A new approach for medical image enhancement based on luminance-level modulation and gradient modulation. (February 2019)

• Record Type: Journal Article
• Authors: Zhao, Chenyi; Wang, Zeqi; Li, Huanyu; Wu, Xiaoyang; Qiao, Shuang; Sun, Jianing
• Is Part Of: Biomedical signal processing and control. Volume 48(2019)
• Page Start: 189
• Page End: 196
• Publication Date: 2019-02
• Subjects: Luminance-level modulation -- Gradient modulation -- Medical images -- Image enhancement; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2018.10.008
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: A novel medical images enhancement method based on luminance modulation and gradient modulation (LM&GM) is proposed. The LM is adopted for adjusting luminance and increasing contrast by shrinking the global dynamic range of input image. The GM is used for enhancing details and textures of the images on the basis of the luminance modulation. The performance of the proposed method is compared with the popular enhancement methods. Quantitative analyses are applied to demonstrate the significance of LM&GM superior to other methods. Abstract: Medical images play a significant role in modern diagnosis but often suffer from non-uniformity and low-luminance, which always affects the results of diagnosis and even leads to misdiagnosis in real applications. To obtain a clear and accurate view of the medical images, a new method based on luminance-level modulation and gradient modulation (LM&GM) is proposed in this paper. LM&GM is a two-stage approach that, first, increases the visual perception using the luminance-level modulation (LM) operation by compressing the range of luminance levels of the input image, and second, uses the gradient modulation (GM) operation to enhance the details of the previous step result. Experimental results on CT images, X-ray images and MRI images from medical image datasets and quantitative analyses by structural similarity index measurement (SSIM), average gradient (AG), relative enhancement in contrast (REC) and information entropy (IE) demonstrate that the results of the proposed method are competitive and overwhelm those of the existing methods.