Towards a Transferable Curriculum in the Training of Thoracic Epidural and Thoracic Paravertebral Blockade Using a Mixed Reality Simulator

By Barys Ihnatsenka, MD, and Linda Le-Wendling, MD    Nov 18, 2016

Barys Ihnatsenka, MD

Nearly 10 years ago, we came up with a novel idea to build a simulator for training on procedures in regional anesthesia (RA). This idea was inspired by a navigation and image fusion technology employed by the Neurosurgery and Ear, Nose, and Throat departments. We believed that by using a similar tracking device, we could track needle movements in a high-fidelity phantom (based on real patient imaging), similar to the way a neurosurgeon can track his instrument in a patient while watching the virtual counterpart of the instrument on the virtual image of the patient’s anatomy.

Linda Le-Wendling, MD

At the beginning, our ambition was to build a comprehensive simulator for all RA procedures, but we narrowed our focus to one region due to the immensity of the task. We decided to focus on anesthesia and analgesia of the thorax. At the time, the University of Florida (UF) Health hospital system had just become a level I trauma center and we had a clinical need to provide good analgesia for patients with multiple rib fractures.

We initially received a local grant from the I. Heermann Anesthesia Foundation to create a prototype in collaboration with our UF engineering team (UF has a very strong tradition in simulation training in anesthesia), which has been instrumental in the undertaking of this project. We created a physical phantom of part of the upper back using the chest CT scan of one of our UF physicians. We constructed the bony structures of the T2-9 bony spine and ribs using a 3D printer and soft tissues out of ballistic gel. We then fused the physical phantom with a 3D virtual image of the bony anatomy, manually adding additional virtual structures of interest such as lung, ligaments, spinal cord, and dural sac, with a focus on keeping the 3D virtual image anatomically correct.

Thus, we created our first version of the mixed reality simulator for thoracic RA. It had both a physical component (needle and phantom) and a virtual component (virtual 3D anatomy image, virtual needle). When the trainee advances a physical needle into the phantom and “lands” on the transverse process (TP), for example, he or she experiences a realistic feeling of hitting bone while simultaneously seeing on the screen an image of the virtual needle touching the virtual TP. If the trainee physically advances the needle past the TP into the area where we programmed our virtual superior costotransverse ligament (SCTL), he or she would notice a thump as the needle engages in ligament and a realistic loss of resistance (LOR) as he or she penetrates the virtual SCTL. The thump and LOR would be triggered at the moment when the physical needle is advanced to the spots in the physical phantom that correspond to the correlating virtual anatomical structure.

With time, our engineers added a greater number of features to the simulator. Early on, we equipped the simulator with a tangible user interface such as a camera that allowed the user to view the virtual anatomy or procedure from different perspectives. One of the greatest additions to the simulator was a mixed reality ultrasound (US) consisting of a physical “dummy” US transducer, whose position is tracked in space, and its virtual counterpart, a virtual US probe with a semitransparent insonating plane displayed in real-time on the computer screen that interacts with the 3D virtual anatomical structures producing a virtual US image. Another cool feature is the ability to replay the procedure with “inside look,” observing and analyzing procedural steps such as US-image acquisition and needling, as if one can see through the skin. Later on, our engineers also added the effect of angle of incidence on US visualization of anatomic structures (such as pleura) and needle. See Figures 1 and 2 for links to videos showing the simulator and its features.

Figure 1. US Assisted TPVB: Right Side TH6

Our mixed reality simulator has been extremely well received. We received a first-place prize for best scientific exhibit from the American Society of Anesthesiologists in 2014 as well as excellent feedback from many national and international experts in the field of regional anesthesia and acute pain medicine.

We began using the simulator to teach thoracic paravertebral (TPVB) and thoracic epidural (TE) blockade. All techniques could be taught in several patient positions (sitting, prone, lateral) and in three versions (landmark-based, US-assisted, and US-guided).

