Clinical research is a slow process by its very nature. Several regulations have been implemented to ensure that there is no harm done with potential new products and that they indeed provide therapeutic value to patients. This need has made research time consuming, expensive and extremely complicated, especially considering developments such as biologic treatments and genetic medicine.
Yet, in the past decade or so, the world has witnessed a tsunami that has revolutionized every segment of the economy: Artificial Intelligence (AI). From being science fiction lore to becoming an aide in sectors as different as recruitment, farming, marketing or automotive, every day the frontiers of AI are pushed to new limits. Today, the healthcare industry is in the process of implementing or experimenting with applications such as data mining, neural networks, and machine learning, always aiming at improving peoples’ lives.
According to CB Insights, in the past 5 years, investment in AI applications for healthcare has topped $4.3bn USD. Clinical research is a well-suited industry for AI disruption for several factors; first, because it is data driven, with the core product being precisely data, which in turn allows computers to identify patterns (data mining) and improve algorithms (deep learning). So far, the industry has produced very interesting results, with data mining aiding in the identification and prioritization of targets for clinical development and deep learning helping to create image diagnostic tools or helping configure and operate conversational bots for support in mental healthcare.
Yet, while AI certainly represents an extraordinary opportunity to fasten clinical research, it comes with caveats. The most relevant ones being the privacy of information, accuracy, and reliability. As for the former, big data and artificial intelligence are new domains of knowledge, and researchers, companies, and regulators are still debating on their ethical implications, particularly who owns the information, to what extent can it be used and if and how to ensure consent is properly given to commercially exploit it. As for the latter, it is important to recognize the current limits of technology. While the newest image diagnostic tools are as precise as humans, they are still accurate only on approximately 85% of cases. Similarly, several instances in which conversational bots were unable to offer adequate responses to input have been documented. In a field as delicate as mental health, it is crucial to have reliable tools to provide effective care.
Technology has certainly come a long way and the tools have improved tremendously within the last decade, and thus we can only theorize how far will they go. In the initial stages of development, innovation came from North America, however, today China has started to emerge as a center for AI due to the combination of several factors such as availability of data, costs and friendly governmental policies. In opposition, North American and European players face mounting costs and tighter regulations, particularly regarding privacy. Nonetheless, even if the speed of improvement remains constant, we are certain the landscape will be entirely different within the next 5 years. Some signs of this are already visible with early AI applications that do not aim to improve the provision of healthcare but aim at becoming therapies by themselves.
In future #KCR_Trends we will continue discussing how will AI change the game for clinical development.
Associate Director, PR & Marketing
mobile: (+48) 605 053 164