Computerized Medical Image Analysis for Optimization of Cancer

Abstract

The medical imaging has a significant role in the diagnosis and management of cancer, but the images produced by medical imaging modalities like MRI and CT usually contain noise, low contrast, and visual artifacts. These artifacts may conceal valuable tumor data and complicate the correct interpretation of clinicians. This thesis is aimed at enhancing medical image quality by means of computerized analysis of images so as to aid in enhancing the evaluation of cancers. Images of different patients acquired by MRI and CT were utilized in standard DICOM format in this study. Ten image enhancement procedures have been used in this study including sharpening, noise reduction, edge retention, contrast enhancement and visualization in three dimensions. The tools that were utilized to complete the analysis were RadiAnt DICOM Viewer and 3D Slicer, which provided the visualization options of the images, multiplanar reconstruction, tumor segmentation, and quantitative analysis. The results shows that image clarity and contrast and noise reduction are greatly enhanced by the use of appropriate enhancement methods. These enhancements contribute to the better visualization of tumor borders and the adjacent organs, resulting in more accurate segmentation and analysis. The improved quality of images also means that repeat scans can be reduced hence the cost and patient exposure can be minimized. On the whole, this study indicates that computerized medical image analysis may be used as a quality tool in streamlining the cancer imaging process and assisting in better diagnosis and treatment planning.