Breast Microcalcification Visualization Using Pseudo-Color Image Processing
Tanaporn Pipatnoraseth, Sukanya Phognsuphap, Cholatip Wiratkapun, Rawesak Tanawongsuwan, Petch Sajjacholapunt, Ikuko Shimizu · 2019
In this study, we aimed to find an effective method for improving visibility of microcalcifications on breast mammograms by using pseudo-color image processing techniques. We had considered twelve pseudo-coloring algorithms or color scales including autumn, bone, cool, copper, hot, rainbow, pink, spring, summer, winter, springtime-1, and springtime-2 color scales. Experiments were performed on 20 mammograms including 10 benign images, and 10 malignant images. The resultant images of using pseudo-color image processing with the twelve color scales were evaluated by radiologists. They determined that at what degree the pseudo-color images are helpful for improving visibility of micro-calcium spots on the mammograms by using 5-level score (5=very high, 4=high, 3=neutral, 2=low, 1=very low). The results using Ian McDowell's rating scale showed that the performances of bone and autumn color scales are very good. The performances of copper, winter, summer, springtime-1, pink, and hot color scales are good, and the performances of springtime-2, spring, cool, and rainbow color scales are fair. The bone color scale, which varied from black, through gray-blue, to white, outperformed other color scales, followed by autumn color scale which the important benefit is that the micro-calcium spots were illustrated shinier and clearer.