Exploring the Narratives of Patients With Cancer Using Large Language Models: Topic Modeling and Social Network Analysis
-
By
-
Xinyu Feng
-
Hin Chi Kwok
-
Ching Kok Chung
-
Janelle Yorke
-
Vivian Hui
-
July 6, 2026
-
Clinical Scorecard: Analyzing Cancer Patient Experiences Through Large Language Models: A Study of Topic Modeling and Social Network Dynamics
At a Glance
| Category | Detail |
|---|
| Condition | Cancer |
| Key Mechanisms | Natural language processing, topic modeling, psychosocial analysis |
| Target Population | Patients with cancer |
| Care Setting | Psycho-oncology |
Key Highlights
- Global cancer burden estimated at 20 million new cases in 2022.
- 69.9% of cancer patients survive at least 5 years post-diagnosis.
- Psychosocial stressors significantly impact cancer patients' quality of life.
- Traditional research methods are limited in capturing evolving patient experiences.
- TopicGPT offers a novel approach to analyze large-scale patient narratives.
Guideline-Based Recommendations
Diagnosis
- Utilize advanced natural language processing techniques for psychosocial assessments.
Management
- Incorporate patient narratives into psycho-oncology services for better support.
Monitoring & Follow-up
- Regularly analyze patient experiences to identify emerging psychosocial issues.
Risks
- Be aware of the psychosocial stressors affecting treatment adherence and quality of life.
Patient & Prescribing Data
Individuals diagnosed with cancer
Focus on holistic well-being alongside traditional oncological therapies.
Clinical Best Practices
- Employ scalable analysis methods to understand patient experiences.
- Integrate findings from patient narratives into clinical practice.
- Utilize network analysis to visualize psychosocial challenges.
Related Resources & Content
