Source: Deep Learning on Medium
Introduction to Biomedical Imaging using Deep Learning:
Learning In Chunks-
This blog post serves as a quick introduction to biomedical imaging using Deep Learning, where we will discuss how artificial intelligence (AI) will shape our future and will bring tremendous impact and applications in health industry. Deep learning has the potential to revolutionize disease diagnosis and management by performing classification and many other difficult tasks for human experts and by rapidly reviewing immense amounts of images.
Biomedical images are measurements of the human body on different scales.
(i.e. microscopic, macroscopic, etc.). They come in a wide variety of imaging modalities (e.g. a CT scanner, an ultrasound machine, etc.) and measure a physical property of the human body (e.g. radiodensity, the opacity to X-rays). These images are interpreted by domain experts (e.g. a radiologist) for clinical tasks (e.g. a diagnosis) and have a large impact on decision making of physicians.
Medical sub-fields were Deep Learning can be applied:
Neuro, retinal, pulmonary, digital pathology, breast, cardiac, abdominal, musculoskeletal.
Sources of Imaging Data:
- US: Ultrasound
- MR/MRI: Magnetic Resonance Imaging
- PET: Positron Emission Tomography
- MG: Mammography
- CT: Computed Tomography
- H&E: Hematoxylin & Eosin Histology Images
- RGB: Optical Images.
Applications of Medical Imaging:
- Detection/ Localization
- Image Reconstruction and Post-Processing
Datasets to Start with:
Manually annotated radiological data of several anatomical structures (e.g. kidney, lung, bladder, etc.) from several different imaging modalities (e.g. CT and MR). They also provide a cloud computing instance that anyone can use to develop and evaluate models against benchmarks.
Cancer imaging data sets across various cancer types (e.g. carcinoma, lung cancer, myeloma) and various imaging modalities.
A collection of biomedical imaging challenges in order to facilitate better comparisons between new and existing solutions, by standardizing evaluation criteria.
A collection of diagnostic and lung cancer screening thoracic CT scans with annotated lesions.
Diabetic retinopathy is the leading cause of blindness in the working-age population of the developed world. It is estimated to affect over 93 million people. High-resolution retinal images that are annotated on a 0–4 severity scale by clinicians, for the detection of diabetic retinopathy. This data set is part of a completed Kaggle competition, which is generally a great source for publicly available data sets.
A collection of brain MRI scans to detect MS lesions.
Large data set of brain tumor magnetic resonance scans. They’ve been extending this data set and challenge each year since 2012.
In second part will dive into code with a dataset and detailed Analysis.
Keep Reading !!!