Research
The research conducted in the T.E.C.I. Center has revealed previously unknown performance metrics regarding mastery in bedside procedures and surgical operations. The T.E.C.I. Center team aims to transform human health and welfare through advances in data science and personalized, data-driven performance metrics for healthcare providers.
Stars of all Colors: Quantifying Variability in Visual-Motor and Cognitive Expertise in Demanding Surgical Tasks
In high-stakes environments, such as high-caliber athletics, flying a fighter jet, and performing minimally invasive surgery, individuals with talent and elite training are expected to perform their best – reacting to unexpected circumstances with both physical and cognitive agility. Improving performance assessment and training requires an understanding of the mechanisms underlying how variability in specific cognitive, technical, and psychomotor actions relates to outcomes. Our research infrastructure allows detailed capture of surgeon movements, forces, sensory feedback, and machine learning to identify features in time-series data, and performance augmentation capabilities.
This project is funded by Stanford Wu Tsai Human Performance Alliance
The Sensor Enabled Smart Ultrasound
With the increased use of ultrasound technology for diagnostics (especially at the bedside), a wide variety of physicians who historically were not operating these devices now use them on a daily basis. Ultrasounds are only as useful as the quality of the image-derived, and that quality is highly dependent on the ultrasonographer. There is an increasing need to provide proper training on the use of ultrasound across many medical specialties. Problematically, this increasing need is met with a lack of effective training programs for physicians and sonographers, which has led to low-quality ultrasounds and ultimately misdiagnoses.
This project is funded by Beijing Institute for Collaborative Innovation (BICI)
Individualized performance optimization for surgeons: Using biometric analytics to understand the impact of stressors and recovery on surgeon performance and skill acquisition
The objective of this proposal is to quantify the effect of stressors on the physiology and subjective experience of surgeons in the operating room. Our overall hypothesis is that a significant proportion of attending and trainee surgeons are under physiologic stress that may affect their operative performance.
This project is funded by Intuitive Foundation
Leveraging Embedded Haptic Sensor Technology for Force Vector Mapping in Orthoses for Adolescent Idiopathic Scoliosis
Adolescent idiopathic scoliosis (AIS) is a condition involving excessive coronal plane curvature of the spine and without effective treatment, severe AIS can cause significant morbidity. Treatment by operative intervention is commonly associated with adverse events and high rates of revision. As such, the gold standard of non-operative management is correction by means of long-term utilization of a custom-fitted rigid brace, however, the effectiveness of orthoses is limited by the qualitative nature of brace fitting and adjustment. Current practice is based on clinician judgment which is variable and dependent on decisions that may not mirror those made in well-controlled trial settings by presumptive experts. The purpose of the proposed research is to explore the use of flexible force sensor technology to augment the efficacy of scoliosis bracing by measuring the force profile of the orthotic across the torso, mapped to individual patient anatomy. Objectives of the proposed research are (1) to refine technology designs for clinical introduction, (2) pilot clinical introduction of the technology, and (3) conduct longitudinal clinical monitoring and determine the usefulness of this technology in the treatment of AIS. The proposed work aims to improve orthotic interventions, potentially resulting in better outcomes and fewer surgeries among AIS patients.
Project funded by Scoliosis Research Society
Quantifying Metrics of Surgical Mastery: An Exploration in Data Science
This project aims to develop a dataset of multiple synchronized data modalities on surgical performance to develop and validate metrics of surgical mastery, and to identify an implementation strategy for using a metrics-based tool for surgeons.
