⚡ Quick Summary
This study introduces KSrisk-GPT, a model utilizing a large language model (LLM) to identify potential kidney stone risk factors from user experiences shared on social media. The model achieved an impressive 95.9% accuracy in detecting risk factors, highlighting the potential of social media data in understanding health risks.
🔍 Key Details
- 📊 Dataset: 11,819 user comments from Zhihu over the past 5 years
- 🧩 Categories: Six common risk factors for kidney stones
- ⚙️ Technology: GPT-4.0 with least-to-most prompting
- 🏆 Performance: Accuracy 95.9%, Precision 95.6%, Recall 96.2%, F1-score 95.9%
🔑 Key Takeaways
- 💧 Insufficient water intake and high-protein diets are significant risk factors for kidney stones.
- 🧬 Genetic factors also play a role in susceptibility to kidney stones.
- 🍽️ Dietary habits such as high calcium intake were frequently mentioned by users.
- 🔍 New potential risk factors identified include excessive use of vitamin C and calcium supplements.
- 💡 LLMs can effectively summarize and identify health risk factors from social media data.
- 🌐 Social media serves as a valuable resource for understanding patient experiences and disease prevention.
- 📈 The study demonstrates the efficacy of using AI in health-related research.
📚 Background
Kidney stones are a common urinary disease that can lead to severe health complications. Factors such as insufficient hydration and dietary choices significantly increase the risk of developing kidney stones. With the rise of social media, patients now have a platform to share their experiences, which can provide valuable insights into managing these risk factors.
🗒️ Study
The study focused on analyzing user comments from Zhihu, a popular Chinese social media platform, over the last five years. By employing the KSrisk-GPT model, researchers aimed to extract and categorize risk factors associated with kidney stones based on real patient experiences. The model utilized advanced prompting techniques to enhance its analytical capabilities.
📈 Results
The KSrisk-GPT model demonstrated remarkable performance, achieving a 95.9% accuracy in identifying comments that contained risk factors for kidney stones. The model also recorded a precision of 95.6% and a recall of 96.2%, indicating its reliability in extracting relevant information from user-generated content.
🌍 Impact and Implications
The findings from this study underscore the potential of leveraging social media data for health research. By utilizing large language models like KSrisk-GPT, researchers can gain deeper insights into patient experiences and identify new risk factors that may not be well-documented in traditional medical literature. This approach could pave the way for improved disease prevention strategies and enhance the quality of life for individuals at risk of kidney stones.
🔮 Conclusion
This study highlights the transformative potential of artificial intelligence in health research, particularly in understanding kidney stone risk factors through social media analysis. The successful application of the KSrisk-GPT model demonstrates how patient experiences can inform medical knowledge and improve health outcomes. Continued exploration in this area could lead to significant advancements in disease prevention and patient care.
💬 Your comments
What are your thoughts on using social media data for health research? We would love to hear your insights! 💬 Leave your comments below or connect with us on social media:
Identifying Kidney Stone Risk Factors Through Patient Experiences With a Large Language Model: Text Analysis and Empirical Study.
Abstract
BACKGROUND: Kidney stones, a prevalent urinary disease, pose significant health risks. Factors like insufficient water intake or a high-protein diet increase an individual’s susceptibility to the disease. Social media platforms can be a valuable avenue for users to share their experiences in managing these risk factors. Analyzing such patient-reported information can provide crucial insights into risk factors, potentially leading to improved quality of life for other patients.
OBJECTIVE: This study aims to develop a model KSrisk-GPT, based on a large language model (LLM) to identify potential kidney stone risk factors from web-based user experiences.
METHODS: This study collected data on the topic of kidney stones on Zhihu in the past 5 years and obtained 11,819 user comments. Experts organized the most common risk factors for kidney stones into six categories. Then, we use the least-to-most prompting in the chain-of-thought prompting to enable GPT-4.0 to think like an expert and ask GPT to identify risk factors from the comments. Metrics, including accuracy, precision, recall, and F1-score, were used to evaluate the performance of such a model.
RESULTS: Our proposed method outperforms other models in identifying comments containing risk factors with 95.9% accuracy and F1-score, with a precision of 95.6% and a recall of 96.2%. Out of the 863 comments identified with risk factors, our analysis showed the most mentioned risk factors for kidney stones in Zhihu user discussions, mainly including dietary habits (high protein, high calcium intake), insufficient water intake, genetic factors, and lifestyle. In addition, new potential risk factors were discovered with GPT, such as excessive use of supplements like vitamin C and calcium, laxatives, and hyperparathyroidism.
CONCLUSIONS: Comments from social media users offer a new data source for disease prevention and understanding patient journeys. Our method not only sheds light on using LLMs to efficiently summarize risk factors from social media data but also on LLMs’ potential to identify new potential factors from the patient’s perspective.
Author: [‘Mao C’, ‘Li J’, ‘Pang PC’, ‘Zhu Q’, ‘Chen R’]
Journal: J Med Internet Res
Citation: Mao C, et al. Identifying Kidney Stone Risk Factors Through Patient Experiences With a Large Language Model: Text Analysis and Empirical Study. Identifying Kidney Stone Risk Factors Through Patient Experiences With a Large Language Model: Text Analysis and Empirical Study. 2025; 27:e66365. doi: 10.2196/66365