Researchers have developed a DL model to classify Tuberculosis (TB).
Localization tasks like object detection and semantic segmentation require strong supervision of bounding boxes pixel-wise labels. Large-scale labeling by an expert radiologist involves lots of effort and time. A drawback of these algorithms is that they are highly data-hungry and require large-scale labeled datasets. When trained on large datasets, these algorithms give accurate results and have shown scalable performance with increased training data. DL has dominated medical imaging for many tasks such as classification, detection, segmentation, etc. It is finding its place in radiology centers as a supporting and assisting tool for radiologists. Recently Deep Learning (DL) algorithms have led to rapid advances in computer vision and medical imaging. Our model trained with the configurations mentioned in this paper outperformed benchmark results. A radiologist subjectively validated the results. The model was developed and tested using the cloud computing platform Google Colaboratory (NVidia Tesla P100 GPU, 16 GB of RAM). The method was evaluated on CXR from the clinical center in the ratio of 70:20 for training and testing. During PCM training, the network is a weight-shared siamese architecture where the first branch applies the affine transform to the image before feeding to the network, while the second applies the same transform to the network output. The model consists of a modified Resnet50 backbone for extracting feature corpus from the images, a classifier, and a pixel correlation module (PCM). This paper proposes a deep learning model for the classification and localization of chest diseases by using image-level annotations. Therefore, there is a drift toward developing computer-aided assistive systems to help radiologists. However, close to fourteen commonly occurring diseases are sometimes difficult to identify by visually inspecting the images.
Therefore, a 2D view of the 3D anatomy is sometimes sufficient for the initial diagnosis. Chest X-Ray (CXR) images provide most anatomical details and the abnormalities on a 2D plane.
0 kommentar(er)
