Factors Affecting Cardiovascular Physiology in Cardiothoracic Surgery: Implications for Lumped-Parameter Modeling

• Main Authors: Joshua Kaufmann, Ethan Kung
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2019-11-01
• Series: Frontiers in Surgery
• Subjects: multi-scale modeling; patient-specific modeling; hemodynamics; virtual surgery; surgical planning
• Online Access: https://www.frontiersin.org/article/10.3389/fsurg.2019.00062/full

Cardiothoracic surgeries are complex procedures during which the patient cardiovascular physiology is constantly changing due to various factors. Physiological changes begin with the induction of anesthesia, whose effects remain active into the postoperative period. Depending on the surgery, patients may require the use of cardiopulmonary bypass and cardioplegia, both of which affect postoperative physiology such as cardiac index and vascular resistance. Complications may arise due to adverse reactions to the surgery, causing hemodynamic instability. In response, fluid resuscitation and/or vasoactive agents with varying effects may be used in the intraoperative or postoperative periods to improve patient hemodynamics. These factors have important implications for lumped-parameter computational models which aim to assist surgical planning and medical device evaluation. Patient-specific models are typically tuned based on patient clinical data which may be asynchronously acquired through invasive techniques such as catheterization, during which the patient may be under the effects of drugs such as anesthesia. Due to the limited clinical data available and the inability to foresee short-term physiological regulation, models often retain preoperative parameters for postoperative predictions; however, without accounting for the physiologic changes that may occur during surgical procedures, the accuracy of these predictive models remains limited. Understanding and incorporating the effects of these factors in cardiovascular models will improve the model fidelity and predictive capabilities.