Integration of Median Filter and K-means Clustering Technique for Ultrasound Tomography

Ultrasound tomography is a powerful imaging modality widely employed in medical diagnostics due to its non-invasiveness and small object detection capabilities. However, the reconstructed images often suffer from the presence of noise, compromising the accuracy and interpretability of diagnostic information. In this paper, we propose and evaluate the application of a median filter to address noise artifacts in the reconstructed images obtained through the distorted Born iterative method (DBIM). Quantitative assessment of the proposed approach involves the utilization of evaluation metric, normalized error, to measure its effectiveness. Furthermore, the application of a basic machine learning technique, K-means clustering, assists in automatically dividing the reconstructed image into distinct segments representing objects, background, and noise. The outcomes demonstrate that K-means clustering effectively delineates these regions in images generated through the proposed method, with minimal presence of noise in the recovered images.