Abstract
Medical imaging, spanning radiography, computed tomography, magnetic resonance imaging, ultrasonography, and histopathological microscopy, generates an enormous and continuously growing volume of visual data that must be interpreted, typically under significant time pressure, by a comparatively limited pool of trained radiologists and pathologists. Machine learning, and specifically deep convolutional neural networks trained on large annotated image datasets, has emerged over the past decade as a technology capable of automating specific, well-defined components of this interpretive workload, with demonstrated applications spanning lesion detection, anatomical segmentation, disease classification, and image quality enhancement [1].
References
This work is licensed under a Creative Commons Attribution 4.0 International License.