Room: Exhibit Hall | Forum 1
Purpose: Mammography has been proven as the most effective screening method. However, several significant features of breast cancer are rarely detected on mammograms due to the slight differences among the densities of breast tissues. For enhancing the solution, this study proposed a novel Laplacian pyramid integrated with dynamic unsharp masking method for mammogram enhancement.
Methods: The proposed method utilizes dynamic weight factors to adaptively enhance details and suppress noise. Laplacian pyramid is utilized to preserve the fine structure at each scale. Dynamic unsharp masking is presented to adaptively enhance details and suppress noise at the moment. Our enhancement method consists of five steps: 1) Sample image down to obtain Gaussian pyramid images; 2) Calculate global mean, neighborhood mean, and neighborhood variance of each scale image to determine the weight dynamic unsharp masking. 3) Process each scale image of Gaussian pyramid with dynamic unsharp masking enhancement; 4) Subtract the enhancement results with Gaussian pyramid images to acquire Laplacian pyramid images; 5) Restructure the final enhanced image with LP images and Gaussian pyramid images according to the restructuration principle of Laplacian pyramid.
Results: To evaluate the proposed method for enhancing image contrast, quantitative studies including information entropy were performed on realistic mammography cases. Furthermore, peak signal-to-noise ratio was applied for assessing the effect of suppressing noise. Results show that both IE and PSNR are increased with the proposed method, compared with the current state-of the art methods.
Conclusion: Our method has demonstrated that different types of regions are enhanced by different degrees with the aid of a regional adaptive evolution, while higher IE and PSNR value indicates that the method is suitable for adaptively enhancing the details and suppressing the noise of digital mammography
Funding Support, Disclosures, and Conflict of Interest: National Natural Science Foundation of China (81301940 and 81428019), National Key Research and Development Program (2016YFA0202003), Guangdong Natural Science Foundation of China (2016A030310388 and 2017A030313692), and Southern Medical University Startup fund (LX2016N003).
Not Applicable / None Entered.