JMIR Publications

ABSTRACT

Background:

Cardiotoxicity is a major concern in heart disease research, as it can lead to severe heart damage, including heart failure and arrhythmias.

Objective:

This study aimed to explore the capability of ChatGPT in generating innovative research hypotheses to address five major challenges in cardiotoxicity research: complexity of mechanisms, variability among patients, lack of detection sensitivity, lack of reliable biomarkers, and limitations of animal models.

Methods:

ChatGPT4.o was employed to generate multiple hypotheses for each of the five challenges. These hypotheses were then independently evaluated by three experts for novelty and feasibility. ChatGPT subsequently selected the most promising hypothesis from each category and provided detailed experimental plans, including background, rationale, experimental design, expected outcomes, potential pitfalls, and alternative approaches.

Results:

ChatGPT generated 96 hypotheses, with 13 (13.5%) rated as highly novel and 62 (64.6%) as moderately novels. The average group score of 3.85 indicates a strong level of innovation in these hypotheses. Literature searching found that 28 (29.2%) hypotheses had at least one relevant publication. The selected hypotheses included using single-cell RNA sequencing to understand cellular heterogeneity, integrating AI with genetic profiles for personalized cardiotoxicity risk prediction, applying machine learning to ECG data for enhanced detection sensitivity, utilizing multi-omics approaches for biomarker discovery, and developing 3D bioprinted heart tissues to overcome the limitations of animal models. Our group's evaluation of the 30 components of the experimental plans for five hypotheses revealed consistent strengths in the background, rationale, and alternative approaches from ChatGPT, with most hypotheses receiving scores of 4 or higher in these areas. While the hypotheses were generally well-received, the experimental designs were often deemed overly ambitious, highlighting the need for more practical considerations.

Conclusions:

Our study demonstrates that ChatGPT can generate innovative and potentially impactful hypotheses for overcoming critical challenges in cardiotoxicity research suggesting that AI-driven hypothesis generation could play a crucial role in advancing the field of cardiotoxicity, leading to more accurate predictions, earlier detection, and better patient outcomes.