Mpirik’s Cardiac Intelligence® Automated Screening Demonstrates 98.6% Accuracy with Moderate and Severe Aortic Stenosis – Medical Alley Association

Mpirik’s Cardiac Intelligence® Automated Screening Demonstrates 98.6% Accuracy with Moderate and Severe Aortic Stenosis

MILWAUKEE, Dec. 14, 2021 /PRNewswire/ — Mpirik, an A.I. healthcare technology company that develops automated patient screening models for cardiovascular disease to drive actionable insights, quality and research, has published research in the Annals of Thoracic Surgery about their screening success.

The study looked at undertreatment of aortic stenosis, a cardiovascular disease with 50% mortality rate at two years after onset of symptoms. Mpirik’s software, Cardiac Intelligence® identifies patients with cardiovascular diseases directly from diagnostic data.

Cardiac Intelligence® analyzed +77,000 echocardiograms using NLP to identify patients with AS with 98.6% accuracyTweet this

Dr. Daniel O’Hair, Chief of Cardiac Surgery at Boulder Community Hospital and Chief Medical Officer for Mpirik explains, “The power of AI is precision at scale. Cardiac Intelligence® succeeds by analyzing vast amounts of real-world electronic health record data in a manner which far exceeds human capacity. Patients at risk from cardiovascular disease can be instantly identified, enabling delivery of the right care at the right time. Complete accounting of every patient enables timely treatment, optimizes patient outcomes and drives value at every level of healthcare.”

The study assessed Mpirik’s Cardiac Intelligence® solution and its ability to reliably identify patients with moderate or severe aortic stenosis from echocardiograms at two geographically disparate healthcare organizations. Over 60,000 patients with over 77,000 echocardiograms were analyzed using Mpirik’s proprietary natural language processing (NLP) engine to identify patients with either textual findings and/or discrete measures meeting guideline-derived criteria.  Results showed the algorithms produced a remarkably high sensitivity (99.1% under-estimation), and high accuracy (98.6% correct classification). (Published June 14 2021, DOI: https://doi.org/10.1016/j.athoracsur.2021.04.106)

“Real world healthcare data today is massive but muddy. AI is a sophisticated solution to compensate for the randomness in much of the data. This understanding of real-world situations inspired us to develop a series of algorithms that uniquely reads text in reports and interprets millions of discrete measurements collected on patients with cardiovascular disease everyday. Our mission is to ensure all patients receive timely and appropriate expert clinical evaluation. This study highlights our unique team which blends clinical expertise and artificial intelligence to solve real-world problems,” says Chris Rogers, Director of Research.

About Mpirik

Mpirik is the developer of a healthcare A.I. platform designed to identify critical findings, filter extraneous data, and facilitate communication by aggregating health data and tracking follow-up. The platform algorithms augment the physician’s ability to deliver excellence by objectively screening patients with heart disease at scale and ensuring their connection to timely care

Keep your finger on the pulse of Medical Alley from anywhere with the ALLEY NEWSLETTER