Specular reflections removal in colposcopic images based on neural networks: Supervised training with no ground truth previous knowledge
This work addresses a domain-specific issue for medical professionals by improving the observability of cervical anatomy in colposcopic images, potentially aiding in early cervical cancer detection, though it is incremental as it builds on existing neural network approaches for image restoration.
The paper tackles the problem of specular reflections in colposcopic images, which can obscure regions and lead to misdiagnosis, by introducing a neural network-based strategy that removes these reflections and restores hidden areas without ground truth data, achieving elimination in 21 out of 22 images and near-complete removal in the remaining one.
Cervical cancer is a malignant tumor that seriously threatens women's health, and is one of the most common that affects women worldwide. For its early detection, colposcopic images of the cervix are used for searching for possible injuries or abnormalities. An inherent characteristic of these images is the presence of specular reflections (brightness) that make it difficult to observe some regions, which might imply misdiagnosis. In this paper, a new strategy based on neural networks is introduced for eliminating specular reflections and estimating the unobserved anatomical cervix portion under the bright zones. For overcoming the fact that the ground truth corresponding to the specular reflection regions is always unknown, the new strategy proposes the supervised training of a neural network to learn how to restore any hidden regions of colposcopic images. Once the specular reflections are identified, they are removed from the image, and the previously trained network is used to fulfill these deleted areas. The quality of the processed images was evaluated quantitatively and qualitatively. In 21 of the 22 evaluated images, the detected specular reflections were eliminated, whereas, in the remaining one, these reflections were almost completely eliminated. The distribution of the colors and the content of the restored images are similar to those of the originals. The evaluation carried out by a specialist in Cervix Pathology concluded that, after eliminating the specular reflections, the anatomical and physiological elements of the cervix are observable in the restored images, which facilitates the medical diagnosis of cervical pathologies. Our method has the potential to improve the early detection of cervical cancer.