Hysteroscopy is a branch of laparoscopic surgery that plays a major role in OBGYN practice. However, hysteroscopic procedures are plagued by high rates of complications. Hysteroscopy simulators have been developed in an effort to improve the training of medical residents, but currently available models do not incorporate haptic feedback with quantitative measurements of the surgeon’s competence. Our design team has identified and defined the major needs for a 3D haptic hysteroscopic surgery simulator with quantitative feedback and positive assessment and improved surgery results from medical students and residents at Vanderbilt Medical Center.

We have purchased capacitive touch sensors and resistance touch sensors. We will embed these sensors into the clamshell uterus design in order to create the hardware for surgical tasks. The sensors will serve as surrogate tumors and other obstacles that the surgeon would encounter during the course of a hysteroscopy. One design we have conceived of requires a binary discrimination of successful target identification and unintended scratching and touching of the healthy uterus interior. One touch sensor will be affixed to the tip of the hysteroscopy surgical rod. If signals from both the hysteroscopy sensor and the uterus are concurrent, the signal will be recorded as a successful attempt. If the hysteroscopy sensor signals without the uterus wall, the trial will indicate that the hysteroscopy instrument struck an unintended portion of the uterus. The number of successes and failures as well the surgeon’s time to accomplish the task will be quantitative measures that we can track for each trial. Integrating the camera, touch sensors and processing the inputs has led us to the next step.

We have purchased a Raspberry Pi computer, which is a small and affordable Linux-based system. We hope to use the Raspberry Pi with our previously-purchased camera to provide a video feed of the simulation in a compact and self-contained system. In addition, the Pi has several input and output pins for interfacing with sensors. These pins will allow us to monitor the surgery environment and keep track of certain events that are detected by electronic sensors. A capacitive touch sensor and pressure sensor were purchased. In the future, these will be used to report contact of the hysteroscope with objects in the surgery environment.

As the design has developed, we have experienced a few minor hurdles that have helped us refine our design. Our first hurdle was deciding how we wanted to implement a touch interface within the closed product. In hysteroscopic surgery, it is critical to hone a resident’s ability to navigate the uterus without puncturing the uterine wall. In order to avoid this costly mistake, it is important to implement touch sensing on the walls of the training device to provide training feedback. We have explored the possibilities of using electrically conductive materials, touchpads akin to iPhone screens, and mechanical pressure sensors. For the first design iteration, we have decided to utilize pressure sensors due to the low learning curve and affordable pricing.

The second major hurdle we’ve encountered is deciding how to implement a visual system that can provide real-time images from within the device. The addition of a visual system is important because it will allow the design to be completely self-sustained and not require external peripherals like a monitor. The primary obstacle with implementing a screen is interfacing it with the device’s processing unit, the Raspberry Pi. A Raspberry Pi is a cheap processor with basic functionality that make it ideal for low level interfaces and computing. At the present time, we are utilizing a laptop to function as the screen but are hoping to find a suitable screen before our next report.

Finally, we are restricted by our software experience as no one in the group has experience creating a SoftWorks 3D model. Using SoftWorks is critical to designing a device that accurately replicates the setup of the uterus and requires substantial training in order to create a usable model. To overcome this hurdle ,we have enlisted the help of a fellow Vanderbilt mechanical engineering student who has enough training to put together a rudimentary model. With his help, we will be able to print a first prototype and manipulate the existing SoftWorks model as we see fit. Collaboration is an important part of proper engineering and something that Team Valor strives to implement in all iterations of its design.
Moving forward, we are anticipating future hurdles regarding our testing protocol. We will need to work with our advisors to determine the most effective means of testing. Having the physical model will also give us a better idea of what tasks we should consider for the testing protocol. All hurdles up to this point were anticipated and, at the moment, no major setbacks have interrupted the entire project.

Now that we have brainstormed and come up with a workable idea for a 3D printed uterus model, a way to gather touch sensing data via pressure sensors, and the technology needed to display the camera output using a raspberry pi, the next step will be acquiring the supplies and implementing the ideas. Our initial prototype of the 3D printed uterus gave us a general idea of shape and size that will be required for a working model. With this information we will obtain help from a resource with SoftWorks experience and 3D print our design in Vanderbilt’s Design Studio. We have spent much time in our meetings doing research and discussing the various options for how to detect where in the model the hysteroscope touches to give feedback on surgeon accuracy. Small pressure sensors will be placed inside the uterus model so that when the scope bumps them it will be recorded if and where it touched. We will need supplies to install the sensors and a device such as an arduino or raspberry pi to record the data from them. We have also spent much time discussing how to display the live camera feed in order to see inside the uterus model while practicing the hysteroscopic surgery. Already having the camera to attach to the surgical instrument, we chose a raspberry pi to attach it to do process the data and display the feed. The next steps will include programming the device and deciding how to attach it and display it along with the 3D model.