AF and AI

Q: Computer (AI) based interpretation for Atrial Fibrillation (AF) of EKG is usually accurate?

A) True

B) False

Answer: B

To date, the data does not support the reliability of automated computer or Artificial intelligence-based readings of AF. Ten to thirty percent of computer ECG interpretations may misdiagnose AF. This is also because the algorithms used by various computers and AI are not standardized, validated, or well-clarified.

Said that some of the newer studies claimed (reference # 3 and 4) that a new deep densely connected neural network (DDNN) based on different principles than the usual method via machine learning, and Supervised Contractive Map (SVCm), reached an overall mean accuracy of 95%. This claim needs to be validated and endorsed by relevant societies. The best we can say is that "the Jury is still out".

To date, each ECG computer should be read by a trained clinician, although computer reading may be of assistance.

#cardiology

#AI

References:

1. Bogun F, Anh D, Kalahasty G, et al. Misdiagnosis of atrial fibrillation and its clinical consequences. Am J Med 2004; 117:636.

2. Lindow T, Kron J, Thulesius H, et al. Erroneous computer-based interpretations of atrial fibrillation and atrial flutter in a Swedish primary health care setting. Scand J Prim Health Care 2019; 37:426.

3. Buscema PM, Grossi E, Massini G, Breda M, Della Torre F. Computer Aided Diagnosis for atrial fibrillation based on new artificial adaptive systems. Comput Methods Programs Biomed. 2020 Jul;191:105401. doi: 10.1016/j.cmpb.2020.105401. Epub 2020 Feb 19. PMID: 32146212.

4. Cai W, Chen Y, Guo J, Han B, Shi Y, Ji L, Wang J, Zhang G, Luo J. Accurate detection of atrial fibrillation from 12-lead ECG using deep neural network. Comput Biol Med. 2020 Jan;116:103378. doi: 10.1016/j.compbiomed.2019.103378. Epub 2019 Aug 2. PMID: 31778896.