Figure 2. Basic Set Up – Test Mode: No Visualization

Still, we lacked a curriculum to optimize trainee learning using the simulator. We therefore applied for and were awarded the prestigious ASRA Carl Koller Memorial Research Grant in 2014 to further develop the simulator, create a teaching curriculum, and conduct outcomes-based research on the use of simulation in clinical education.

In the past two years, our work on this project can be divided into the following categories

  • Further simulator development
  • Technique refinement for TPVB and TE
  • Curriculum development for TPVB and TE with the creation of an integrated tutor and a focus on transferability of procedural skills into real practice
  • Data acquisition on outcomes with our simulator-based training, which includes development of a testing algorithm to assess technical competence.

Simulator Development

Simulator improvements were based on the following three goals: to improve clinical utility, to enhance educational experience, and to improve product features.

Upgrades to the clinical utility of our simulator included the addition of hydrolocation with US-guidance (ability to visualize anterior pleural displacement when the needle tip is correctly placed in the thoracic paravertebral space), the addition of more anatomical structures such as the internal intercostal membrane and intercostal muscles for the study of more lateral approaches to the TPVB, the creation of two levels of difficulty by decreasing the dimensions of the epidural and paravertebral spaces, and the addition of the presence of a false LOR when the epidural needle exits the interspinous ligament.

To enhance the educational experience of our trainee, we built a new physical simulator (and its virtual correlate) with different anatomical dimensions. This version is intended for testing and was developed in order to assess the trainee more accurately by avoiding biased scoring due to the trainee’s increasing familiarity with the simulator during the learning phase. We also added multiple cognitive aids to help the trainee improve precision when manipulating the US transducer and the needle. These cognitive aids include probe perpendicularity indicator, needle trajectory projection and depth marker, needle and US beam alignment indicator, and labeling of anatomic structures. We created a scoring algorithm based on the different approaches to TPVB and TE in order to help the trainee recognize inadequacies or errors that require attention. For the simulator use we designated two modes of function (training and testing) with new more user friendly interfaces for instructors and trainees.

Product feature improvements were aimed at increasing durability, portability, and dependability of our simulator units so that we can ship the simulator across the country and even across the ocean, allowing us to run workshops with potentially hundreds of simulated procedures in any given day with less hardware and software malfunction.

Technique Refinement

In order to refine our procedural techniques for both TPVB and TE, we had to update our knowledge of the published literature. In addition, we initiated several studies to clarify important anatomical questions and merits/pitfalls of previously described techniques in TPVB/TE that have not been clearly defined in the current body of literature. Our study of CT scans examining the dimensions and distances of the thoracic paravertebral space earned us a first place in category out of 725 abstracts at the International Anesthesia Research Society’s 2016 meeting. We are in the process of completing two other studies that will help us refine our techniques for TPVB and TE placement.

We incorporated our findings into our teachings in order to improve accuracy of needle placement as well as safety (avoiding inadvertent dural or pleural puncture). We tested our revised techniques and elicited feedback from novice and expert learners. We generated multiple hours of lecture material that was eventually condensed into a basic curriculum and a bonus curriculum. We created a multimedia library of videos, photos, drawings, animations and images to aid in explaining our refined procedural techniques. We ran multiple pilot studies to test our approach for failures and inaccuracies.

Curriculum Development

We believe that discovery learning is not as effective as an organized curriculum when trying to master complicated procedures such as TPVB and TE placement. We have made great progress in the development of our integrated tutor (virtual instructor with some basic components of intelligent tutor). The integrated tutor deconstructs these complicated procedures into the basic component steps and helps the learner understand the best way to fine-tune their technique, step-by-step.

The integrated tutor allows for independent learning in the most efficient way possible. Our integrated tutor not only disseminates information using multimedia (videos/photos/animations) but also is intelligent enough to give feedback to the learner. In addition, the use of a virtual tutor allows the learner to study at their own pace without the need to invite experts to his or her institution, which can be time-consuming for the expert and costly for the learner.

