From accelerating, drug development to automating literature reviews and prioritizing patients at-risk – new advances in natural language processing (NLP) based on deep learning and transfer learning have made a whole set of software use cases viable. The Healthcare NLP Summit – a free online conference on April 6th and 7th – brings together 30+ technical sessions from across the community that works to apply these advances in the real world. Here are some of the promising applications for which current state-of-the-art results will be discussed.

Accelerating drug discovery

Traditional drug discovery processes take one to two decades. The recent initial success with the COVID-19 vaccines raises the bar – why aren’t we moving as fast to deliver cures for cancer, Alzheimer, kidney failure, and heart disease? NLP can accelerate many aspects of the drug discovery process – from patient recruitment to preparing regulatory documents for submission.

Finding & Coding Patient Risk Levels

Amidst the COVID-19 pandemic, under-reporting of cases is a concern for healthcare agencies. This is a broad issue – undiagnosed diabetes, depression, or flu also result in needless suffering. NLP algorithms can help healthcare leaders stay ahead of the pandemic by predicting patient risks, predicting outcomes, and anticipating future disease hotspots worldwide. Similar techniques are used to reduce administrative burden by semi-automating medical coding, risk adjustment, and pricing for accountable care organizations.

Answering Medical Questions

NLP can improve both the patient and provider experience by reducing transcription costs and delays, improving health records’ accuracy, and improving access to information. Automated speech-to-text and transcription systems are helping clinicians increase the quality of treatment while also protecting physicians from excessive burnout. Providers attribute that NLP algorithms help them relax and spend more time with their patients during appointments.

Summarizing medical records

Summarizing long medical charts accurately into a structured and normalized list of critical data points is applicable in a broad set of use cases, from building real-world datasets to clinical trial cohorts. Historically – and in most cases still today – manually reviewing and processing stacks of chart notes from clinical records could take weeks or months. NLP aims to execute this process within seconds, which frees up clinicians’ time to concentrate on more complex situations while also reducing the time & cost spent on administrative tasks.

 
Learn more about how state-of-the-art NLP is applied in the real world at the Healthcare NLP Summit by John Snow Labs on April 6-7. This virtual event features 30+ talks and 5 days of training focused on NLP best practices, real-world case studies, and challenges in applying deep learning and transfer learning in practice – and the latest open source libraries, models and transformers you can use today. Register now to hear from teams who are leading the Healthcare NLP space at Google, Roche, Microsoft, 3M, Stanford, IBM, Merck, Berkeley, Cigna, Amazon, IQVIA, John Snow Labs, and many more!