Advancing the Practice of Heart Surgery to Improve Safety


The Department of Veterans Affairs, in its mission to deliver the highest quality care for our nation’s Veterans, strives to advance the state of heart surgery. At the forefront of these advances are centers like the Medical Robotics and Computer-Assisted Surgery (MRCAS) Laboratory, a joint effort by the VA and Harvard Medical School. MRCAS is led by Dr. Marco Zenati, the Chief of Cardiac Surgery for the VA Boston Healthcare System, and an innovator of surgical practices for over 30 years. Under Dr. Zenati, MRCAS continues to be a leading center for collaborative research on surgical performance, with various studies funded by the National Institutes for Health and industry partners. For the first time, MRCAS heart surgery data collected at the VA is available to outside researchers.
To understand how researchers study heart surgery requires a brief introduction. Heart surgery, or cardiac surgery, is a lifesaving procedure for many patients. Two of the most common procedures performed are Coronary Artery Bypass Grafting (CABG) and Aortic Valve Replacement—which is often a minimally invasive procedure known as Transcatheter Aortic Valve Replacement (TAVR). As many as 400,000 patients undergo CABG each year, and by 2020 more than 276,000 patients in the US had received TAVR since its approval in 2011. As with any heart surgery, these procedures come with their associated risks and require the focus of a coordinated team to perform.
Researchers who study the performance of heart surgery refer to the mental effort of the surgical team as cognitive load. In certain situations, the high demand imposed by surgical tasks and other contributing factors like teaching and flow disruptions may exceed the team members' cognitive capacity, leading to a potentially risky state of cognitive overload. Several recent studies in the literature have demonstrated a direct relationship between cognitive overload and patient outcomes. Although there are several methods of cognitive workload assessment, the majority of existing tools are administered after a procedure and do not allow recognition or course correction of performance in real-time. Therefore, developing a real-time cognitive assessment tool is beneficial since it would allow for the detection of states of cognitive overload that predispose surgeons to errors as well as improve both situational awareness and communication of potential hazardous states.
This data story marks the release of a dataset representing part of an ongoing MRCAS study of the cognitive load experienced by surgical staff during aortic valve replacement and coronary artery bypass graft procedures. The team employs video, audio, biopotential signals, and various other measurements to study cognitive load in cardiac surgery. The primary objectives of this study are as follows:
1. To monitor and evaluate the cognitive workload of surgical teams during cardiac surgeries and its impact on patient outcomes during the operations.
2. To create an advanced cognitive guidance system (CGS) tailored to the surgical process, patient conditions, and the cognitive demands of clinicians. This system will also identify situations of high cognitive workload.
3. To conduct simulations with real cardiac surgery teams to test the CGS's feasibility, usability, and effectiveness in a high-fidelity operating room setting.
The dataset available here is a subset that will be expanded as more data is collected and de-identified. It is currently comprised of 40 surgery records that each have:
· Sound level measurements (SLM) in the operating room,
· Electrocardiogram (ECG) measurements for four key members of the surgical team (i.e., Surgeon, Nurse, Perfusionist, Anesthesiologist), and
· Time-stamped, textual annotations made by team members over the course of each procedure.
Each surgery record has been anonymized and packaged into a JSON file that can be extracted and analyzed using common programming languages such as Python. For more information, or to request access, contact