Enhancing US Thyroid Images for Cancer Detection: Pre-Processing Techniques

Thyroid cancer detection using ultrasound (US) images requires efficient preprocessing techniques to enhance image quality and suppress noise. This research introduces a preprocessing technique for US thyroid images in MATLAB, aiming to apply and measure several filtering methods. The filters are chosen respectively as average, adaptive median, mean, and Wiener to handle different types of noise and preserve necessary details for analysis. Applying these filters one by one enables us to identify and highlight clear information while maintaining the accuracy of medical testing. The effectiveness of filtering techniques for noise reduction and image enhancement is shown in experimental results, which prepares the way for accurate thyroid cancer detection. According to the findings, preprocessing plays a key role in enhancing the diagnostic ability of US thyroid machines.