Figure 6.

AI-ECG visualization of a 46-year old female with older ECG-age. A 46-year-old woman who had type 2 diabetes mellitus, hyperlipidemia under regular OPD follow up. Her ECG-age is 61.4 years old which is much higher than her chronological age. Using the class activation mapping and attention mechanism to explain the AI-ECG prediction, we used white-to-red gradient to indicate the importance of each lead, and the darker-to-light gradient to indicate the contribution of each position in prediction of ECG-age. Light green and yellow mean older and younger rhythms. The most important part in this case is aVL, which accounts for 18.6%, while AI considered it was an old feature with widely green color. Although it presents sinus rhythm, we can see the part emphasized by AI-ECG shows relative irregular baseline, which may be caused by a patient with muscle tremor, muscle tension, dry skin turgor, and Parkinson's disease. We considered that AI-ECG may acquire information from these tiny changes. The patient was diagnosed of coronary artery disease, single vessel disease status has post-percutaneous coronary intervention with a drug-eluting stent implanted at left circumflex artery after 3 years.