Accessible computing power and storage has driven new opportunities for artificial intelligence (AI), with the technology finally bringing machine learning (ML) to fruition. ML has demonstrated success in clinical research and to support clinical trial, design, execution and analysis. 

ML algorithms require large, quality datasets for both a training and test set of data. Training data is a set of example data used to fit the ML model and the test set is a set of previously unseen data used to evaluate the performance of the model.

Data used for ML must be factual, insightful and facilitate decisions in the real world.  This avoids biased results, which can lead to bad decisions that impacts people’s health and safety. A Contract Research Organization will have experience working with clinicians and scientists to formulate specific questions, identifying appropriate data sets to address the questions and will have expertise in processing, integrating and analyzing diverse datasets to maximize its value.

ML in Clinical Trials

Efficiency improvements in clinical trials equal savings. For example, an ML platform could allow for running pre-clinical trials to allow early identification of demographics that are most likely to respond to a drug and pinpointing biomarkers that show the most promise for patient response thus refining the compound and the trial design.

ML-based predictive analytics are being used in recruitment, retention and patient engagement.  For example, Identifying and recruiting the right candidates accelerates R&D timelines.  Patient engagement during the trial is increasingly important as providers expand the use of Health IT, including apps and wearables to manage health. 

The 21^st Century Cures Act enacted in 2016, was sponsored by the FDA and designed to accelerate medical product development, bring innovations to trial designs and outcome assessments to speed the development and review of medical products.  The FDA is encouraging an increase of observational studies utilizing real-world evidence and is developing standards and methodologies for their use. This would allow ML analysis of larger datasets, i.e. the interpretation of wearables, data, or electronic health records.

Real-world data may also be used alongside multi-omic mapping of patients receiving an investigational product to create a “digital twin”: a representation of a specific individual reflecting their physiological and molecular status as well as their lifestyle over time. Digital twins could be used to understand, for example, what would have happened to the individual if they had received placebo or standard of care. This approach shows utility, i.e. where a “digital twin” of a patient’s heart is created using medical data and models the unique characteristics of an individual’s heart. This model can be used by clinicians to test different treatment options by comparing possible outcomes without any real risk to the patient.

Data Mining

Pharma companies have a vast R&D database containing years of data from clinical trials, lab experiments, etc. This data contains potential patient insights awaiting discovery with different approaches, for example, by natural language processing, which can sift through previous unstructured research documents for findings that are relevant to current research or by ML approaches across structured data pooled from many clinical studies.

An experienced team can pool together studies and integrate other data types, harmonizing different versions of data dictionaries and standards. These teams can apply ML across clinical and real-world data identifying patterns to inform future trials and research and generating business value.

ML in Clinical Research

IT technology has been dominated by traditional healthcare companies, but today Google, Apple, Facebook and Amazon have brought unprecedented levels of investment and innovation across the healthcare spectrum. Google is applying AI capabilities in the areas of disease detection, data interoperability and health insurance.  An example is the application of its DeepMind Health to differentiate between healthy and cancerous tissue to improve radiation treatment.

In a speech earlier this year, former FDA Commissioner Dr. Scott Gottlieb noted that new streams of real-world data gathered directly from electronic health records and other data sources, paired with advances in ML, will be crucial for creating the next generation of clinical trials.  He stressed the importance of modernizing the clinical trial process to take advantage of this data as well as IoT devices, claims, lab tests, wearable devices and even social media.

New streams of real-world data can radically alter the efficacy of clinical trials but it’s crucial that data science teams be adept at delivering high quality advice and results, not only for standard clinical trials, observational data and studies, but also to help businesses understand what is possible in this new frontier.

---

Jennifer Bradford, PhD is Head of Data Science for the CRO PHASTAR. She previously worked for the Advanced Analytics Group at AstraZeneca, leading the development of the REACT clinical trial monitoring tool, which she later customized and delivered to other sponsors as part of Cancer Research UK (CRUK).  Within CRUK and in close collaboration with the Christie hospital she worked on EDC, app development and wearables data analytics in the context of clinical trials. She has a degree in Biomedical Sciences from Keele University and a bioinformatics Masters and PhD from Leeds University.