Noise Reduction in Lumbar Sagittal Short Tau Inversion Re-covery (STIR) MRI Images Using a Combination of Wiener Filtering With Contrast Stretching and Unsharp Masking

Background: Magnetic Resonance Imaging (MRI) is one of the most important imaging modalities in medicine, especially for diagnosing spinal disorders. The Short Tau Inversion Recovery (STIR) sequence has the advantage of suppressing fat signals, but often produces images with a low Signal-to-Noise Ratio (SNR), so that noise can interfere with diagnostic quality. Therefore, image optimization methods through noise reduction and image quality enhancement techniques are needed. Objective: This study aims to analyze the effectiveness of combining the Wiener Filter with Contrast Stretching and Unsharp Mask in reducing noise and improving the quality of Lumbar Sagittal STIR MRI images. Method: This study used an experimental quantitative approach with lumbar sagittal STIR MRI images as the research object. The research process included applying the Wiener Filter as a noise removal method, then improving the quality with Contrast Stretching and Unsharp Mask. Image quality was evaluated quantitatively using the SNR, CNR, PSNR, and MSE parameters. Results: The application of a combination of Wiener Filter with Contrast Stretching and Unsharp Mask showed an increase in SNR and CNR values compared to before processing, as well as higher PSNR values with lower MSE. This indicates that this combination method is effective in reducing noise while sharpening the anatomical structure details in Lumbar Sagittal STIR MRI images. Conclusion: The combination of Wiener Filter with Contrast Stretching and Unsharp Mask has been proven to improve the quality of Lumbar Sagittal STIR MRI images with significant noise reduction, contrast enhancement, and image sharpness. This method has the potential to be implemented in radiology practice to improve diagnostic accuracy.