Feature-based ML and Radiomics in radiology image Analysis and computer-aided diagnosis systems.
Computed tomography colonography (CTC) is a medical imaging and diagnostic procedure for finding polyps of different shapes and sizes in the large intestine using image processing techniques. Our research focus on developing an automated computer-aided assessment of polyps using CTC images. The study's objectives were a) colon segmentation, b) electronic cleansing, and c) smaller polyp measurement and radiomics feature analysis. New image processing methods were developed by considering the domain aspects of colon analysis. The significance of the results was proved qualitatively and quantitatively. By considering the requirements of the radiologist from medical imaging applications, the research results were translated to a prototype software, which has features such as loading of CTC images, performing various image processing methods, and automated methods for colon segmentation, electronic cleansing, and smaller polyp measurement. The prototype includes object-oriented design, multithreading, following standard coding guidelines, and integrating volume-rendering frameworks for 3D visualization. This can be used as a basic image processing framework for different medical imaging modalities, such as CT, MRI, and PET. I am extending the work by applying the AI techniques for volume segmentation, shape analysis, and volume rendering.
 SCIE. Semantic segmentation and PSO based method for segmenting liver and lesion from CT images. 10.24425-ijet.2022.141283
 SCIE. Computer Aided Assessment of Colon Polyps in CT Colonography using Image Processing Techniques, Medical Physics
 SCIE. A quantitative validation of segmented colon in virtual colonoscopy using image moments. 10.1016/j.bj.2019.07.006
 SCIE. Computer-aided diagnosis of liver lesions using CT images: A systematic review. 10.1016/j.compbiomed.2020.104035
K N, Manjunath, "Feature-based ML and Radiomics in radiology image Analysis and computer-aided diagnosis systems." (2022). Technical Collection. 52.