The following are key points to remember from a state-of-the-art review about artificial intelligence (AI) in cardiovascular (CV) clinical trials:

1. AI can improve trial design by generating simplified inclusion criteria and more targeted subgroups for analysis, with adaptive eligibility strategies offering an estimated 15-20% reduction in sample size.
2. AI can enhance and improve the cost-effectiveness of eligibility screening compared to human review by more efficiently evaluating and synthesizing information in addition to discrete data.
3. Generative large language models can reduce the length and complexity of informed consent forms and combining with an AI chatbot would make enrollment accessible any time.
4. Automated clinical endpoint adjudication by AI tools using natural language processing may allow for more rapid and consistent identification of benefit or unacceptable risk compared to expensive, labor intensive central clinical events committees; however, further development and validation are needed.
5. Use of AI to analyze digital biomarkers and wearable devices and interpret the vast amount of data being collected will present novel challenges on interpretation of these alongside traditional clinical biomarkers. Of note, these technologies may not be accessible for patients who lack technological knowledge or due to cost.
6. CV imaging traditionally requires centralized review; however, several deep learning models provide automated interpretation of CV imaging with some already validated and shown to be less costly and more reproducible than traditional methods.
7. AI such as chatbots, wearable devices, and outcomes analysis may assist in shifting trials to a decentralized setting and allow for a robust, continuously updated data set.
8. Generative AI has been incorporated in the preparation of manuscripts, with many major journals now allowing AI use for this purpose as long as use is disclosed and authors take full responsibility of the final manuscript.
9. US Food and Drug administration areas of focus regarding the development and use of AI include fostering collaboration to safeguard public health, advance the development of regulatory approaches that support innovation, promote the development of harmonized standards, guidelines, best practices, and tools, and support research related to the evaluation and monitoring of AI performance.
10. Key risks of AI in clinical trials include poor generalizability, data set shift, algorithmic bias, lack of clinical interpretability, patient data privacy, patient access to technology, and loss of human competency, and mitigation strategies must be utilized to ensure patient safety and result integrity.

Clinical Topics: Cardiovascular Care Team

Keywords: Artificial Intelligence, Clinical Trials as Topic, Deep Learning

< Back to Listings