An artificial intelligence program for the radiologic diagnosis of brain lesions

A software program has been developed that uses a frame-based expert system for differential diagnosis in neuroradiology. A frame-based expert system is used to store the magnetic resonance (MR) and computed tomographic (CT) imaging characteristics of over 100 known brain disorders in object-like entities. The frames are organized in a hierarchic structure in which lower order frames inherit attributes from higher order frames, with the highest frame containing information that applies to all the other frames. Program execution follows a consultation paradigm with a dynamic database. A decision tree menu provides a user-friendly interface with which to navigate through the network, based on features of the lesion as depicted on MR and CT images. The system can provide a differential diagnosis based on the MR imaging findings alone with information criteria including the signal intensity of the lesion on T1- and T2-weighted images, the location of the lesion, and the presence or absence of mass effect. The differential diagnosis may be further refined by adding CT-related information, including CT attenuation and the presence or absence of calcification and contrast enhancement.