Deep learning-based image analysis helps automate the search for biological abnormalities
Medical imaging analysis has traditionally required the flexibility of a human inspector and the ability to make qualitative decisions about an unstructured scene. It can be time consuming and difficult to locate an object or region of interest with precision due to either a confusing background or image quality issues. An automated system must successfully identify the region of interest while ignoring irrelevant features. Today, deep learning-based image analysis can automate the search for biological anomalies in radiological x-rays, ultrasounds, and NMRs.
Whether searching for a specific anomaly or any deviation from the body’s normal appearance, Cognex Deep Learning combines the flexibility of a human inspector with the speed and robustness of a computerized system. The location tool identifies the region of interest (e.g. a certain organ), despite the visually confusing and poorly contrasted background, by learning the distinguishing features of that area. The defect detection tool develops a reference model of an organ’s normal appearance, as well as specific types of anomalies, based on training on a set of sample images. Any anomalies which digress from the normal physiology of the targeted zone are flagged as defects for a CAD computer-aided diagnosis by an expert radiologist.