Implementing artificial intelligence for electrocardiogram interpretation: A case study
Background: Artificial intelligence (AI) is expected to have a growing role in medical diagnostic interpretation and existing programs should be challenged with difficult cases in clinical practice senerios. An isolated posterior myocardial infarction (MI) is suggested by ST segment depression in th...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-03-01
|
| Series: | JEM Reports |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2773232024000622 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Background: Artificial intelligence (AI) is expected to have a growing role in medical diagnostic interpretation and existing programs should be challenged with difficult cases in clinical practice senerios. An isolated posterior myocardial infarction (MI) is suggested by ST segment depression in the anteroseptal leads on a standard 12-lead electrocardiogram (ECG) and confirmed by the presence of 0.5mm ST segment elevation in any of the posterior leads (V7-V9). Isolated posterior MI is rare (potentially <4 % of all MIs). Case report: We present a case of a 79-year-old man who presented with intermittent chest pain and subtle ECG changes concerning for a posterior MI. His catheterization images confirm a completely occluded LCx artery. We also present the AI analysis of the ECG's crucial for making the diagnosis in this case.Why should an Emergency Physician be aware of this?Given the diagnostic challenge of posterior wall MIs with a standard 12-lead ECG, clinical suspicion for a posterior MI should remain high with any degree of ST segment depression in the anterior leads and prompt the emergency physician to obtain a posterior ECG. AI-based ECG interpretation was able to determine that this patient was having an occlusive myocardial infarction. We discuss how to utilize the third-party AI for diagnostic aid. |
|---|---|
| ISSN: | 2773-2320 |