Changing Technologies, Changing Industry: How Artificial Intelligence and Machine Learning are Driving Innovation in the Life Science Sector

Published on
June 14, 2023
Written by
Saara Meghji
Read time
4 min

Saara Meghji

Research Analyst

With the rapid uptake and influence of Artificial Intelligence (AI) within nearly all facets of life, now is the time for those of us in the healthcare and life sciences sectors to query the relationship between this tectonic-shifting technology and healthcare professionals. The life sciences sector has often been a pioneer of new technologies, leveraging human innovation to increase efficacy, decrease time to market, and improve data management. The rapid uptake of AI and Machine Learning (ML) across all sectors provides the next opportunity for life science companies to lead charge.  

Major influential bodies have already been exploring and discussing the implications of these advancements in technology; specifically, for AI and Machine Learning (ML). Despite their similarities, AI and ML can pose vastly different implications for the healthcare sector. While AI involves the development of technology designed to imitate human behaviour as effectively as possible, ML is a subset of AI that focuses less on general imitation and instead on training a machine to carry out a specific task. Both, however, carry far-reaching implications for the future of healthcare.  

The United States Food and Drug Administration (FDA) addresses the new technologies in its recently published discussion paper on the use of AI and ML in the development of drugs. This accompanies another whitepaper regarding AI in drug manufacturing, which serves as a starting point for various stakeholders to discuss the potential implications of AI within the pharmaceutical industry and beyond.

So, what does the FDA think about these technologies? In the discussion papers, they anticipate three ways AI will have a direct impact on the advancement in drug development, and by extension, the role of healthcare professionals:  

  1. Aid in the identification and speed-to-market of lifesaving drugs. AI and ML can predict the success rates and points of error of various drugs in relation to different health conditions.  This can speed up the process of compound screening by evaluating datasets faster, thereby ‘fast-tracking’ drugs with a greater propensity to succeed during clinical trials.
  2. Improve the efficacy of clinical trials. Clinical research is one of the most significant applications of AI, given the ease with which the technology is used (via large language models, or ‘LLMs’) to analyze large swaths of data. AI can also play a role in analyzing and interpreting results from trials to conduct entry-level analysis, drawing conclusions that can be used by healthcare professionals in continued drug development.  
  3. Introduce simpler ways of pharmaceutical manufacturing. AI and ML methods can be used to assess the status of medical equipment, perform quality control measures on various products, and alert healthcare providers. These methods speed up innovation and optimize efficiencies for healthcare professionals. Gone are the days of being bogged down with quality control, as AI can do the dirty work for scientists who can now prioritize developing better drugs at faster rates.

Despite outlining these various benefits, the FDA has expressed significant concerns for a world with blind faith in AI. Emphasizing the vital role that humans play in the effective implementation of AI, the FDA believes that AI should not be completely unregulated, but rather, that regulations should work under the influence of humans to carry out various tasks more efficiently.

In particular, the discussion paper notes that human-led governance is necessary to maintain accountability and transparency over AI when it comes to Life Sciences. Initial excitement over the capabilities of AI may result in a world with lax regulation and inaccuracies in drug development. The FDA stresses the importance of careful, human-led planning when using AI in drug development. Discontinuation may be necessary if the maintenance of AI becomes counterproductive to an institution’s ability to benefit people. Further, any benefits of efficient scraping and data analysis are mitigated when datasets themselves are flawed. AI may not always see through these flaws, thus generating biased, unrepresentative, or inaccurate results. The FDA urges the importance of human eyes thoroughly looking over data collected and selected by AI to ensure that bias is eradicated from the beginning.

Beyond this whitepaper, life sciences professionals are presently evaluating the immediate impact of Large Language Models, such as Chat-GPT and GPT 4, in healthcare. In a webinar run by FirstWord HealthTech about the use of Generative AI in healthcare, Najat Khan, Janssen’s Chief Data Science Officer, discussed how AI can effectively address pressing problems with greater efficiency than humans. Machine learning can generate reports and knowledge graphs with greater ease and accuracy, allowing professionals to focus their energy on achieving large-scale innovations. AI also significantly helps to tailor drugs to patient needs by scouring through electronic records more quickly than a human. However, Dr. Khan expressed the same warning shared by the FDA; namely AI, within the context of healthcare, is only truly as good as the data it is fed.  

As the world embraces AI and ML, healthcare professionals and data scientists still play a vital role in ensuring that data sets are accurate and diverse.  

At Delphic Research we are encouraged by the advances that the life science industry is experiencing as a result of these new technologies. We continue to be vigilant, monitoring for developing technologies and their applications, keeping clients updated on the trends that they need to know.  

Subscribe to our newsletter

Thanks for joining our newsletter
Oops! Something went wrong while submitting the form.