagnosis deals with different medical image modalities and comprises two main functionalities. One is image retrieval, i.e. retrieving similar kinds of images and the second is classification for a given input image. Medical images come in various modalities (e.g., X-ray, MRI, CT scans) and formats, leading to challenges in developing a unified approach for feature extraction, retrieval, and classification across diverse image types which are essential for developing CAD systems. Traditional retrieval methods such as text annotation and histogram-based approaches suffer from several critical drawbacks. This work focuses on developing Content-Based Medical Image Retrieval (CBMIR) and Classification models which can be used in developing efficient CAD systems and in turn can provide complete assistance to radiologists and doctors in diagnosing diseases effectively.

Pagina's: 200, Paperback, LAP LAMBERT Academic Publishing