Automatic design of an effective image filter based on an evolutionary algorithm for venous analysis

Medical doctors and clinical technologists operate specific, complicated diagnostic systems to assess venous diseases. Instead of using such expensive equipment, low-cost infrared cameras can capture vein images noninvasively and simply. However, the recorded image has a possibility to result in low contrast and a low signal-to-noise (S/N) ratio. An effective image filtering method to estimate venous changes will solve this problem and enable the early detection of disease. For this study, a novel filtering method based on the genetic algorithm (GA) with the expectation–maximization algorithm was proposed for the visualization of vein shapes; its effectiveness was evaluated by images acquired from a near-infrared (780 nm) camera. The novel filter was able to be automatically designed by the GA to improve the worse S/N ratio of vein images, with an unknown correct answer image. If the proposed filtering method is incorporated into e-healthcare applications, it could be widely distributed through smartphones or tablets and facilitate finding abnormal veins at an early stage. © 2015, Springer-Verlag Wien.