From July to August 2021, 60 medical staff and general practice trainees in primary care institutions were randomly assigned in two groups with 30 in each group. The control group received conventional ECG training and the study group received ECG training with artificial intelligence image recognition system. The quantitative assessment was conducted before and 2 months after training. There was no significant difference in baseline scores between two groups (43.33±9.94 vs. 44.33±10.33, P>0.005); while the scores of the study group was significantly higher than that of the control group after training (76.00±11.92 vs. 68.00±9.61, P<0.05). Compared with the control group, the correct rates of atrial fibrillation/atrial flutter, sinus arrhythmia and atrioventricular block in study group were significantly higher than those in control group (100.0% vs. 70.0%, 96.7% vs. 73.3%, 86.7% vs. 43.3%; all P<0.05). It is suggested that the application of artificial intelligence ECG recognition system can significantly increase the teaching effect in primary medical training.

Objective:To evaluate the relationship between regional coronary flow reserve (CFR) obtained from cadmium-zinc-telluride SPECT(CZT SPECT) myocardial functional perfusion imaging (MFPI) and invasive fractional flow reserve (FFR) measured during coronary angiography (CAG) and its clinical value in guiding coronary interventions.Methods:Forty-two patients (30 males, 12 females, age (63.3±9.8) years) who completed CZT SPECT MFPI in the First Affiliated Hospital of Nanjing Medical University from June 2022 to September 2022 and underwent CAG within 3 months were included retrospectively. The concordance of CFR and FFR for diagnosing myocardial ischemia (CFR<2.0 and FFR<0.8) was calculated at the vascular level. The diagnostic efficacy of coronary stenosis≥70% for decreased myocardial blood flow (CFR<2.0) was calculated. Kappa test was used to analyze the data. Results:A total of 126 major coronary arteries were identified in 42 patients, of which 30(23.8%) had a CFR<2.0 by CZT SPECT and 33(26.2%) had stenosis≥70% in CAG. A total of 32 coronary vessels were performed with MFPI CFR and FFR measurements, of which 6 were both decreased and 21 were both normal, so the concordance rate was 84.4%(27/32)( Kappa=0.612, P<0.001). Among 33 coronary vessels with stenosis≥70%, 13 were with CFR≥2.0. Among 30 coronary vessels with CFR<2.0, 10 were with stenosis<70%. When using stenosis≥70% to diagnose CFR decreasing, the sensitivity was 66.7%(20/30), specificity was 86.5%(83/96), positive predictive value was 60.6%(20/33), negative predictive value was 89.2%(83/93), and accuracy was 81.7%(103/126). Conclusions:The concordance between CFR and FFR for the diagnosis of myocardial ischemia is good. Nearly 1/3 of the coronary arteries with decreased CFR have stenosis<70%, whereas nearly 40% of the coronary arteries with stenosis≥70% are not result in myocardial ischemia. Regional CFR determined by CZT SPECT may have potentially significant clinical value in the diagnosis of coronary artery disease and decision-making of coronary intervention.