Promising Role Of Artificial Intelligence In Clinical Research - Nidhi Sharma
We are living in the new reality where things are moving rapidly and adapting fast to the new technology. From
speech recognition to driving direction to the smart reply to emails, artificial intelligence (AI) has made our lives
much easier. We all have experienced that phase when hospitals were closed for months due to pandemic and in
the race of developing the effective vaccine, there has been a huge focus on clinical trials in a ways that was never
seen before. In the recent past, AI techniques have brought the largest paradigm shift to healthcare particularly in
the field of clinical research. Machine learning (ML), a rapidly growing field of AI that allows super computers to act
as a guide for clinical diagnosis of a patient with specific indication or disease through fine image analysis has
brought a huge revolution in the clinical trial industry.
Two major factors that have made AI-ML so impactful are: firstly, due to high availability of medical data in the
form of medical history at the healthcare settings and secondly, the introduction of complex algorithms in order to
process and analyze data consisting of vast attributes. There is an enormous opportunity in clinical research
though emphasizing on some of the key focus areas of AI Application in clinical trials like Patient recruitment,
Patient compliance, AI-ML based software as medical device, and drug discovery.
Achieving subject recruitment targets and patient compliance are the major contributors of clinical trial success
and most often, clinical trials fail due to inadequate subject enrollment and/or due to non-adherence observed
from the subject’s side. AI systems built through vast amount of electronic and medical record data as well as the
protocol of interest, quickly identify patients that would be eligible for the protocol thereby, vastly improving
patient recruitment time. Likewise, the potential clinical trial patients can share their digital data with the
investigators with help of technology. Sensor evolution in pills to track adherence to medication taken by the
patient and electronic devices help coach trial participants effectively during the course of clinical trial project.
One of the finest benefit of AI-ML software exists in its ability to auto-learn from real-world practice and
experience, and its capability to expand its performance. From the identification of diabetic retinopathy on fundus
screening images to simplifying the complex MRI scans analysis, AI/ML has taken a center stage. There are many AI
based wearable health trackers (EKG, Heart rate, sleep cycle, breathing rate, activity level, blood pressure) that are
being used in the clinical trials to monitor health of trial participant. Emotional state / mental health and vitals
monitoring at home really make new ways in crowdsourcing of the data. Virtual health care assistants just like Siri
& Alexa are helping out trial subjects in Self-care, basic clinical advice and scheduling a medical appointment.
Home based or decentralized clinical trials have been the means the clinical research industry has responded to be
able to meet patient safety and continuity on trial medication during pandemic.
AI is also being used in different domains of Drug discovery including drug design and screening to discover ligands
of special interest in order to identify treatment that are targeted to specific biologics of different diseases. AI/ML
beyond thinking of just disease identification to subset diseases at molecular and other levels will be helpful. It is a
well-known fact that several biopharma players have collaborated with the software companies to advance better
drug candidates into the clinic, validate targets better, and improve patient recruitment and progress how clinical
trials were conducted conventionally. However, responsible data sharing could avoid duplication of efforts in
collecting the data and could also bring down the cost of running a clinical trial. Data sharing will also strengthen
the collaboration between institutions & the scientific community leading to reshaping of the future directions in
the improved patient health outcomes. Although data transfer laws need to be more stringent in order to refrain
from any legal liabilities and obligations.
There has been an overestimation about the fact that the AI machines will be soon replacing the human
component. We should not be skeptical about that fact since role of physician can never go away because of the
human element and moreover doctor’s empathy for their patients is not something that an algorithm can
replicate. However, we should not forget the fact that certain tasks that are currently performed by highly
specialized physicians can be completely replaced by super computers that run AI systems but only to an extent
where the time-consuming & repetitive cognitive functions could be automated. Medical diagnosis without human
interpretation is quite unrealistic as the treating physician has to take the ultimate responsibility of validating a
clinical decision or the risk forecasting. With the robust AI-ML systems, Physicians partnering with AI as a decision
making aid in their medical practice will see their healing power enhancing to ten folds in the coming times.
Futuristic AI-ML aims to identify drug faster, patients faster and getting through the clinical trial in less time & less
However, like other emerging technologies, AI systems for automated risk detection require robust evaluation for
clinical effectiveness before broad adoption. In the beginning of 2021, the U.S. Food and Drug Administration
issued its long-awaited action plan concerning the regulation of artificial intelligence (AI) and machine learning
(ML)-based Software as a Medical Device (SaMD) when intended to treat, diagnose, cure or prevent medical
condition. In short, this plan is a reflection of FDA’s commitment to encourage and harmonize the development of
Good Machine learning Practice (GMLP) and providing guidance to clinical scientists, data scientists, machine
learning researchers & the software engineers. The new considerations will largely support AI researchers to foster
highly collaborative and inter-disciplinary robust model in delivering highly effective clinical trial outcomes. Indeed,
AI-ML based technology platform hold an incredible promise by catering the research innovations for several
medical specialties and has so far commendably improved the quality and value of care one can offer to the clinical
Nidhi Sharma (MBA, MSc, PGDPM)
The author is trained clinical research professional with extensive project management experience. She has
successfully led sites through several regulatory audits and has been associated with training the research teams. Formerly working with Thumbay Research Institute of Precision Medicine at the Gulf Medical University, Ajman (UAE) in the capacity of Associate manager – Clinical Research Operations, she has demonstrated ability to deliver administrative and research support and direction to the clinical and academic research operations. She has also served as Member secretary for Institutional Review Board/ IEC. She is passionate about application of artificial intelligence in clinical research and quality assurance.
Residence address: Apartment 207 – Silicon Residence 2, block C, Dubai Silicon Oasis- Silicon Heights II, Dubai, United Arab Emirates
LinkedIn address: www.linkedin.com/in/nidhi-sharma-ab287618