In this Q&A, she tells us about future trends and opportunities.
Contextual data integration involves merging diverse data sets while preserving and utilizing the context surrounding that data. Could you elaborate on why this approach is important in personalized medicine? How does it overcome traditional data silos in healthcare and positively impact the end user?
Imagine trying to piece together a puzzle with half the pieces missing—that’s what happens when data is siloed. Without integrated, centralized data across sources, data decisions are made using incomplete, fragmented datasets. Contextual data integration keeps every critical piece of the puzzle intact, bringing data together to form a complete picture.
In personalized medicine, which leverages data from devices, sensors and wearables, omics data, imaging, electronic health records (EHR), and other sources, more data equals more complexity. The more data we bring in, the more advanced statistical methods and robust data management systems and techniques are needed to process and interrogate that data while also maintaining data integrity and compliance with regulatory standards.
Overcoming traditional data silos with modern, tech-driven methods for contextual data integration not only achieves better data quality, but is critical for enabling biopharma companies to process and extract insights from the big data of precision medicine. Only then can these novel data types be translated into personalized medicines that benefit patients.
What are the biggest challenges faced when integrating diverse datasets, and how are you solving them?
There are many challenges today when integrating diverse datasets, from dealing with varied data formats and sources, to managing the sheer volume of data being collected in modern clinical trials. In addition to data from traditional sources like electronic data collection (EDC), there is now a surge of data from wearables, images and labs, and genomic biomarker data, to name a few. Some of this is driven by technology advancements within data collection, making it easier to automate collection of varied, continuous data streams, as in the case of wearables and sensors. Another fueling factor is recent and rapid scientific advancement, such as genome sequencing and innovations in imaging data.
In the case of precision medicine, which requires high volumes of patient data to personalize outcomes, all of these factors are coming together to create an incredible pace of rising data volume and variety. Where data was once primarily coming from EDC, we’re now seeing many other sources of data acquisition, all of which must be harmonized to gain usable insights for decision making. As data volume has increased, processing these high-volume datasets and managing data flows has become much more complicated, because with more data, systems and sources, there are more stakeholders and decision-makers to manage. And even with all of these data challenges, there is added pressure on biopharma to perform data processes faster and more efficiently, often with fewer resources.
To begin to solve these problems, leveraging a foundational data infrastructure that incorporates artificial intelligence (AI) and machine learning (ML) can automate time-consuming tasks for data ingestion, integration and standardization. In data management and data review, AI/ML can be used to detect anomalies in the data, making it easier for the human in the loop to leverage their expertise where it’s needed most.
How can AI-driven data integration and review empower more targeted therapy development?
AI-driven data integration enables teams to quickly aggregate and analyze diverse datasets—including clinical data, genomics, and real-world data—to uncover patterns and correlations. This is increasingly important in the development of targeted therapies, where clinical teams need holistic insight into the data. This enables teams to identify the right patient populations for specific therapies, optimize trial designs, and predict outcomes more accurately.
AI also significantly reduces the burden on people. By automating manual or routine tasks within data standardization, cleaning, review and analysis, AI frees clinical and data teams to focus their domain expertise on critical decision-making instead of non-critical data processing activities. For example, AI can support automated mapping of data to standards, surface outliers in data analytics and visualizations, and detect data discrepancies for human review. This kind of AI-supported data management is as much a technical upgrade as it is a way to reduce fatigue and optimize resources. The amount of data and the vital role of data precision medicine isn’t slowing down; it’s only increasing. Adopting AI to scale data processes in line with rising data collection will result in a more efficient, quality data pipeline and the ability to generate insights faster, for a higher success rate in this area of drug development.
What types of data are most critical in an anticipatory approach to patient care, and how do you ensure they are effectively utilized?
For an anticipatory approach to patient care, the most critical data types include genomic data, biomarkers, patient-reported outcomes, and real-world evidence from digital health tools like wearables and remote monitoring devices. These data types provide a multi-dimensional view of a patient’s health trajectory, which allows for earlier detection of risks and proactive management of potential issues.
However, to effectively and accurately utilize these data types, advanced data platforms need to support real-time data integration and analysis so that teams can continuously process this data, identify trends, and generate actionable insights. This grants stakeholders the ability to make informed decisions in real time, anticipate and address risks or issues, and accurately leverage data for improved patient outcomes.
What advancements in data solutions do you think will most significantly impact personalized medicine in the next 5-10 years?
We’re on the brink of a shift where the value and potential of data will be increasingly unlocked by the next phase of tech advancement, particularly AI/ML driven opportunities. The next big leap in personalized medicine will rely on intelligent data solutions that do more than store and analyze information. One of the most significant changes will be the development of “living datasets” that continuously update in real time, pulling in new information from multiple sources like patient devices, lab results, and clinical trials without any lag. Datasets will not be static snapshots but evolving entities that adapt instantly to new findings, reshaping how we approach everything, from diagnostics to drug development.
We’ll also continue to see AI and machine learning handle routine, non-critical data tasks, allowing human experts to focus on high-impact data-driven decision-making. Data solutions that provide real-time data integration and analysis will power opportunities across the data ecosystem by providing a foundation of holistic, quality data for advanced applications, from accuracy of patient selection to optimization of clinical trial designs. Advancements in data modeling techniques will also help simulate patient populations, helping biopharma companies better predict treatment outcomes and tailor therapies. By leveraging these data models, researchers can accelerate clinical development and improve the precision of personalized medicine, at scale.