The utilization of artificial intelligence (AI) in predicting cardiovascular disease (CVD) risk is rapidly transforming the healthcare landscape. A systematic review published in the Journal of Medical Systems highlights significant advancements in AI-based predictive models, particularly emphasizing the importance of accounting for right-censored data. This comprehensive analysis sheds light on the strides made by machine learning (ML) and deep learning (DL) models, revealing their efficacy in survival outcome prediction for CVD.
AI Models in CVD Prediction
The review encompasses 33 studies that employ various ML and DL models for predicting CVD outcomes. The majority of these studies were conducted in the United States, with nearly half published in 2023. Among the ML models, Random Survival Forest (RSF), Survival Gradient Boosting models, and Penalized Cox models were identified as the most frequently used. DeepSurv emerged as the leading DL model, demonstrating superior predictive capabilities compared to ML models.
Explainable AI Techniques
In terms of explainability, permutation-based feature importance and Shapley values were the most commonly utilized eXplainable AI (XAI) methods. These techniques provide insights into how AI models make predictions, which is crucial for clinical acceptance and integration. The review also noted a significant gap in the inclusion of social determinants of health (SDoH) and gender-stratification in these predictive models, with only one in five studies performing gender-stratification analysis.
From a market access perspective, the findings underscore the potential for these AI models to be integrated into clinical practice, enhancing early detection and personalized treatment plans. However, the limited consideration of SDoH and gender factors could pose barriers to widespread adoption and equitable access to these advanced predictive tools.
Key Inferences
Key takeaways from the review include:
- DeepSurv models outperform ML models in predicting CVD outcomes.
- RSF stands out as a robust ML model for survival analysis.
- XAI methods such as permutation-based feature importance and Shapley values are essential for model transparency.
- There is a critical need to incorporate SDoH and gender-stratification in future AI models to ensure comprehensive risk assessment.
In conclusion, this systematic review highlights the current state of AI-based models in CVD prediction and points to areas needing further research and refinement. Ensuring that these models account for a broader range of health determinants and gender differences will be pivotal for their clinical utility and acceptance.
Original Article:
J Med Syst. 2024 Jul 19;48(1):68. doi: 10.1007/s10916-024-02087-7.
ABSTRACT
Artificial intelligence (AI) based predictive models for early detection of cardiovascular disease (CVD) risk are increasingly being utilised. However, AI based risk prediction models that account for right-censored data have been overlooked. This systematic review (PROSPERO protocol CRD42023492655) includes 33 studies that utilised machine learning (ML) and deep learning (DL) models for survival outcome in CVD prediction. We provided details on the employed ML and DL models, eXplainable AI (XAI) techniques, and type of included variables, with a focus on social determinants of health (SDoH) and gender-stratification. Approximately half of the studies were published in 2023 with the majority from the United States. Random Survival Forest (RSF), Survival Gradient Boosting models, and Penalised Cox models were the most frequently employed ML models. DeepSurv was the most frequently employed DL model. DL models were better at predicting CVD outcomes than ML models. Permutation-based feature importance and Shapley values were the most utilised XAI methods for explaining AI models. Moreover, only one in five studies performed gender-stratification analysis and very few incorporate the wide range of SDoH factors in their prediction model. In conclusion, the evidence indicates that RSF and DeepSurv models are currently the optimal models for predicting CVD outcomes. This study also highlights the better predictive ability of DL survival models, compared to ML models. Future research should ensure the appropriate interpretation of AI models, accounting for SDoH, and gender stratification, as gender plays a significant role in CVD occurrence.
PMID:39028429 | DOI:10.1007/s10916-024-02087-7
This article has been prepared with the assistance of AI and reviewed by an editor. For more details, please refer to our Terms and Conditions. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author.
