From entertainment to healthcare, ecommerce to finance – it’s undeniable that the way we do business changes when artificial intelligence is involved. AI is the buzzword du jour – seated currently at the top of Gartner’s well-known hype cycle.
In the life sciences sphere, the promise is there, but the widespread integration of AI technology has yet to be realized. This idea was supported by a survey released by Pistoia Alliance this year, which highlights both the adoption of these technologies and the challenges associated with getting the most out of them.
The survey of more 350 life science professionals found that nearly half (44%) are experimenting with AI-based solutions.
Here’s a look at the emerging AI solutions in the life science industry described in the survey:
Natural Language Processing (NLP): Life sciences organizations can collect, annotate, and integrate unstructured text from large data sources, such a biomedical literature.
At CCC, we see life sciences organizations using text analytics and natural language processing to get actionable insight from real world data – and as a result they’ve found valuable intelligence that can inform commercial business strategies.
Here is a look at two use cases from our partners at Linguamatics, where text mining has transformed real world data to real world evidence.
Machine Learning (ML) In the life sciences setting, machine learning can be used to quickly and accurately identify disease phenotypes, to learn and predict from structured biological data and image-based data, and to improve patient safety and drug development.
But – a lack of access to full-text scientific literature vs. abstracts, inconsistent terms of use, and formatting discrepancies can lead to machine learning difficulties, as CCC’s Doug Knight points out here.
2 Milestones before AI reaches breakout velocity
AI is already being integrated into the lifecycle of pharmaceutical and healthcare companies in the areas of drug target discovery and clinical diagnosis, as well as the discovery of biomarkers and drug targets. Still, the survey outlined limiting factors before AI reaches breakout velocity in the life science industry.
The two primary limitations are the lack of in-house expertise and lack of access to quality data to develop domain-specific AI solutions. Because of this, we’re seeing more and more companies focus on building partnerships with organizations that have expertise in AI.
Lack of technical expertise is the most cited barrier for AI (30%) and for ML/NLP (28%)
There are several open-source platforms for developing AI solutions, but the ability to tailor in-house solutions to emerging biological data sources such as single-cell sequencing, proteomics, metabolomics, or 3D spatial/image-based biological data is more complex. Integrating this information with existing knowledge will be a necessary precursor to maximize the benefits of AI.
Lack of access to data (24%) and data quality (26%) were the next biggest barriers to using AI.
Many researchers cite not having access to quality data as a barrier to developing AI solutions. We know that intelligent data and information integration is what drives innovation, and business and scientific users need to search across, access, and analyze internal and external content and data. That’s difficult to do when data and content metadata often lack standardization, limiting the ability to process and analyze them.
Ultimately, we should look forward to cross-disciplinary partnerships that can bring together experts from complementary specialties to ask novel questions and deliver insights that directly impact human health.