• Strategy
  • Payers
  • AI
  • Cybersecurity
  • Device Development
  • Studies/Published Research
  • Ethics
  • Regulatory
  • Digital Biomarkers
  • Data Science
  • Companion Diagnostics
  • Clinical Care
  • Digital Therapeutics
  • Academic Research
  • Virtual Reality
  • RWE
  • Digital Endpoints
  • Medical Devices (Class II and Class III)
  • Commercialization
  • Clinical Trials
  • Drug-Device Combinations
  • SaMD
  • Remote Patient Monitoring
  • Patient Engagement
  • Profiles/Q&A

Generating Pharma’s AI: Q&A With Masanori Ito, Senior Director, Head of Enterprise Insights and Digital Solutions, Digital, Analytics, and Technology at Astellas

Feature
Article

Ito discusses how data-driven AI can be utilized by the pharma industry.

Masanori Ito

Masanori Ito
Senior Director, Head of Enterprise
Insights and Digital Solutions,
Digital, Analytics, and Technology
Astellas

Pharmaceutical Executive: How can data-driven AI be utilized by the pharma industry?

Masanori Ito: Today, companies use data and analytics to help address business challenges, measure success, and conduct trend forecasting at the industry level, all of which can help improve the overall health of the business. When thinking about the pharma industry adopting AI, we have additional challenges to navigate, given the highly uncertain nature of the industry and the significant financial investments involved to launch and sustain projects. However, with careful consideration, pharmaceutical companies can greatly benefit from data-driven AI and its ability to revolutionize the industry.

At Astellas, we utilize data and analytics to help inform strategic decision-making processes. Specifically, our teams use data and analytics as a tool to help choose which business portfolios to invest in, and the right time for doing so. Data-driven methods have helped us make objective and accurate decisions for portfolio profiling and determine optimal business models. We have also used data and analytics to help optimize and accelerate business development by using objective data to evaluate the strengths and opportunities in partnerships.

For business models concentrated in specific disease areas, data-driven methods enable leaders to predict insights in drug discovery, design clinical trials, and forecast development timelines using historical R&D data and real-world data/evidence. Additionally, this can help estimate the speed of market penetration and predict sales, by utilizing market data. All of the above can help inform strategic business decisions, and I believe this is truly just the beginning when it comes to AI opportunities in pharma.

PE: What regulatory concerns do companies have regarding the use of data-driven AI?

Ito: When new technologies rapidly grow and evolve, it can be challenging for regulatory protections to match the pace. Current regulatory concerns in AI, such as data privacy, adds an extra layer of challenges when pharmaceutical companies are utilizing data-driven AI. The GxP (Good Manufacturing Practice, Good Clinical Practice, etc.) perspective can also impact AI’s use, as it requires a very high level of accuracy and background explanation. Additionally, because the use of AI requires human oversight, and since users may deal with interpretation issues, questions are raised on whether productivity in the workplace is increased significantly when data and analytics are implemented.

PE: How is Astellas implementing data and analytics into its business?

Ito: At Astellas, we apply data and analytics in various ways to help inform different strategic business and R&D decisions, especially those focused on transformative therapies and new technologies. Not only that, the scope of our use of data analytics extends to all areas of our value creation, including supply chain management and insight into innovative organizational structures.

We are using generative AI, one of the most well-known technologies among the data-driven AI, to supplement individuals’ creativity and assist in finding smart solutions. For example, generative AI can summarize large amounts of texts, and we are using this tool to evaluate large quantities of publicly available documents (such as corporate reports) to maintain a competitive, intelligent perspective of industry issues. This, in turn, makes our decision-making process inherently more data driven.

Another example is the development of our hypothesis-oriented simulation, which can bolster strategic decision making by creating sample scenarios and deductive guesses that mimic real-world processes. Any past data can be utilized appropriately for predictions as far as the structure of the business problem does not change. However, the internal and external environment of the pharmaceutical business is changing rapidly nowadays, and it is often difficult to predict the future and create value based on past data. This is where deductive, hypothesis-oriented simulation comes into play. We are constantly evaluating and updating our simulation hypotheses, to improve the quality of these measures and our overall strategic thinking and planning.

We use both data-driven and hypothesis-oriented simulations, depending on the challenge we are tackling, and we believe this gives us a competitive advantage.

PE: What is the future of data and data-driven AI in the pharma industry?

Ito: Currently, the utilization of AI, machine learning, and large language models are peaking in terms of inflated expectations. As the industry learns more about what these technologies can and cannot accomplish, we will see a widespread implementation of practical applications. As data-driven AI becomes more widely used and a variety of tasks are performed at high speed and high quality, decisions about what challenges to address and what actions to take will become increasingly important. In addition to data-driven AI, hypothesis-oriented simulation will be needed to support these decisions by mitigating the future risks and uncertainties. Outside of research, business leaders will likely expand their use of such methods to help inform business decisions, as well as predicting and managing uncertainties in the future, from the mid-to-long-term.

Related Videos
© 2024 MJH Life Sciences

All rights reserved.