Logan Wright

Andrew Thomson

With the emergence of more digital health technologies in recent years, artificial intelligence (AI) has rapidly expanded in healthcare for a range of applications including handling of administrative logistics at hospitals and other medical facilities, assessment of patient health data, and analysis of novel molecules in drug discovery. AI applications are also helping to provide a better understanding of disease progression, leading to improvements in diagnosis, monitoring, and treatment of specific diseases and conditions. Multiple sclerosis (MS) is one disease area where AI is providing benefit to physicians and patients. Given the unknown cause and the lack of reliable biomarkers, MS presents significant challenges in diagnosis and treatment. Several companies are incorporating AI solutions into their drug development and commercialization strategies to try to address these challenges.

New computer-aided diagnosis solutions have shown that they can support earlier detection of MS by classifying, quantifying, and identifying diagnostic patterns in medical images, including disease characteristics that may be too marginal for physicians to detect. In one example, researchers showed that a vector machine algorithm can accurately diagnose MS based on levels of plasma selenium, vitamin B12, and vitamin D3 in patients’ blood. This approach allowed researchers to accurately differentiate between patients who were previously diagnosed with MS and a control group based on levels of these biomarkers. C Light Technologies’ eye-tracking platform coupled with machine learning to detect MS using unique algorithms and instruments provides another example of how manufacturers are using AI to diagnose MS.

Earlier detection of MS may improve patient outcomes and allow physicians to potentially treat more patients as well as help decrease the cost burden on healthcare systems. It will be essential to understand who is responsible for managing patient data generated from AI solutions and how these data can be efficiently collected and stored so that health networks or large payer systems can access them while maintaining data privacy.

Lev Gerlovin, Life Sciences Practice, CRA

Innovations in digital disease tracking technology including applications for smartphones are also emerging and enabling the collection of real-world MS patient data in real-time. These technologies allow physicians to monitor their patients remotely, only requiring in-office visits at specific milestones during disease progression and providing added convenience for patients. Monitoring in real-time can also provide a more comprehensive understanding of disease etiology. Researchers at AbbVie and the University of California San Francisco are using wearable biosensors to track MS disease progression to better capture patient prognosis, track their disease progression, and select appropriate clinical interventions.

More companies are also integrating AI solutions into drug discovery platforms to expedite the identification of novel biologic therapies for MS. In one example, AxoSim is leveraging its NerveSim technology to identify neurological therapies for diseases, including MS, based on an in vitro model to monitor the impact of drug candidates on the electrophysiological properties and cell-cell interaction of Schwann cells. This platform more accurately represents human physiology and has the potential to reduce clinical failures and allow companies to develop potentially effective drugs more rapidly and at lower costs.

Effective application of AI technologies could lead to improved clinical outcomes, reduce healthcare costs, provide a platform for precision-based medicine, and accelerate the development of the next generation of MS treatments. Drug developers must understand how to effectively incorporate AI into their drug discovery platforms and approach to developing diagnostic tools. They must also work to consolidate data generated from AI solutions and make these data accessible to key industry stakeholders. But AI applications are not limited to MS – drug developers are working to harness AI in other disease areas, especially those with high unmet needs and poorly understood etiologies, to improve patient care in the years ahead.

The views expressed herein are the authors’ and not those of Charles River Associates (CRA) or any of the organizations with which the authors are affiliated