CT brain image advancement for ICH diagnosis

Area of Science:

• Medical Imaging
• Image Processing
• Radiology

Background:

• Accurate detection of primary intracerebral haemorrhage (ICH) is crucial for patient diagnosis.
• Computed tomography (CT) brain image quality directly impacts diagnostic accuracy.
• Current imaging methods may require enhancement for optimal visualization of ICH.

Purpose of the Study:

• To develop and evaluate an image enhancement algorithm for CT brain scans.
• To improve the clarity of edges and visualization of ICH in medical images.
• To assess the reliability and performance of the proposed enhancement technique.

Main Methods:

• CT brain images (n=40) with primary ICH were processed.
• Images were denoised using Wiener filter, non-local means, and wavelet methods.
• A modified unsharp masking (UM) algorithm was applied for image enhancement.

Main Results:

• The combination of Wiener filter and modified UM algorithm yielded optimal results.
• Performance metrics included low mean square error (2.89) and high peak signal-to-noise ratio (31.72).
• Blinded assessors validated the algorithm's reliability with a median score of 65%.

Conclusions:

• The proposed algorithm effectively enhances CT brain images for ICH detection.
• This technique offers a reliable approach to improve medical image analysis.
• Enhanced image clarity can potentially lead to more accurate and timely diagnoses of ICH.