Image Enhancement for Visualization Jaw Bone in DICOM CT Scans using Windowing Strategies, Adaptive Denoising Bilateral Filtering, and Sharpening

Digital Imaging and Communications in Medicine (DICOM) has become the standard for medical imaging in the healthcare field. The open-source nature of this data model facilitates its application in manufacturing, healthcare providers, and research scientists. This study focuses on DICOM data for scanning the human jaw region. Image processing research requires datasets, such as ground truth. however, there is currently a lack of ground truth data for jaw structures in DICOM files. Therefore, manual annotation for creating ground truth is necessary. To support this endeavor, image enhancement aimed at improving the visualization of the jawbone. This paper proposes an image enhancement technique for visualizing jawbone structures in DICOM-CT scans using windowing strategies, adaptive denoising bilateral filtering, and sharpening. Windowing strategies are applied to enhance the bone area based on the values of window width and window level. The quality is then further refined using bilateral filtering and sharpening techniques. Our experiments, conducted on a dataset of DICOM files, confirmed the effectiveness of the proposed technique through quantitative metrics such as the Perceptual Image Sharpness Index (PSI) and Contrast Improvement Index (CII), alongside visual comparisons using histograms. The results indicate that our method not only enhances the visual quality of CT images but also facilitates the subsequent annotation process by streamlining the creation of ground truth (GT) data.