Abstract: A frequency-analyzed signal from a pulsed Doppler ultrasound examination is processed in a dedicated computer using statistical pattern recognition to assess the presence and extent of arterial disease. The Doppler blood flow signal is derived from a suitable utlrasound scanner, is frequency analyzed with a real-time, fast Fourier Transformer spectrum analyzer, and is processed by the computer to achieve a diagnosis of the degree of stenosis of the carotid artery through pattern recognition between the Doppler signal and a database of known preclassified spectra. The diagnosis ordinarily follows a hierarchical decision making format to objectively classify the condition of the patient.