ST-segment isolation and quantification as a means of improving diagnostic accuracy in treadmill stress testing

Abstract

A new method of ST-segment analysis utilizing computer-analytic techniques has been employed in treadmill exercise testing, with the aim of enhancing diagnostic sensitivity and specificity. One hundred thirty-three individuals were studied, including 62 normal subjects (Group I), 29 patients with coronary disease and clear, “ischemic” ST-segment responses to exercise testing (Group II), and 42 patients with coronary disease but normal or nondiagnostic exercise tests (Group III). The techniques used included: computer averaging, to minimize motion artifact and baseline drift; a means of isolating the ST-segment from the T-wave and quantifying ST-amplitude and slope (the isolated ST integral, IST); and the relating of the IST to a given heart rate, thus taking cognizance of the dependency of ST-depression on heart rate and level of exercise. These methods resulted in a test specificity exceeding 90 per cent and a sensitivity of over 85 per cent. Further evidence of the improved sensitivity achieved using these techniques included a 79 per cent (33 of 42) recognition of abnormalities in Group III, patients having normal or nondiagnostic visually interpreted treadmill stress tests (i.e., no flat or downsloping ST-segments of 1 mm. or greater). Moreover, 15 of 29 patients in Group II (52 per cent) manifested abnormal IST's before development of a typical “ischemic” ST, and in 17 patients (59 per cent), the IST continued to remain positive after disappearance of the characteristic flat or downsloping ST-segment. It is concluded that this type of computation analysis adds appreciable diagnostic sensitivity and specificity to treadmill stress testing.