This project is funded by the NIH
Utilizing Digital Technology to Foster Community Engagement and Increase Breast Health Awareness Among Diverse Populations
The causes of breast cancer disparities arise from a complex interplay of multiple factors that impede awareness and limit access and engagement in preventative and follow-up care. The goal of the proposed research is to build on our prior success at women's wellness events and use this information to take a deeper dive into the multiple factors that contribute to the underuse of preventive breast health services among subgroups in our catchment area. This proposal aims to: (1) Develop and implement a multi-pronged educational experience that includes immersive community engagement at two levels: (A) a brief educational video featuring a digital breast exam simulator to facilitate breast health knowledge and engagement in preventive care; (B) a culturally tailored and competent survey to assess breast health experience and knowledge, as well as barriers to engagement in collaboration with community liaisons in three community health education venues; and (2) Use interviews and survey data to identify aids and barriers to engagement in preventive behaviors and acceptance of the immersive learning environment and simulator technology. Results of this program will inform the design and testing of a full-scale, culturally tailored community-based intervention to improve engagement in breast cancer preventive practices and services among a wider variety of communities at high risk.
Project funded by Stanford Spectrum Community Engagement Program
Surgical Metrics Project
The Surgical Metrics Project was introduced to attendees during The American College of Surgeons Clinical Congress 2019. The Surgical Metrics Project booth provided participants with the opportunity to engage in an exploration of the use of wearable technologies to measure surgical decision making and surgical technique.
Assessment of Cardiac Surgical Competence in a High-Fidelity Simulation Environment
The T.E.C.I. Center team has partnered with Cardiothoracic Surgeon, Dr. Joshua Hermsen, at the University of Wisconsin – Madison to evaluate the cognitive and technical aspects of heart surgery. In a simulated operating room, participants use the Cardiac Surgery Simulator (KindHeart, Chapel Hill, NC) to perform an open-heart operation requiring atrial to aortic CPB and cardioplegic arrest. A surgical assistant, an anesthesiologist, a perfusionist, a scrub technician, and all necessary equipment needed to cannulate the patient, initiate CPB, cross-clamp the aorta and arrest the heart with antegrade cardioplegia are provided during the surgery. Electromagnetic motion sensors were placed on both hands pf the surgeon and surgical assistant while operating in order to capture real-time hand movements and quantifiable performance metrics. Findings from these objective metrics will lead to improved resident training, better patient outcomes and an understanding of surgical expertise.
Motion Tracking During Intubation on a Sensor-Enabled Manikin
The T.E.C.I. Center has worked with a wide range of clinical specialties to capture haptic performance data during high risk bedside procedures by applying motion tracking and sensor technology physicians and their instruments.
In October 2018, the T.E.C.I. team held a major performance data event at the annual American Society of Anesthesiology Annual Meeting. During this meeting, over 130 anesthesiologists volunteered to be instrumented with sensors and perform two simulated intubations (normal patient & critically burned patient). Sensors on the manikin, physicians and their instruments enabled digital documentation of a variety of approaches to intubation.
Pictured above is a bird’s eye view of the data collection environment (left) and the motion tracking output of the laryngoscope (right).
Participants lined up to contribute their motion and sensor data to our intubation database.
Motion Tracking During a Portion of Simulated Cataract Surgery
Surgical procedures involve a wide variety of complex technical and cognitive decisions that must be carefully choreographed to achieve the highest quality outcomes for patients. Motion tracking technology provides a unique opportunity to digitize every critical maneuver a physician makes to complete an operation. The resulting “Digital Choreograph” can be used to make major advancements in surgical documentation, communication and training.
The T.E.C.I. Center is collaborating with a variety of surgical specialists to digitize both rare and common surgical procedures.
Pictured to the left is T.E.C.I. team members working with Dr. Robert Chang of Stanford Ophthalmology.
Motion Tracking in a Simulated Microsurgery Environment
The T.E.C.I. Center partnered with University of Wisconsin-Madison Plastic Surgeon, Dr. Samuel Poore, to collect motion, video, and audio data from practicing physicians during a simulated microsurgery procedure. Data collection participants were asked to conduct an anastomosis of a divided dorsal artery on a chicken foot model using a 10-0 nylon suture under an operating microscope. Participants were also given an assistant to help with the procedure. With this data collection, we are able to capture psychomotor skill from expert microsurgeons gaining some fundamental insight into objective skill metrics for expertise.
