“Medical-grade” AI is poised to help health systems manage their data

“Medical-grade” AI is poised to help health systems manage their data

There has been a lot of hype around generative AI tools like ChatGPT over the past year or so. But for all the emerging healthcare use cases that involve large language models, it’s worth noting that one of the main values ​​of LLM is its ability to improve on natural language processing capabilities that have been around for decades.

Healthcare organizations need AI systems that can ingest and understand a patient’s entire chart, or the entire organization’s EHR system — and genomic AI tools are well-positioned to help.

We interviewed Kim Perry, Chief Growth Officer at emtelligent, a medical NLP company, who spoke about the power of… Generative AI, large language models and their relationship to NLP, AI systems that can understand an entire chart or entire EHR system, and how “clinical-grade AI” can improve patient outcomes and prevent physician burnout.

Question: There’s something of a misunderstanding about the power of generative AI and large language models — they actually support natural language processing, which is another form of AI. Please talk about this relationship.

a. It is amazing that less than a year ago, ChatGPT and Generative AI were not known to the general public. ChatGPT now has more than 100 million users around the world, including consumers and entire industries. Unfortunately, many people form opinions about emerging technologies based on first impressions and casual reading. But genetic AI and big language models are new topics for many consumers and businesses, so some confusion is inevitable.

Natural language processing Generative AI predates recent developments in LLMs. However, we are now seeing genAI and LLMs being used to create what we call “medical-grade AI” or medical-grade NLP. LLM students are specifically trained on clinical data, and unlike traditional NLP, medical-grade AI is able to unlock 80% of medical data currently hidden in unstructured text.

The ability of medical AI to process and understand millions of documents will change the way doctors do their jobs at the point of care. But their value will extend across the healthcare continuum to benefit not only providers, but also payers, pharmaceutical companies, life sciences, and academic researchers.

Q: AI like ChatGPT is great for answering one question, but you suggest that service organizations need systems that can understand an entire schema or EHR system. Please detail.

a. GenAI is great for answering a single question, but that’s assuming the answer is actually correct. Often times, apps like ChatGPT make up facts, or “hallucinate.” This is simply unacceptable in a situation where clinicians must make evidence-based decisions at the point of care.

If providers cannot trust the source of the information, they will stop using it. Furthermore, genAI can bombard users with a barrage of keyword-based data, making it difficult for doctors to find the information they need.

Providers need tools that can access, understand, and contextualize patient information from a single chart or across an electronic health record system. This is where traditional NLP fails; It can’t make sense of all that unstructured data.

Medical-grade artificial intelligence benefits from advances in Machine learning and the LLMs I mentioned earlier allow provider organizations to unlock the value of unstructured data such as free text notes. For example, medical-grade AI can understand medical terminology, including abbreviations and slang, that cannot be understood in traditional NLP.

Since medical-grade AI is designed for enterprises, it can process and understand millions of documents. This ability to scale is critical as the volume of health data continues to grow.

Q: How can what you call “medical-grade AI” improve patient outcomes?

a. Medical-grade AI can improve patient outcomes by providing doctors with the information they need, when they need it, and in an easily consumable form, thus enabling them to deliver more effective care.

When doctors are unable to access patient information in unstructured data — about a past procedure, for example, a chronic condition or a severe drug allergy — they lack a comprehensive view of the patient. This can lead to errors in diagnosing and treating patients leading to negative outcomes.

Conversely, when medical AI generates automated summaries of a patient’s history, doctors at the point of care have immediate access to information that enables them to gain insights into a patient’s health and well-being. This is especially important when a doctor sees a patient for the first time.

As I mentioned earlier, doctors won’t use tools they don’t trust because they can’t verify the information given to them. Medical AI overcomes this concern by linking the information in a patient’s summary to the original data in that patient’s chart, allowing doctors to review context and check sources for accuracy.

Q: How can Medical AI prevent physician burnout?

a. Doctors spend a lot of time searching and sifting through patient data. This creates a lot of pressure on providers because they want to interact with patients, not stare at a computer screen or search for specific information buried in a long patient chart.

Medical-grade AI provides powerful capabilities like context-sensitive search and automated summaries that dramatically improve workflows. By striking the right balance between recall and precision, medical AI enables doctors to be more efficient and effective in treating patients.

Improving workflow helps Reduced burnout because doctors do not feel as if they are constantly struggling to keep up with patient workloads. It stands to reason that when doctors have smarter digital tools in their offices and on the exam table, they will be less frustrated and more able to practice at the top of their license. This is what all doctors want to do.

Follow Bill’s HIT coverage on LinkedIn: Bill Siwicki
Email him: bsiwicki@himss.org
Healthcare IT News is a HIMSS media publication.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *