A non-uniform quantization scheme for visualization of CT images

Area of Science:

• Medical Imaging
• Image Processing
• Radiology

Background:

• Medical image quality is crucial for diagnosis and treatment planning.
• Image noise degrades visual quality, potentially leading to misinterpretations.
• Medical image enhancement is vital for improving diagnostic accuracy.

Purpose of the Study:

• To present a novel clustering-based contrast enhancement technique for computed tomography (CT) images.
• To improve the visual quality and diagnostic utility of CT scans.
• To develop an efficient method for converting high-bit-depth CT images for standard displays.

Main Methods:

• A recursive data splitting into clusters is employed to maximize error reduction.
• The technique groups similar pixels, maximizing inter-cluster and minimizing intra-cluster similarities.
• 256 clusters are used to convert 16-bit CT scans to 8-bit images for visualization.

Main Results:

• The proposed method effectively enhances contrast in CT images.
• It demonstrates superior execution efficiency compared to existing algorithms.
• Enhanced images show improved quality suitable for standard displays.

Conclusions:

• The clustering-based contrast enhancement technique offers a valuable tool for medical image processing.
• The method provides an efficient and effective solution for CT image enhancement.
• This approach aids in improving diagnostic accuracy by enhancing image quality.