Knowledge discovery approach to automated cardiac SPECT diagnosis

摘要: The paper describes a computerized process of myocardial perfusion diagnosis from cardiac single proton emission computed tomography (SPECT) images using data mining and knowledge discovery approach. We use six-step process. A database consisting 267 cleaned patient SPECT (about 3000 2D images), accompanied by clinical information physician interpretation was created first. Then, new user-friendly algorithm for computerizing the diagnostic designed implemented. were processed to extract set features, then explicit rules generated, inductive machine learning heuristic approaches mimic cardiologist's diagnosis. system is able provide computer diagnoses studies, can be used as tool cardiologist. achieved results are encouraging because high correctness diagnoses.