Back to search

FRINATEK-Fri prosj.st. mat.,naturv.,tek

Personalized Virtual Heart Models for Diagnosis and Treatment Planning in Patients with Heart Failure

Alternative title: Personaliserte virtuelle hjertemodeller for diagnose og behandlingsplanlegging hos hjertesviktpasienter

Awarded: NOK 8.0 mill.

Arrhythmias, characterized by irregular beating of the heart, persist as a prominent cause of death in Norway. However, accurately assessing the risk of lethal outcomes in patients remain a challenge. MyVirtualHF envisions an innovative approach to the identification of individuals at high risk for these life-threatening arrhythmias by leveraging cutting-edge computer models of the heart. The project addresses the current limitations in diagnosis and treatment planning, which often involve costly and invasive medical procedures. MyVirtualHF has made significant strides in advancing the creation of personalized heart models derived from magnetic resonance images (MRI). A semi-automatic pipeline has been developed to streamline the generation of finite element models from MRIs, ensuring reproducibility and consistency while minimizing user input. An added breakthrough involves the introduction of a novel coordinate system, simplifying the precise localization of diseased regions within the heart. This coordinate system not only aids in pinpointing pathological areas but also facilitates the creation of new models, enabling systematic exploration of the relationship between various aspects of heart disease and their contributions to arrhythmia. Building upon these advancements, MyVirtualHF has successfully generated a comprehensive library of heart models, each exhibiting distinct types of scars. Through simulations utilizing these virtual hearts, the project has gained valuable insights into the intricate relationship between patient-specific heart structures and the propensity for deadly arrhythmias. Importantly, machine learning techniques have been incorporated to identify specific imaging features that serve as predictive indicators of patient risk. Expanding the scope of its impact, MyVirtualHF has applied these sophisticated modeling techniques to create accurate representation of pregnant women's and the developing fetus's hearts. This extension aims to significantly enhance the monitoring of fetal cardiac health, showcasing the versatility and potential of this technology across various life stages. As MyVirtualHF continues to refine and improve these complex computer models, it remains committed to unlocking the full potential of this technology. The project stands at the forefront of reshaping the diagnosis of arrhythmic risk, offering transformative insights and applications that span the entire spectrum of life, from prenatal stages through adulthood.

The prevalence of Heart Failure (HF) in Europe is 5-8% and a leading cause of hospitalizations. However, the diagnosis and management of HF patients remains a challenge. The vision of MyVirtualHF is to reshape the diagnosis and treatment of HF by the development of personalized virtual heart computer models for therapeutic decision making. This vision addresses the current lack of specificity in designing care for the individual HF patient. Current medical standards rely on imprecise and insensitive biomarkers (such as ejection fraction or ECGs) to make predictions about patient trajectory. While more invasive measurements (such as pacing induced arrhythmia) have shown promise as better outcome predictors, the high risk for complications associated with the procedures makes implementation in the clinic inadvisable. Finally, predicting an individual patient’s response to therapies remain difficult. Often, drugs or procedures need to be fine-tuned or repeated in order to reach optimal therapeutic response from patient. The aim of MyVirtualHF is to address this limitation in current clinical care by developing personalized virtual heart models that can be used by physicians to non-invasively and safely diagnose and predict therapeutic response in HF patients. In order to do so, MyVirtualHF will develop efficient personalization methods that incorporate an individual HF patient’s geometry, genetics, electrophysiological changes, etc. into virtual models. The heterogeneity and complexity of remodeling that occurs under HF makes personalization an inherently risky endeavor. However, the success of MyVirtualHF could lead to a paradigm shift in treating and ensuring health in HF patients.

Funding scheme:

FRINATEK-Fri prosj.st. mat.,naturv.,tek