Artificial Intelligence Serving Medical and Pharmaceutical Research
Artificial intelligence is increasingly being deployed in the healthcare world to assist caregivers in their practice or to optimize patient care. Another area significantly impacted by AI is medical and pharmaceutical research. Let's decrypt.
In the healthcare sector, R&D teams, whether in fundamental research or pharmaceutical research, have quickly embraced AI-based solutions to speed up the discovery of new treatments, optimize clinical trials, personalize treatments, or shorten development phases.
More broadly, AI can manage and analyze large volumes of medical data, enabling research teams to save time or make informed decisions. It's important to keep in mind that AI-based solutions developed for research are just tools to simplify the mission of research teams and are not meant to replace them.
Machine Learning to Accelerate Clinical Phases
One of the components of artificial intelligence is machine learning, also known as automatic learning, which aims to give machines the ability to 'learn' from data, using mathematical models. (1) It facilitates the acceleration of processes and the handling of a massive volume of data and information, enabling the development of predictive analysis. This technology is now at the heart of R&D strategies.
Artificial intelligence models help speed up development phases and reduce timeframes and, obviously, associated costs.
In the search for new drugs, one of the major challenges faced by scientists in designing active principles is the multitude of data available for analysis (2). To address this issue, drug design that utilizes combinatorial chemistry is now ubiquitous. Artificial intelligence brings new perspectives as it can analyze vast sets of biological, genetic, and molecular data to identify new potential targets for drug development. This accelerates the drug research process, which can take years.
AI systems can automatically generate candidate molecules for new drugs, taking into account chemical properties, bioactivity, and ADME properties (Absorption, Distribution, Metabolism, Excretion). This speeds up the drug discovery process.
The use of AI can predict the potential pharmacological activity of chemical compounds based on their molecular structure, allowing for quicker identification of the most promising candidates. For example, companies like Iktos or Aqemia offer concrete solutions to assist R&D teams. They collaborate with numerous pharmaceutical laboratories to help them accelerate and optimize drug design phases (Janssen, Merck, Servier, etc.).
Another interesting aspect of artificial intelligence is the possibility of developing personalized treatment plans based on individual genetic and medical characteristics of patients, thus maximizing efficacy and minimizing side effects.
Finally, by automating certain steps in the drug design process, AI can reduce drug development costs and accelerate the time needed to bring a drug to market.
Rise of In Silico and Digital Twins
For many years, research and development around new molecules have primarily relied on experiments conducted in the laboratory ("in vitro") and then on living beings ("in vivo") through clinical trials.
Artificial intelligence offers new possibilities with the deployment of "in silico." This refers to the use of simulations and computer models to, for example, identify and design new drugs, study molecular interactions, simulate clinical trials, or analyze huge sets of medical data, including medical imaging data, genomic sequencing, and electronic medical records, to identify trends and correlations that might elude manual analysis.
An increasing use of digital twins is observed in public or private research teams to work on these simulations. These are computer models incorporating artificial intelligence that accurately reproduce the characteristics and behaviors of a specific patient or individual. Researchers use digital twins to test scenarios in a safe and cost-effective manner before trying them in real environments or situations (3).
Various applications include:
Optimization of Clinical Trials
In clinical trials, artificial intelligence is used at various levels: the design of clinical trials, the identification of suitable patients, and real-time monitoring.
For identifying eligible patients, AI can quickly pinpoint patients suitable for a clinical trial by analyzing large amounts of medical data, such as electronic medical records, medical imaging, and genomic data. This also simplifies the selection and recruitment of candidates.
Research teams use AI to design more efficient clinical trial protocols by optimizing randomization, sample size, trial duration, etc. This brings various benefits such as cost reduction and acceleration of the process.
During clinical trials, AI solutions facilitate the analysis of collected data to predict and understand outcomes. For example, electronic data capture (EDC) is used, a technology that streamlines data collection while minimizing human errors and ensuring data security (4).
It then becomes possible to monitor data in real time, detect abnormal trends, and identify potential issues, which can lead to quicker adjustments in the protocol or the discontinuation of the trial if necessary. Potential side effects of the tested drugs can also be identified more quickly by analyzing trial participant data.
But beyond this optimization, artificial intelligence could eventually completely replace certain clinical trials. A recent example is an AI developed by Novasdiscovery that performed as well as a 3-year clinical trial: "On September 11, the pharmaceutical group AstraZeneca published the results of its clinical trial on a treatment for lung cancer. Three days before, thanks to artificial intelligence, a prediction of the results of this trial was published by the deputy director of the Institute of Cancerology at the Hospices Civils de Lyon. A prediction that proved accurate" (5).
Challenges to Overcome
In this digital transformation of health and research sector, artificial intelligence raises numerous challenges and issues:
The new regulation on the use of artificial intelligence in the European Union, named the AI Act, which will come into effect in January 2024, will provide a new regulatory framework for the marketing of solutions with a focus on safety, health, and fundamental rights.
Today, research, whether fundamental or pharmaceutical, is the field in healthcare that has most integrated artificial intelligence into its operations. This technology brings numerous benefits in clinical phases, the discovery of new treatments, and the optimization of clinical trials. It also holds promise in the fight against cancer and rare diseases. However, as we have seen, there are still challenges and issues to be addressed for optimal and ethical use of this technology.
It's fast, very fast, perhaps too fast for some?
At Digital Pharma Lab, we are enthusiastic revolutionAIres, and we love speed!
The current transformation is explained by the convergence of several factors: the mastery of technology concurrent with a wide global use of connected devices, thus promoting the exponential growth of data collection (+ 40% annually). This is how traditional programs are disappearing in favor of much more powerful AI programs using these vast volumes of data.
Of course, we must collectively ensure an ethical and secure data collection, and regulation must adapt in real time to avoid being overtaken…
So, this is just the beginning. In the face of fears and controversies that this digital revolution arouses, let's be confident that technology will rhyme with hope (for new treatments) and that Research will give AI its rightful place of honor!
Rémy Teston
Digital Consultant / E-Health Expert – Buzz E-santé
.png)
.png)
.jpeg)
