Automated Fusion of Multi-modality Biomedical Images

A novel automated approach of multi-modalitybiomedical image control point detection,registration, and fusion has been successfullydeveloped. The new algorithm, which consists of theAdaptive Exploratory Algorithm for the controlpoint/feature detection and Heuristic OptimizationAlgorithm, is reliable and time efficient. The newapproach has been applied on three ophthalmologicmodalities of nonhuman primate eyes - angiogram,fundus, and oxygen saturation retinal images. It hasachieved an excellent result by giving thevisualization of fundus or oxygen saturation imagewith a complete angiogram overlay. By locking themulti-sensor retinal images in one place, thealgorithm allows ophthalmologists to match the sameeye over time to get a sense of disease progress andpinpoint surgical tools. The new registration andfusion algorithm can be easily expanded to the eye,brain, or body images of human or animals'. Thetarget readers are graduate/undergraduate students,faculty, and research scientists who are interestedin the image registration and fusion algorithms;ophthalmologists and physicians; image processingsoftware developers; and whoever wants to learn imagefusion.