Neural network approach for the computer-aided diagnosis of coronary artery diseases in nuclear medicine

The computerized scheme developed can aid radiological diagnosis in the detection and classification of coronary artery diseases in ²⁰¹Tl myocardial single photon emission computed tomography bull´s eye images by use of artificial neural networks. The multilayer feedforward neural network used with a backpropagation algorithm has 41 256-input units, 50 to 100 units in a single hidden layer, and eight output units. The neural networks, consisting of two major networks for extent and severity images, were trained using pairs of training input data and desired output data. The results show that the propagation performance of the neural-network-based system was comparable to that of experienced radiologists