– Machine learning has quickly become of the most pervasive trends in the healthcare IT industry, opening up endless opportunities for vendors, providers, and patients alike.
From personalized clinical decision support for cancer and chronic diseases to virtual assistants that engage patients and streamline provider workflows, the idea that true artificial intelligence could one day overhaul the entire delivery system seems tantalizingly close.
Reaching that goal, however, will require developers to slowly unlock a long series of machine learning accomplishments in a number of highly targeted areas, building upon increasingly sophisticated algorithms until, perhaps, their actions are one day indistinguishable from those of humans.
One of the foundations of this complex task is natural language processing. The ability to programmatically extract meaning from text and speech – and in some cases, return naturalistic responses to the user – is a core element required to understand human input, squeeze relevant information out of huge volumes of data, and bridge the gap between man and machine.
In healthcare, the rapidly rising interest in machine learning and artificial intelligence is creating a lucrative market for products and services that can tackle these issues to support precision medicine, decision support, computer assisted coding (CAC), consumer engagement tools, and smarter electronic health records.
Globally, the healthcare-specific NLP marketplace is predicted to be worth $4.3 billion by 2024 – more than a four-fold increase since 2015.
The anticipated 18.8 percent compound annual growth rate (CAGR) will come from advances in text mining, voice processing, and machine translation improvements, Transparency Market Research (TMR) indicates.
The NLP market represents a significant portion of the overall healthcare artificial intelligence segment, which is likely to be valued at $22.7 billion by 2023, says Research and Markets in a separate assessment, representing a 48 percent CAGR over the next five years.
“Used as a part of artificial intelligence systems, applications of NLP technologies are being deployed for predictive analysis and clinical decision support systems,” TMR said in a previous report.
“The global healthcare natural language processing market is expected to receive an impetus from the uptake of these technologies by several companies for extracting knowledge from several clinic documents via machine learning or deep learning applications. The growing volume of unstructured clinical data and the unstoppable penetration of EHR systems are expected to fuel the growth of this market in the coming years.”
The natural language processing market is currently dominated by three major players who hold more than 40 percent of the sector: IBM, Apple, and Microsoft, notes TMR.
Massive investments by these companies have positioned them as early leaders in cognitive computing and machine learning, but they will need to work hard to hold on to their territory as startups with deep healthcare expertise enter the competition and EHR ecosystem vendors form partnerships to help improve their user experiences.
The complexity of clinical language and providers’ propensity for acronyms and abbreviations calls for natural language processing tools that are highly sensitive to disambiguation and understanding context.
Large volumes of unstructured data generated within electronic health records or held in photocopied reports, PDFs, and other static representations, can present an additional challenge to NLP tools.
Overcoming these hurdles will be important for the ultimate success of machine learning and artificial intelligence technologies.
Without reliable, complete, and accurate data extracted at a pace and volume that far exceeds the capacity of humans, AI and machine learning tools will not be able to function to a high degree of usefulness.
As NLP tools become more sophisticated, however, they will support the continued growth of decision support, CAC, and provider engagement options that aim to reduce documentation burdens, improve the personalization of care, enhance patient safety, and bring a more collaborative approach to quality care.