This article is more than 5 years old. Information may no longer be current.
AI accurately identifies cardiac rhythm devices
Click Here to Manage Email Alerts
Artificial intelligence via a neural network exceeded human performance by accurately identifying the manufacturer and model of pacemakers and defibrillators in patients who underwent radiography, according to a study published in JACC: Clinical Electrophysiology.
“Unfortunately, current models are slow and outdated, and there is a real need to find new and improved ways of identifying devices during emergency settings,” James P. Howard, MA, MB BChir, clinical research fellow at Imperial College London, said in a press release. “Our new software could be a solution, as it can identify devices accurately and instantly. This could help clinicians make the best decisions for treating patients.”
Researchers analyzed radiographic images of 1,676 devices from 1,575 patients that were implanted between February 1998 and May 2018. Devices included in this study consisted of 45 models from five manufacturers: Biotronik, Boston Scientific, Medtronic, Sorin and St. Jude. A training set of 1,451 images was used to develop the neural network. An additional 225 images with five examples of each model were also included in the testing set.
The final neural network had an accuracy of 99.6% to identify the manufacturer of a device (95% CI, 97.5-100) and 96.4% to identify the model (95% CI, 93.1-98.5).
The median identification of the manufacturer of a device was 72% when five cardiologists used the flowchart (range, 62.2%-88.9%). These cardiologists were unable to identify the model group.
The neural network was superior in identifying the manufacturer of devices compared with all of the cardiologists (P < .0001 compared with median human identification; P < .0001 compared with best human identification).
“This study demonstrates this neural network has superior accuracy in identifying the manufacturer of a device compared with that of human cardiologists and electrophysiologists using a flowchart approach,” Howard and colleagues wrote. – by Darlene Dobkowski
Disclosures: Howard reports no relevant financial disclosures. Please see the study for all other authors’ relevant financial disclosures.
Click Here to Manage Email Alerts