Applying AI to EHR for Drug Repurposing

Main idea: Drug development is expensive, risky and long. It takes an average of 13 years and 2–3 billion dollars to bring a new drug from bench to bedside. Drug repurposing refers to the discovery of new medical indications for an approved or investigational drug, which mitigates risks across all 3 dimensions. The most critical task of drug repurposing is to identify new associations between drugs and diseases and the analysis of heterogeneous data based on Artificial intelligence (AI) methods allows for this analysis to be done computationally in a time and cost-efficient manner.

Why it matters: Electronic Health Records(EHRs), which have rich records of patient and population data over large time spans, can serve as a valuable data source for applying AI/ML algorithms for the drug repurposing.

What you should watch out for:

• Applications to use cases such as rare diseases where investments in discovering new drugs might not be financially feasible

• As AI/NLP methods improve due to the path breaking advances, the opportunities for identifying drugs that can be repurposed grow significantly

Home