The curriculum developed is a fusion of multimedia-based presentations combined with sets of drills and tests designed to constantly educate and assess the learner for their mastery of each of the steps. Due to the depth and breadth of the material, the presentations were divided into more easily digestible blocks:

  • The Fundamentals (overview in clinical practice, anatomy, basics of ultrasound-guided regional anesthesia
  • Entry-Level Techniques (description and demonstration)
  • Troubleshooting (difficult procedures)
  • Skills and Drills (description of basic skills in needle manipulation and US image acquisition including drills for skill mastery)
  • Tests (understanding theory and principles and technical skills practice and assessment of competence)

This curriculum allows flexibility for the learner to repeatedly practice specific skills for his or her level of training and based on feedback from the integrated tutor. The test component of this curriculum includes both a knowledge test and a procedural skills test with immediate feedback to the learner. A score is given to them for both test components. The knowledge test is a series of multiple-choice questions that assess the learner’s grasp of anatomy, procedural technique, block indications/contraindications, and complications. The technical skills test is conducted on a different physical phantom to avoid falsely high scores due to familiarity with the training phantom. The technical skills test is equipped with a scoring algorithm that has been checked for accuracy using experts in the field of RA. Algorithms were tailored to specific techniques (landmark based vs ultrasound assisted vs ultrasound guided).

Data Acquisition

We are in the early phase of data collection acquisition as we are working hard to smooth out glitches and validate our scoring algorithm. Most of the data gathered currently are on test subjects. Enrollment has begun on our IRB-approved study. In our study, we have divided subjects into 3 groups.

  • Group A has access only to traditional lecture material and the simulator without visualization of the 3D virtual anatomy and without the immediate feedback on the quality of performed blocks (Discovery learning).
  • Group B has access to visualization of the 3D anatomy, cognitive aids, and immediate feedback but still utilizes discovery learning without the assistance of the integrated tutor (IT).
  • Group C has access to all the same features as group B but with the addition of the IT.

All subjects regardless of group designation learn TPVB and TE placement without the assistance of an expert in regional anesthesia.

We propose that a curriculum with the IT will be superior to discovery learning without specific objectives and without feedback on skill mastery. We also eventually will test the clinical transferability of the knowledge and skills obtained through this curriculum by asking subjects who had independently trained with our simulator (even novices without previous training in RA) to perform blocks on cadavers and compare their competency with those individuals trained on cadavers by experts in TPVB and TE block performance.

Ultimately, we hope to obtain data on learning curves for TPVB and TE procedures as well as long-term knowledge and skill retention. For those who have mastered basic techniques in TPVB and TE placement, we are in the process of creating a bonus curriculum that covers techniques not covered in the basic curriculum, which includes several variations of US-guided TPVB and of TE placement with patients in different positions (lateral, for example). Our bonus curriculum also covers the fundamentals for out-of-plane US-guided technique.


While we embarked on a journey to instruct trainees in the acquisition of the complex skills of TPVB and TE placement, we as educators and researchers have gathered new insight into WHAT we teach and HOW we teach it. More specifically, we found that it is much harder to teach US-based TPVB techniques (especially US-guided procedures) than landmark-based techniques, moreso in individuals with “US dyslexia.” We discovered that while US was a harder skill to acquire, its mastery helped complete more complicated tasks more successfully. And we learned that even in the advanced practitioner, it takes at least 1-2 hours of practice on the simulator to enhance the ability to efficiently manipulate the needle and to effectively use US as a tool to improve RA success and safety.

We thank ASRA for its generous support, our engineering team for its endless labor and ingenuity, and all of our residents/fellows/faculty for their feedback and data points! We intend to share our educational materials for free with all ASRA members and will be happy to help with courses and workshops for those who are interested in using our simulator in teaching and learning thoracic regional anesthesia. We also believe that our current work is only the beginning of the use of mixed reality simulators with structured curriculum in clinical education and are open for collaboration and further research.

Barys Ihnatsenka, MD, is an assistant professor, and Linda Le-Wendling, MD, is an associate professor, both in the Department of Anesthesiology, College of Medicine, at the University of Florida in Gainesville.

Note: This article originally appeared in the ASRA News, Volume 16, Issue 4, pp. 9-11 (November 2016).

Read more ASRA Blog entries.