Paper Type

Short

Paper Number

1478

Description

The digital transformation within the healthcare industry not only marks a significant transition towards enhancing patient care and opens new opportunities for health research. Recognizing the unique and important role clinician notes play in enhancing treatment processes and outcomes, we introduce a method for constructing patient representations through semi-supervised topic modelling. This approach effectively addresses the challenges related to unlabeled data and the lack of ground truth. We compare our method with traditional topic modeling and show our method more accurately predicts patient discharge decisions. This method not only provides a novel application of natural language processing for improved clinical decision-making but also offers a new approach to constructing research constructs from clinician notes.

Comments

Healthcare

Download

Share

COinS
 
Jul 2nd, 12:00 AM

Constructing Patient Representation through Semi-Supervised Topic Modeling

The digital transformation within the healthcare industry not only marks a significant transition towards enhancing patient care and opens new opportunities for health research. Recognizing the unique and important role clinician notes play in enhancing treatment processes and outcomes, we introduce a method for constructing patient representations through semi-supervised topic modelling. This approach effectively addresses the challenges related to unlabeled data and the lack of ground truth. We compare our method with traditional topic modeling and show our method more accurately predicts patient discharge decisions. This method not only provides a novel application of natural language processing for improved clinical decision-making but also offers a new approach to constructing research constructs from clinician notes.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.