New AI Platform Simulates Thousands Of Clinical Trials At Once

An Israeli startup has announced a new version of its AI platform that runs clinical trial simulations on drugs under development, predicting their risks and outcomes. The new version will let the platform run thousands of trials at once. 

QuantHealth says its AI simulator can optimize clinical trials and help identify how participating subjects would respond to treatment. 

The new product, Katina, can simulate hundreds of thousands of protocol combinations (more than one trial element) of patient groups in a clinical trial, in order to maximize the probability of its success. 

Katina is integrated into the QuantHealth platform and is powered by the startup’s proprietary generative AI, which is trained on real world data from 350 million patients and data from hundreds of thousands of clinical trials. 

Once a trial has been simulated, the platform gives different scores to different protocols to help researchers decide which ones will work best for their specific trial. The company says its AI platform has an accuracy of 86 percent.

According to the company, 90 percent of drugs fail the clinical stage, which costs pharma companies $50 billion.

“This release is a game changer for protocol design committees and portfolio optimization teams,” said Orr Inbar, Co-Founder and CEO of QuantHealth.

“With the power of QuantHealth’s generative AI, and now further enhanced by monte-carlo workflow [a model used to predict the probability of a variety of outcomes when there is a potential for random variables], the Katina platform fires off thousands of protocols at once and allows each stakeholder to evaluate the simulations based on what matters to them most,” he said.  

“This high-throughput, holistic approach ensures that when it comes to protocol selection and development strategy, no stone is left unturned, and all voices in the room are heard.”