The use of artificial intelligence in clinical research shows great promise for transforming drug and device development.
AI refers to algorithms that train software to perform certain tasks and improve upon them over time by constantly analyzing new data.
For instance, AI could enhance the performance of medical devices, allowing them to assist with image analysis for earlier detection. It could improve study designs and assist with recruitment, combing through large volumes of medical records to more quickly identify ideal patient populations. It could enhance data collection and data quality, leading to faster drug and device development. AI could also lead to more targeted treatment and better patient outcomes.
What Is The FDA’s Stance On Artificial Intelligence In Clinical Research?
The FDA recognizes the role AI can play in spurring medical innovation. In 2019, the FDA announced steps it would take to consider a regulatory framework for the development of medical devices that use AI algorithms.
The FDA had already authorized an AI-based device for detecting diabetic retinopathy and a second AI-based device for alerting providers of a potential stroke in patients. Former FDA Commissioner Scott Gottlieb called the authorization of these technologies “a harbinger for the confluence of AI and medical devices” in his statement.
AI-enabled medical devices use “locked” algorithms, meaning they don’t continually adapt or learn. Instead, they’re modified by the manufacturer at intervals, then “trained” using new data. However, that could change in the near future.
For example, an algorithm that detects cancer could become more accurate at identifying different types using real-world data that would take a physician an entire lifetime to analyze.
The FDA will require careful oversight to ensure the benefits of these advanced technologies outweigh the potential risks to patients.
“With artificial intelligence, because the device evolves based on what it learns while it’s in real world use, we’re working to develop an appropriate framework that allows the software to evolve in ways to improve its performance while ensuring that changes meet our gold standard for safety and effectiveness throughout the product’s lifecycle,” Gottlieb stated.
How Could AI Impact Recruitment for Clinical Trials?
Recruiting participants for clinical trials is often the most time-consuming and expensive stage of a study, according to Chunhua Weng, a biomedical informaticist at Columbia University, in a Nature article.
With that in mind, AI could be put to work searching doctors’ notes and pathology reports for people eligible to participate in a clinical trial.
Clinical researchers don’t do this because—aside from the sheer number of notes—the text of these documents is often unstructured, which means it requires background knowledge or context to understand. One doctor might describe a heart attack as a myocardial infarction while another notes it as an MI. A branch of AI called natural language processing (NLP) enables computers to analyze the written and spoken word. An NLP algorithm can be trained to recognize synonyms of the same condition, making it easier for researchers to identify eligible patients.
Another way AI might aid recruitment is by helping patients find clinical trials for which they might qualify.
Weng and her colleagues developed an AI-enabled software that reads the more than 300,000 global clinical trials on ClinicalTrials.gov and generates basic questions to assess patients’ eligibility.
How Can Artificial Intelligence Improve Trial Accuracy?
Once a patient enrolls in a study, he or she receives the experimental study drug or device. Each day, they fill out a patient diary, noting when they took the drugs or used the device and any adverse reactions. But relying on self-reporting is inefficient and introduces a greater potential for error.
Patients who don’t adhere to the rules of the clinical trial put their health at risk and can also interfere with the accuracy of study outcomes.
AI startups are working to solve this challenge.
The mobile platform AiCure uses image and facial recognition algorithms to ensure patients follow instructions. Patients use their phones to take a video of themselves swallowing a pill, and AiCure confirms that the right person took the right pill.
Another startup in this category, Catalia Health, is developing a healthcare companion and coach using AI. Catalia hopes to enforce behavioral changes in patients by asking specific questions, setting reminders, and tailoring conversations to each patient.
An AI assistant’s ability to successfully enforce lifestyle changes largely depends on patients’ willingness to interact with it on a daily basis. The interaction itself could be monitored using both AI and Internet of Things (IoT) sensor technology.
How Can AI Improve Treatment?
Certain types of AI algorithms are used to detect patterns in vast amounts of data and interpret its meaning. An algorithm could comb through data to make an informed prediction of a particular cancer type based on a combination of genes. It could help target the success of a particular therapy, choosing the right dosage based on the treated patient statistics and other information.
AI could even help physicians make more informed, data-driven decisions about what drug or device is best for a patient’s specific needs.
One AI platform, Abtrace, is designed to help doctors prescribe the right antibiotic for each patient by analyzing millions of data points using natural language processing and recognizing patterns.
According to the company, 30% of antibiotic prescriptions are inappropriate, which contributes to a growing resistance to antibiotics.
It’s not hard to imagine artificial intelligence being used in similar ways to help medical devices offer more precise, targeted treatment.
AI algorithms could comb through data from a patient’s wearable heart monitor or sleep tracker, for instance, and help physicians recommend a personalized plan for them.
Artificial intelligence could even analyze data to predict when a medical device will fail.
What’s Next For AI In Clinical Research?
AI takes traditional data analytics to a new level with the ability to review large volumes of data, make conclusions and generate more reliable results over time.
This can help clinical researchers recruit patients faster, design better trials and improve accuracy. It can also help them improve patient care by developing more personalized drugs and devices.
It’s clear that artificial intelligence has the potential to revolutionize the way we plan and conduct clinical trials. Exactly how researchers will harness its power and how regulators will oversee its use remains to be seen.
Our team takes pride in keeping up with the latest developments in clinical research, including the use of artificial intelligence in clinical trials and medical devices.
We can help your team take advantage of new technologies while ensuring you maintain regulatory compliance.
To learn more about other emerging medical device trends, check out our latest whitepaper.