NeuroBlu NLP: Curating data on symptomatology to provide a more complete picture of the patient journey

April 2, 2024

Real-world data (RWD) has the unique and powerful potential to provide insights about patient outcomes in routine, clinical settings and is inclusive of patients who are not typically represented in clinical trials. Traditional RWD sources such as claims do not include more granular clinical information, such as symptomatology or disease severity. This additional granular information is key to identifying subgroups of interest and assessing their clinical outcomes, especially in the context of novel therapeutic development.

Holmusk has developed a novel natural language processing (NLP) approach called NeuroBlu NLP that extracts key information from clinical notes and transforms it into structured, analyzable, and actionable data. Once these NLP models are applied to unstructured data, the resulting insights are made available in NeuroBlu, a powerful analytics tool built to explore the world’s richest real-world database for behavioral health.

Holmusk has developed a transformer architecture-based NLP model to capture core clinical features of patients with depressive disorders, namely anhedonia, suicidal ideation with plan or intent (SP), and suicidal ideation without plan or intent (SI) or where the plan or intent are unknown. Read the use case to learn about this unique model, its development, and ability to uncover previously inaccessible symptomatology insights with up to 99% accuracy.


Interested in learning more? Contact us to schedule a demo.

Back to top
Contact us