Computing patient similarity based on unstructured clinical notes
By: Petr Zelina , Marko Řeháček , Jana Halámková and more
Potential Business Impact:
Finds similar patients for better cancer treatment.
Clinical notes hold rich yet unstructured details about diagnoses, treatments, and outcomes that are vital to precision medicine but hard to exploit at scale. We introduce a method that represents each patient as a matrix built from aggregated embeddings of all their notes, enabling robust patient similarity computation based on their latent low-rank representations. Using clinical notes of 4,267 Czech breast-cancer patients and expert similarity labels from Masaryk Memorial Cancer Institute, we evaluate several matrix-based similarity measures and analyze their strengths and limitations across different similarity facets, such as clinical history, treatment, and adverse events. The results demonstrate the usefulness of the presented method for downstream tasks, such as personalized therapy recommendations or toxicity warnings.
Similar Papers
Patient Trajectory Prediction: Integrating Clinical Notes with Transformers
Machine Learning (CS)
Helps doctors predict sickness better using notes.
Bridging Electronic Health Records and Clinical Texts: Contrastive Learning for Enhanced Clinical Tasks
Computation and Language
Helps doctors predict patient return to hospital.
Classification of Psychiatry Clinical Notes by Diagnosis: A Deep Learning and Machine Learning Approach
Machine Learning (CS)
Helps computers find mental health problems in notes.