A Phenotyping of Diastolic Function by Machine Learning Improves Prediction of Clinical Outcomes in Heart Failure

Diastology in the intensive care unit: Challenges for the assessment and future directions

Myocardial dysfunction is common in patients admitted to the intensive care unit (ICU). Septic disease frequently results in cardiac dysfunction, and sepsis represents the most common cause of admission and death in the ICU. The association between left ventricular (LV) systolic dysfunction and mortality is not clear for critically ill patients. Conversely, LV diastolic dysfunction (DD) seems increasingly recognized as a factor associated with poor outcomes, not only in sepsis but also more generally in critically ill patients. Despite recent attempts to simplify the diagnosis and grading of DD, this remains relatively complex, with the need to use several echocardiographic parameters. Furthermore, the current guidelines have several intrinsic limitations when applied to the ICU setting. In this manuscript, we discuss the challenges in DD classification when applied to critically ill patients, the importance of left atrial pressure estimates for the management of patients in ICU, and whether the study of cardiac dysfunction spectrum during critical illness may benefit from the integration of left ventricular and left atrial strain data to improve diagnostic accuracy and implications for the treatment and prognosis.

The Heart and Artificial Intelligence-How Can We Improve Medicine Without Causing Harm

PURPOSE OF REVIEW

The introduction of Artificial Intelligence into the healthcare system offers enormous opportunities for biomedical research, the improvement of patient care, and cost reduction in high-end medicine. Digital concepts and workflows are already playing an increasingly important role in cardiology. The fusion of computer science and medicine offers great transformative potential and enables enormous acceleration processes in cardiovascular medicine.

RECENT FINDINGS

As medical data becomes smart, it is also becoming more valuable and vulnerable to malicious actors. In addition, the gap between what is technically possible and what is allowed by privacy legislation is growing. Principles of the General Data Protection Regulation that have been in force since May 2018, such as transparency, purpose limitation, and data minimization, seem to hinder the development and use of Artificial Intelligence. Concepts to secure data integrity and incorporate legal and ethical principles can help to avoid the potential risks of digitization and may result in an European leadership in regard to privacy protection and AI. The following review provides an overview of relevant aspects of Artificial Intelligence and Machine Learning, highlights selected applications in cardiology, and discusses central ethical and legal considerations.

Left Atrial Strain in the Assessment of Diastolic Function in Heart Failure: A Machine Learning Approach

BACKGROUND

Diastolic dysfunction (DD) assessment in heart failure is still challenging. Peak atrial longitudinal strain (PALS) is strongly related to end-diastolic pressure and prognosis, but it is still not part of standard DD assessment. We tested the hypothesis that a machine learning approach would be useful to include PALS in DD classification and refine prognostic stratification.

METHODS

In a derivation cohort of 864 heart failure patients in sinus rhythm (age, 66.6±12 years; heart failure with reduced ejection fraction, n=541; heart failure with mildly reduced ejection fraction, n=129; heart failure with preserved ejection fraction, n=194), machine learning techniques were retrospectively applied to PALS and guideline-recommended diastolic variables. Outcome (death/heart failure rehospitalization) of the identified DD-clusters was compared with that by guidelines-based classification. To identify the best combination of variables able to classify patients in one of the identified DD-clusters, classification and regression tree analysis was applied (with DD-clusters as dependent variable and PALS plus guidelines-recommended diastolic variables as explanatory variables). The algorithm was subsequently validated in a prospective cohort of 189 heart failure outpatients (age, 65±13 years).

RESULTS

Three distinct echocardiographic DD-clusters were identified (cluster-1, n=212; cluster-2, n=376; cluster-3 DD, n=276), with modest agreement with guidelines-recommended classification (kappa=0.40; P<0.001). DD-clusters were predicted by a simple algorithm including E/A ratio, left atrial volume index, E/e' ratio, and PALS. After 36.5±29.4 months follow-up, 318 events occurred. Compared to guideline-based classification, DD-clusters showed a better association with events in multivariable models (C-index 0.720 versus 0.733, P=0.033; net reclassification improvement 0.166 [95% CI, 0.035-0.276], P=0.013), without interaction with ejection fraction category. In the validation cohort (median follow-up: 18.5 months), cluster-based classification better predicted outcome than guideline-based classification (C-index 0.80 versus 0.78, P=0.093).

CONCLUSIONS

Integrating PALS by machine learning algorithm in DD classification improves risk stratification over recommended current criteria, regardless of ejection fraction status. This proof of concept study needs further validation of the proposed algorithm to assess generalizability to other populations.