Automatic identification of brain metastasis development in patients with lung cancer using natural language processing
Poster Number: P50
Presentation Time: 05:00 PM - 06:30 PM
Abstract Keywords: Large Language Models (LLMs), Natural Language Processing, Real-World Evidence Generation, Population Health
Primary Track: Foundations
Programmatic Theme: Clinical Informatics
Brain metastases (BM) lead to increased mortality. However, lack of data in cancer registries on BM limit real-world studies that could inform management. We developed a BERT-based NLP model to accurately detect BM in MRI reports of cancer patients. Aggregating report-level predictions, the model accurately identified patients with BM in a validation cohort and also revealed additional cases not recorded in the registry, demonstrating its potential in enhancing BM detection and facilitating epidemiological research.
Speaker(s):
Julie Wu, MD PhD
Palo Alto Veterans Affairs Healthcare System
Julie Wu
Author(s):
Aparajita Khan, PhD - Stanford University; Julie Wu, MD PhD - Veterans Affairs Health System Palo Alto; Chloe Su, PhD - Stanford University; June Corrigan, BS - Boston University School of Public Health; Rika Terashima, MD - Stanford University; Megan Chang, BS - Stanford University; Emily Rodriguez, BS - Stanford University; Christopher J. Shin, BS - Stanford University; Akash Shah, BS - Stanford University; Rakshit Kaushik, BS - Stanford University; Allison Kurian, MD - Stanford University; Heather Wakelee, MD - Stanford University; Curtis Langlotz, MD, PhD - Stanford University; Leah Backhus, MD - Stanford University; Michael Kelley, MD - VA Health System, Durham; Nathanael Fillmore, PhD - VA Boston Healthcare System; Summer Han, PhD - Stanford University;
Poster Number: P50
Presentation Time: 05:00 PM - 06:30 PM
Abstract Keywords: Large Language Models (LLMs), Natural Language Processing, Real-World Evidence Generation, Population Health
Primary Track: Foundations
Programmatic Theme: Clinical Informatics
Brain metastases (BM) lead to increased mortality. However, lack of data in cancer registries on BM limit real-world studies that could inform management. We developed a BERT-based NLP model to accurately detect BM in MRI reports of cancer patients. Aggregating report-level predictions, the model accurately identified patients with BM in a validation cohort and also revealed additional cases not recorded in the registry, demonstrating its potential in enhancing BM detection and facilitating epidemiological research.
Speaker(s):
Julie Wu, MD PhD
Palo Alto Veterans Affairs Healthcare System
Julie Wu
Author(s):
Aparajita Khan, PhD - Stanford University; Julie Wu, MD PhD - Veterans Affairs Health System Palo Alto; Chloe Su, PhD - Stanford University; June Corrigan, BS - Boston University School of Public Health; Rika Terashima, MD - Stanford University; Megan Chang, BS - Stanford University; Emily Rodriguez, BS - Stanford University; Christopher J. Shin, BS - Stanford University; Akash Shah, BS - Stanford University; Rakshit Kaushik, BS - Stanford University; Allison Kurian, MD - Stanford University; Heather Wakelee, MD - Stanford University; Curtis Langlotz, MD, PhD - Stanford University; Leah Backhus, MD - Stanford University; Michael Kelley, MD - VA Health System, Durham; Nathanael Fillmore, PhD - VA Boston Healthcare System; Summer Han, PhD - Stanford University;
Automatic identification of brain metastasis development in patients with lung cancer using natural language processing
Category
Poster - Regular
Description
Date: Tuesday (11/12)
Time: 05:00 PM to 06:30 PM
Room: Grand Ballroom (Posters)
Time: 05:00 PM to 06:30 PM
Room: Grand Ballroom (Posters)