Computer Vision is Providing a New Lens Into Biopharma Innovation | American Pharmaceutical Review
By Carlos Melendez, COO,Wovenware
Computer vision (CV), a form of AI that uses an algorithm to identify, process and analyze vast amounts of text, images and videos that would be impossible to address with the human eye, has been gaining traction in many industries. It’s helping city planners identify traffic patterns by counting vehicles on highways; or for government workers to read forms and documents and identify missing or questionable information; and more recently to identify faces with masks to ensure COVID-19 compliance in public areas. In fact, according to a Reports and Data report, the global computer vision market is forecast to reach USD 25.69 Billion by 2028.
The Science Lab of the Future
Computer vision is also playing a major role in pharma and life sciences and ushering in new standards for intelligent medical devices, diagnostics and treatments that will have lasting repercussions on the people and the world at large. The scientific lab of the future will be one where human intelligence is amplified by artificial intelligence and most notably computer vision. Given the need to more quickly commercialize pharmaceuticals and medical devices, enable faster diagnoses and better outcomes, digital transformation via computer vision will become the norm in life sciences. Consider the following applications:
Vaccine delivery. Since science is grounded in data, computer vision can sort through thousands of biological images to identify those that are most likely to trigger a robust immune response. It’s able to identify the proteins that make up a virus, as well as those that are mutating.It’s also accelerating the discovery of new molecular combinations, tracing toxicity markers and disease triggers.
Rapid at-home testing. Computer vision is being applied to rapid at-home tests to identify markers to indicate the presence of COVID-19 and other viruses, or to confirm pregnancy results. In the very near future, a mobile app for a rapid antigen single-use, self-test for COVID-19 will be available via a mobile app that can not only provide immediate diagnosis, but also remove the need to have an overtaxed lab or trained professional confirm the diagnosis.
Medical diagnostics. New computer vision models are being used with X-ray and MRI images to automatically read scans with extreme accuracy, augmenting the work of medical professionals by helping them more quickly identify the presence of disease, recognize anomalies and assess the severity of the disease in order to make faster treatment decisions.
Clinical trials. Using computer vision in clinical trials could help assess efficacy of drugs and their impact on patients by tracking and identifying the vast amounts of datasets collected during the clinical trial process.Researchers spend inordinate amounts of time observing how specific treatments are working. Computer vision apps, for example could be used on smart phones to monitor trial participants, comparing health conditions to past data.
Assessing biomedical data. Beyond clinical trial data, most biomedical data is unstructured, residing in different systems and formats, such as scientific publications, clinical records, and other systems. Computer vision can help to analyze documents and images and identify biologically relevant information and correlations between data.
Regulating AI in Biopharma
As computer vision and other forms of AI emerge as key tools in life sciences, the regulatory environment for AI-driven pharmaceutical methods and medical devices is evolving and it can be questionable as to how forms of AI are regulated by the FDA. According to a Pew report, “currently, the FDA regulates some AI-enabled healthcare products and it is considering how to adapt its review process for AI-enabled medical devices that have the ability to evolve rapidly in response to new data.” The report goes on to explain that the FDA reviews software based on its intended use and the level of risk to patients. Since it’s intended to treat, diagnose, cure or prevent disease, the FDA currently classifies it as a medical device.
Since AI algorithms in general, must be continuously trained based on new data and changing situations, new inspections guidelines are evolving, which may require new FDA approval each time algorithms are significantly modified. Yet, the key goal is to ensure that AI-assisted discovery, diagnostics and treatments are enabling safe and effective outcomes.
Computer Vision in the Future
Despite its nascent status in life sciences today, computer vision has the potential to improve outcomes outside of the traditional lab or clinic. Consider its implications in treating patients in wars or natural disasters. Hand-held MRI or X-ray machines, equipped with AI algorithms, could help detect conditions on the spot so that patients can receive more effective treatment based on highly accurate diagnoses. This type of intelligent diagnosis could also enable better triaging of patients who require urgent care.
Challenges to Adoption
While there are many opportunities for computer vision in life sciences, and it’s already in use today by leading pharma firms, the primary barrier to adoption is currently cost. Computer vision integration requires a transformative shift, as well as significant investment by companies to build the tech infrastructure. Currently there is a lack of affordable CPU power to address the growing need to process the huge datasets required. As major tech players, such as Apple and Intel, work to deliver affordable CPU power, more and more providers will join the effort, driving down CPU costs. Cloud and AI innovation, as well as advanced edge computing capabilities are here today, and will proliferate the market as the costs barriers to entry are lowered.
As life sciences and medical professionals have seen all too clearly over the past two years, innovation abounds in how cures are found, patients are treated and infection spread is managed. Computer vision is playing a vital role, so that overworked clinicians and life science professionals can receive the support they need to deliver highly accurate data-driven results and focus their efforts on delivering the most optimum patient outcomes.
About the Author
Carlos M. Meléndez is the COO and co-founder of Wovenware, a Puerto Rico-based design-driven company that delivers customized AI and other digital transformation solutions that create measurable value for government and private business customers across the U.S.
Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!