Comparative Analysis of Quality Assessment of Fetal Heart Image Enhanced by Morphological Opening and Median Filtering

The quality of fetal heart ultrasound images plays a crucial role in prenatal diagnostics, where accurate image interpretation is vital for detecting congenital heart defects and monitoring fetal development. However, these images are often compromised by noise and artifacts, necessitating effective enhancement techniques. This study presents a comparative analysis of three image enhancement methods: Morphological Opening, Median Filtering, and a combination of both, evaluated using Mean Squared Error (MSE), Peak Signal-to-N oise Ratio (PSNR), and Structural Similarity Index (SSIM). The findings reveal that Median Filtering outperforms the other methods, achieving the lowest average MSE of 10.078442, indicating minimal distortion and higher accuracy in the enhanced images. Furthermore, Median Filtering produces the highest average PSNR of 38.150508, reflecting superior image quality with a higher signal-to-noise ratio. In terms of structural similarity, the Median Filtering method also excels with the highest SSIM value of 0.938693, demonstrating its effectiveness in preserving the structural integrity of the original images. In contrast, Morphological Opening and the combined method showed inferior performance, with higher MSE values of 12.837397 and 14.048525, lower PSNR values of 37.098163 and 36.712228, and lower SSIM scores of 0.918484 and 0.911021, respectively. These results highlight the superior capability of Median Filtering in enhancing fetal heart ultrasound images, making it a more suitable choice for clinical applications where high image quality is essential for accurate diagnosis.