Medical Image Colorization for Better Visualization and Segmentation
| Autor: | Yoshihiko Gotoh, Muhammad Usman Ghani Khan, Nudrat Nida |
|---|---|
| Rok vydání: | 2017 |
| Předmět: |
Modality (human–computer interaction)
genetic structures Pixel business.industry Computer science ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION 030206 dentistry Grayscale Peak signal-to-noise ratio 030218 nuclear medicine & medical imaging Visualization 03 medical and health sciences 0302 clinical medicine Segmentation Computer vision Artificial intelligence Monochromatic color Noise (video) business |
| Zdroj: | Communications in Computer and Information Science ISBN: 9783319609638 MIUA |
| DOI: | 10.1007/978-3-319-60964-5_50 |
| Popis: | Medical images contain precious anatomical information for clinical procedures. Improved understanding of medical modality may contribute significantly in arena of medical image analysis. This paper investigates enhancement of monochromatic medical modality into colorized images. Improving the contrast of anatomical structures facilitates precise segmentation. The proposed framework starts with pre-processing to remove noise and improve edge information. Then colour information is embedded to each pixel of a subject image. A resulting image has a potential to portray better anatomical information than a conventional monochromatic image. To evaluate the performance of colorized medical modality, the structural similarity index and the peak signal to noise ratio are computed. Supremacy of proposed colorization is validated by segmentation experiments and compared with greyscale monochromatic images. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|