Why Medical Imaging is the Next Frontier for Deep Learning
Jorg Vandenhirtz, 04-01-2019
Medical imaging, which includes radiological x-rays, ultrasounds, and NMRs, has traditionally required the flexibility of the human eye to detect anomalies. Computers are famously confused by busy backgrounds and image quality issues, such as specular glare. This makes it incredibly difficult for traditional machine vision algorithms to locate an object or region of interest with precision—especially to identify abnormalities amidst an unstructured scene. It can be time consuming and difficult—if not impossible—for an automated system to successfully identify the region of interest while ignoring irrelevant features.
Today, however, deep learning-based image analysis can automate the search for biological anomalies reliably and repeatedly and robustly. This is changing the role of today’s radiologist, who can now take advantage of a computer-aided diagnosis (CAD) for medical imaging. Whether searching for a specific anomaly, such as tumor, or any deviation from the body’s normal appearance, Cognex ViDi combines the flexibility of a human inspector’s eye with the speed and robustness of a computerized system. Two specialized tools aid this process. The ViDi Blue-Locate tool locates the region of interest, such as a certain organ, even when the background is visually confusing or poorly contrasted. The ViDi Red-Analyze tool uses a set of training images to develop a reference model of that organ’s normal appearance, as well as specific types of anomalies, so that it can flag any anomalies which digress from the normal physiology of the targeted zone as defects.
Some great examples include using deep learning-based tools to localize and identify organs or implants in an x-ray. The Cognex ViDi Blue-Locate tool can locate a specific organ by learning its distinguishing features. To train the Blue-Locate tool, all you need to provide are images where the targeted features are marked. Similarly, deep learning-based defect detection and segmentation tools, like ViDi Red-Analyze, can help identify anomalies in medical image. ViDi Red-Analyze develops a reference model of an organ’s normal appearance, as well as specific anomalies, based on a set of sample images. Any anomalies which digress from the normal physiology of the targeted zone are flagged for a CAD computer-aided diagnosis by an expert radiologist.
To learn more about exciting new deep learning applications in medical imaging, download the free guide: VisionPro ViDi: Deep Learning-Based Solutions for Life Sciences.