Application of Motion Based Technology in the Objective Assessment of Laparoscopic Suturing of a Validated Vaginal Cuff Simulation Model
The purpose of this research study is to evaluate various objective psychomotor skills that differentiate novice versus expert surgeons when performing laparoscopic suturing of a validated vaginal cuff model (pictured left). This information will assist in identifying key steps with the greatest objective difference, and assist with formative feedback. Furthermore, this information will guide teaching and deliberate practice.
Sensor-Enabled Breast Simulator
The T.E.C.I. Center team has developed a sensor-enabled breast simulator (pictured below) to train women how to perform a proper self-breast examination. This health awareness project is based on a database of over 2,000 physician breast examinations confirming the best technique to locate a palpable breast lesion.
Our goal is to empower women to take charge of their breast health by promoting a personally interactive and hands-on experience that uses culturally sensitive haptic technology combined with artificial intelligence.
The T.E.C.I. Center team has established community partnerships disseminate this training and to empirically investigate women’s interest, fears and baseline understanding regarding breast health.
Real-time haptic feedback is displayed as the breast is examined.
Pictured above are sensor maps of effective breast exam technique on the left and ineffective breast exam technque on the left.
Virtual Reality Simulation
In a virtual reality environment, participants were asked to perform a target-tracking task using a haptic device. Participants used a stylus to follow a moving target on a screen. To challenge participants to demonstrate their psychomotor abilities, distracting forces were applied to the stylus throughout the task with varying levels of force. Psychomotor skills demonstrated during this task include hand-eye coordination, motor-control, reaction time and error management. Testing surgical residents using this VR environment setting indicate that residents who left clinical practice and became active in research lost some of their psychomotor skills after two years. Findings in this study will have an impact on the future design of surgical curricula aiming to prevent skill decay during surgical residency.
Tourniquet Master Training System for Junctional and Inguinal Hemorrhage Control Devices
Junctional and inguinal bleeding is a significant and challenging problem on the battlefield. Several inventors have developed new types of tourniquets, including the Abdominal Aortic Tourniquet (AAJT) as well as others, to address groin and pelvic injuries. While these new hemorrhage control technologies have been developed, validated, and approved for use, training systems that teach and refresh skills related to these technologies have not been developed. Training systems for junctional injuries are vital because the injuries are rare on the battlefield and difficult to prepare for.
During this effort, we propose to develop and thoroughly evaluate a full-scope prototype Tourniquet Master Training (TMT) training system. TMT is a scenario-based training system featuring (1) a reconfigurable sensor system linked to a software-based virtual mentor that provides objective assessment during training, and (2) a mobile mentor that provides refresher training. TMT is a sensor-based system usable with any manikin, and because of this flexibility, TMT is an affordable training system that is configurable for future advances, including new application areas for both existing and future devices.
Motion Tracking During a Simulated Laparoscopic Ventral Hernia Repair
Based on our prior work with hands-on medical examinations, we established that top tier performers utilize similar haptic patterns. To translate this to surgical procedures, we utilized motion data during a simulated laparoscopic ventral hernia repair (LVH). We hypothesized that motion data can stratify top and bottom tier performers for streamline video review.
Surgical residents (N=37) from seven Midwest programs performed a partial LVH repair on a validated benchtop simulator. Their hand movements were captured using electromagnetic motion tracking sensors. Each repair was graded with a final product score (FPS).
We then identified the top and bottom ten performers based on FPS and reviewed their motion plots. Finally, two blinded raters independently reviewed motion plots to identify performance. Top performers had significantly higher FPS than bottom performers (23. 3 ± 1. 2 vs 5. 7 ± 1. 6 p<0. 01). Both raters could stratify top performers from bottom performers through motion plots compared to stratification based on FPS, Rater1 chi2 7. 2 p<0. 01 and Rater2 chi2 5. 5 p=0. 019. The motion plots of the bottom tier performers had differences from the top tier performers in which we correlate to relevant portions of the procedure. By reviewing motion plots from the top and bottom tier performers during simulated LVH repair, we can stratify performance levels. Differences in motion data can enable rapid review to identify training needs and feedback. This is a significant step to artificial intelligence in surgical training assessment.