This video demonstrates the operation of a fully automated computer vision diagnostic technology, using FM microscopy. The integration of the equipment, and the development of the detection algorithms and decision support application, has been developed by Signature Mapping Medical Sciences, Inc., a subsidiary of Applied Visual Sciences, Inc.
Signature Mapping Medical Sciences is an image analysis software development company. Using image processing and pattern recognition science, the company has created a fully automated computer vision diagnostic product, TBDx, designed to find mycobacterium tuberculosis in digital images captured from sputum smears, using a camera and laboratory microscope. The technology integrates a 200-slide loader, an automatically focused digital camera, and a microscope with automated slide stage. The camera acquires 100 digital fields of view, which are transferred and stored in a computer workstation. Computer vision algorithms and decision support software analyze these digital images to detect and classify patient cases as TB positive or negative. Each slide can be analyzed in 2 minutes; 200 slides are processed in less than 7 hours. This computer vision diagnostic technology can be used to identify the most probable positive TB cases, which can be confirmed by molecular diagnostics and tested for drug susceptibility. The performance improvements and economic impact of combining computer vision and molecular diagnostic technologies can be substantial. As the microscope is the primary throughput device, this digital image acquisition hardware / software diagnostic platform can be used for multiple infectious disease applications.