Improved classifications of myocardial bull's-eye scintigrams with computer-based decision support system

In a recent study, artificial neural networks were trained to detect corona ry artery disease using scintigraphic data as input. The performance of the networks was better than that of human experts using coronary angiography as a gold standard. in clinical practice, this type of neural networks will not take over the decision-making process from the physician but will assi st by proposing an interpretation of the scintigram. The purpose of this st udy was to assess the influence of such decision support on the interpretat ions of the physicians. Methods: A population of 135 patients who had under gone both myocardial Tc-99m-sestamibi rest/stress scintigraphy and coronary angiography within a 3-mo period was studied. An image set consisting of t he bull's-eye rest, stress, difference and quote images was constructed far each patient. Three experienced physicians independently classified all im age sets regarding the presence and/or absence of coronary artery disease i n two vascular territories using a four-grade scale. The physicians classif ied the image sets twice with and twice without the advice of artificial ne ural networks. Results: The joint evaluation of the three physicians showed significantly improved performance with decision support, measured as incr eases in the areas under the receiver operating characteristic curves from 0.65 to 0.70 (P = 0.018) and from 0.79 to 0.82 (P = 0.006) for two vascular territories. Furthermore, the joint evaluation showed significantly less i ntraobserver and interobserver variability with decision support. Conclusio n: Physicians classifying myocardial bull's-eye images benefit from the adv ice of artificial neural networks. These results show the high potential fo r neural networks as clinical decision support systems.