Improved Approach for Intima-Media Thickness Measurement in Carotid Ultrasound Imaging Using Adaptive Despeckle Filtering

This paper addresses the challenge of Intima-Media Thickness (IMT) measurement, a critical technique for assessing cardiovascular risk, by focusing on the impact of speckle noise in ultrasound (US) imaging. The presence of speckle noise degrades the accuracy of IMT measurements, making effective despeckling essential. However, there is a trade-off between noise reduction, surface smoothing, and structural preservation. This study proposes the efficiency of despeckling filters in improving IMT measurement accuracy. The effectiveness of speckle suppression is quantified using the Normalized Mean Square Error (NMSE), while visual quality is assessed through Peak Signal-to-Noise Ratio (PSNR). Structural preservation post-filtering is evaluated using the Mean Structural Similarity Index (MSSIM). Furthermore, IMT detection accuracy is measured using the Mean Absolute Distance (MAD), representing IMT error. The study presents the Adaptive Regularization Savitzky-Golay (ARSG) filter as the best-performing method. It shows that the ARSG filter reduces IMT measurement error (with a MAD of 9.54) and achieves a high level of structural preservation (with an MSSIM of 89.39%). Compared to other methods, the ARSG filter provides a better balance between noise reduction and structural integrity, leading to improved IMT measurement accuracy. In summary, the paper demonstrates that the ARSG filter is highly IMT measurements by balancing noise suppression and the preservation of image structure, making it a valuable tool for assessing cardiovascular risk through ultrasound imaging.