Answers are grounded in this paper and cite their sources.
Evaluates image quality of chest X-rays, optimizing Gaussian filter for clinical use.
This study aimed to optimize the standard deviation (σ) of the Gaussian low-pass filter used in the Unsharp Masking (USM) process for clinical digital chest radiography. The study was intentionally limited to normal male and female adults in their twenties to minimize physiological and anatomical variability and to establish a standardized baseline for clinical image preprocessing. A total of 241 normal chest X-ray images (122 males and 119 females) obtained from routine health checkups were analyzed. Image quality was quantitatively evaluated using the Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) across a σ range of 0.1–2.0. Both metrics indicated that USM produced superior image fidelity compared with simple Gaussian filtering. The PSNR reached the elbow point of approximately 60 dB at σ=0.5, while SSIM remained consistently high (>0.99). No statistically significant difference was observed between male and female groups, although females showed slightly higher numerical values. The results confirm that σ=0.5 provides an optimal balance between noise suppression and structural preservation. This value, derived from a homogeneous normal population, may serve as a reference parameter for image preprocessing in future studies involving diverse or pathological groups. The results are expected to contribute to improving image consistency and supporting clinical standardization in digital chest radiography.
No takes yet. Share an insight, caveat, or question.
Youl-Hun Seoung (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: