In our latest episode of the Learning Health System podcast, we explored the role of Natural Language Processing (NLP) in healthcare. We discussed how NLP, particularly through the ElasTex platform, is transforming how clinical notes are utilized for research, analytics, and real-time decision-making.
Dr. Mirhaji emphasized that storytelling is central to medicine. Clinicians document their interactions with patients in narratives, capturing rich, detailed accounts that structured datasets often lack. Millions of these clinical notes are created daily, containing valuable insights that can inform care and research. However, traditional data querying methods do not intuitively accommodate these unstructured narratives. This is where NLP comes in—enabling healthcare systems to extract, index, and analyze these notes efficiently and meaningfully
While NLP is not new to healthcare, Dr. Mirhaji pointed out that existing tools often fall short of integrating seamlessly into a Learning Health System. The ElasTex platform was designed to overcome these challenges by providing scalable, real-time, and fully integrated text processing capabilities. It enables healthcare organizations to index and catalog the vast variety of notes generated across disciplines, from clinical summaries to operational records.
One of the standout features of ElasTex is its ability to unify NLP with broader analytics. Instead of treating text analysis as a standalone capability, ElasTex embeds it into the overall analytical workflow, ensuring that insights from clinical notes are readily available for research, predictive modeling, and real-time clinical decision support.
ElasTex is already making a tangible impact in areas such as cohort-based studies and clinical trial matching. Dr. Mirhaji explained how researchers can leverage NLP to identify and track patient histories, treatment responses, and disease progression. For example, tumor staging details, which are often buried in unstructured notes, can now be indexed and queried efficiently—helping match patients to appropriate clinical trials faster and more accurately.
With the rise of generative AI and large language models (LLMs), a natural question arises: Will these advancements make traditional NLP obsolete? Dr. Mirhaji believes that while LLMs offer powerful capabilities, they are computationally expensive and require careful implementation to ensure efficiency, privacy, and scalability. ElasTex takes a hybrid approach, integrating traditional NLP with fine-tuned LLMs to balance accuracy, cost, and real-time application needs.
Healthcare data security remains a critical concern. Clinical notes contain sensitive patient information, making robust privacy measures essential. ElasTex employs sophisticated de-identification techniques using NLP and LLMs to redact personal identifiers while preserving the utility of the data. Additionally, stringent access controls and audit logs ensure compliance with HIPAA and other regulatory frameworks.
As Montefiore and other institutions continue to evolve into fully functional Learning Health Systems, NLP-powered platforms like ElasTex are paving the way. By enabling real-time data integration, cross-disciplinary collaboration, and scalable analytics, NLP is helping bridge the gap between unstructured narratives and actionable insights.
Dr. Mirhaji closed the conversation by highlighting the need for continuous innovation. Future episodes will delve into how large language models and AI-driven analytics are shaping the next phase of healthcare transformation. Stay tuned!