Clinical artificial intelligence (AI) firm, Mendel, has announced results from a trial that found its Neuro-Symbolic system can outperform current systems such as GPT-4 in sorting patient cohorts from electronic medical records (EMR).
Titled ‘ACR: A Benchmark for Automatic Cohort Retrieval’ and conducted alongside Cornell University, the report found that the fact that Mendel’s AI system could outperform the industry standard of GPT4 “highlights the potential of integrating expert knowledge with language learning models (LLM) in healthcare.”
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
Mendel says that its AI approach couples LLMs with a proprietary hypergraph reasoning engine that can sort through patient cohorts and identify what patient may be eligible for a clinical trial. Being able to identify patients appropriate for a certain clinical trial is pivotal for the clinical trials industry as it is the means by which trial cohorts are formed.
As part of the evaluation, the Mendel AI was fed a dataset comprising 1,400 patient records.
Wael Salloum, co-founder and chief science officer at Mendel, said: “Our latest research at Mendel marks a significant milestone in the field of AI in general, and healthcare in particular.
“We are the leader in clinical reasoning by coupling LLMs with our hypergraph reasoning, enhancing both the effectiveness and efficiency of patient cohort retrieval. This work is critical in paving the way for more robust and scalable clinical reasoning. This breakthrough underscores our commitment to advance the AI field to transform clinical research and improve patient outcomes.”
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataIt follows after the company published results of another study utilising its AI software alongside nurses with the goal of speeding up the pre-screening of oncology patients for clinical trials, finding the accuracy for a human alongside an AI was non-inferior to human alone (78.7% vs. 76.7%) and both were greater than AI-alone (63.5%).
An excerpt from the newest Mendel study reads: “The fact that a neuro-symbolic approach can outperform GPT4 highlights the potential of integrating expert knowledge with LLMs in healthcare, a domain rich with explicit knowledge.
“Furthermore, neuro-symbolic approaches could enhance the practical adoption of ACR systems in the real world since healthcare professionals often require system control, reduced hallucinations, and consistent, predictable behaviour.”
