Research On Medical Image Enhancement Technique For Magnetic Resonance Imaging

As the rapid development of medical imaging technology,medical imaging has become an indispensable technology in modern medical diagnosis by virtue of its noninvasiveness and high efficiency.The medical images are able to visualize the internal tissue structure of the human body in a non-invasive manner,facilitating diagnosis and treatment by doctors.Magnetic resonance imaging is the most common imaging modality.In processes of MR image acquisition,some images suffer from low contrast,high dynamic range,and blurred edge details due to interference from various factors,such as light,temperature,and the noise inevitably introduced during the imaging process.The low-quality images not only have an impact on the clinician’s diagnosis,but also negatively interfere with the analysis of medical data based on artificial intelligence technology.For this reason,how to improve the quality of medical images is an urgent technical problem.1)To solve the problem that MR image details are complex and difficult to enhance,an enhancement algorithm based on non-downsampled contour wave transform is proposed.Using Non-Subsampled Contourlet Transform for MR image decomposition.For high-pass images,enhanced edge contour details are positioned from adjacent eight directions using four blurring rules.For low-pass images,contrast enhancement using improved adaptive global and local histogram enhancement algorithms.The Bra TS 2017 dataset was selected as the experimental dataset for validation,and the enhanced results were compared with three image enhancement methods,Lai,DFWT,and FAWDN.The analysis of the results was performed using the rubric of subjective evaluation and quantitative analysis.The results show that the proposed enhancement algorithm can improve the image contrast while enhancing the detail information of the image,which is beneficial to the doctors for accurate diagnosis and subsequent analysis of the disease.2)To solve the problems of blurred details and low contrast in the acquisition of medical images,a domain transform medical image enhancement algorithm based on multi-scale decomposition is proposed.Decompose images into high-frequency images and low-frequency images using multi-scale decomposition.Combining fractional order integration with the Scharr operator for improving the contrast of high frequency components in the contour domain and mining weak edges in images.Using an improved adaptive histogram enhancement algorithm to improve the overall contrast of the low frequency component.Image reconstruction based on inverse nondownsampled contour waves is performed to obtain enhanced images.Both Bra TS2018 and INbreast were selected as experimental datasets to evaluate both image assessment metrics and help for segmentation experiments.The comparison experiments show that the proposed algorithm achieves a good contrast enhancement effect,has good stability and adaptability,and can well meet the needs of medical image enhancement.