ABSTRACT
BACKGROUND AND OBJECTIVE: To develop a healthcare chatbot service (AI-guided bot) that conducts real-time conversations using large language models to provide accurate health information to patients. METHODS: To provide accurate and specialized medical responses, we integrated several cancer practice guidelines. The size of the integrated meta-dataset was 1.17 million tokens. The integrated and classified metadata were extracted, transformed into text, segmented to specific character lengths, and vectorized using the embedding model. The AI-guide bot was implemented using Python 3.9. To enhance the scalability and incorporate the integrated dataset, we combined the AI-guide bot with OpenAI and the LangChain framework. To generate user-friendly conversations, a language model was developed based on Chat-Generative Pretrained Transformer (ChatGPT), an interactive conversational chatbot powered by GPT-3.5. The AI-guide bot was implemented using ChatGPT3.5 from Sep. 2023 to Jan. 2024. RESULTS: The AI-guide bot allowed users to select their desired cancer type and language for conversational interactions. The AI-guided bot was designed to expand its capabilities to encompass multiple major cancer types. The performance of the AI-guide bot responses was 90.98 ± 4.02 (obtained by summing up the Likert scores). CONCLUSIONS: The AI-guide bot can provide medical information quickly and accurately to patients with cancer who are concerned about their health.
ABSTRACT
BACKGROUND: The International Agency for Research on Cancer classifies 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) as a known carcinogen. This study aimed to investigate the association between exposure to secondhand smoke (SHS) and NNAL concentrations in non-smokers. METHODS: This was a cross-sectional study based on data from the 2016 to 2018 Korea National Health and Nutrition Examination Survey. Urine NNAL concentrations were categorized into tertiles of 3,615 study participants who were non-smokers. All sampling and weight variables were stratified, and analyses to account for the complex sampling design were conducted. RESULTS: The overall, male, and female percentages of SHS exposure among non-smokers were 22.4%, 29.2%, and 20.4%, respectively. The geometric means of urine NNAL concentrations were 1.896±0.098 pg/mL and 1.094±0.028 pg/mL in the SHS exposure and non-exposure groups, respectively. After adjusting for confounding variables, in the total group, the geometric mean of urine NNAL concentrations was significantly higher in the SHS exposure group than in the SHS non-exposure group (adjusted P-value <0.001). Compared with the non-exposure group, the adjusted odds ratios (95% confidence intervals) for the highest NNAL tertile group of overall SHS exposure in the total, men, and women groups were 2.44 (1.95-3.05), 1.65 (1.08-2.53), and 2.73 (2.11-3.52), respectively, after full adjustment. CONCLUSION: The urine NNAL concentration in the SHS exposure group was significantly higher than that in the non-exposure group. Exposure to SHS was associated with a higher risk of elevated urine NNAL concentrations in non-smokers.
