Study Shows that Human Technicians Outperform Artificial Intelligence in Reviewing Heart Monitor Data
Dr. Mark E. Willcox of Alaska Heart & Vascular Institute performed a study which discovered that human technicians outperform artificial intelligence (AI) in interpreting data by more than 200 percent.
Initially, the study was on comparing the performance of two common cardiac monitors the Carnation Ambulator Monitor (CAM) and the BioGuardian. Dr. Willcox’s findings were quite significant and were presented at the Heart Rhythm Society’s annual conference. His findings started discussions in the industry about the use of artificial intelligence in cardiac monitors and their relationship with professionals. It also triggered detailed conversations on the future of cardiology.
While this pilot study had a limited size, it found a substantial difference in the diagnostic accuracy between the different screening processes of the different cardiac monitors. When the BioGuardian monitor collects the data, an algorithm first screens it and passes the significant findings to a human. On the other hand, the CAM monitor’s data is first screened by a human technician and AI helps them in recording and tracking any irregularities found in the data.
Dr. Willcox points out that the algorithms are at a disadvantage for performing early detection because their identification system is based on finding pre-defined sequences.
Although this pilot study shows the superiority of human technicians over AI-based monitoring systems, we can’t throw out AI and machine learning-based technologies. Experts believe that AI will become an integral part of modern health monitors as they get more common. Popular health trackers right now are Apple Watch and Fitbit.
Source: Alaska Health & Vascular Institute