1/30/15

 

Meeting with Dr. Anderson

On Tuesday January 27, we had an informative meeting with our project advisor, Dr. Ted Anderson.  The agenda of our meeting included bringing Dr. Anderson up to date on our recent progress, discussing the current state of technology being used in hysteroscopy simulations, and collecting information on the aspects of the project that he would like us to prioritize.  In addition to these topics, Dr. Anderson also gave our team two hysteroscopes to borrow and is interested in scheduling time for us to shadow him and watch him perform a hysteroscopic surgery in the near future.

We showed Dr. Anderson the 3D printed uterus model, the small USB lighted camera, and our ideas to use the Raspberry Pi and touch sensors for a diagnostic task within the simulation.  Dr. Anderson was very pleased with this progress and gave us suggestions for how to improve the design to best suit the needs of the surgeons.  For example, he requested that the size be slightly smaller than our original design to be a more accurate representation of an actual uterus shape.  He also asked us to coat the insides of the model with a more light-reflective material so visualizing the inside of the uterus will be easier. Overall, Dr. Anderson was very pleased with our progress and is excited to see more.

Dr. Anderson is excited about the progress of our work because currently there is nothing else like this on the market.  There are computer simulations to practice hysteroscopic surgeries and physical simulations for other surgeries (for example laparoscopic surgeries) but no physical models that directly correlate to the surgical skills needed in hysteroscopy.  This model will be the first of its kind to offer both physical experience and objective and quantitative means of assessing these skills.

We also discussed the skills being assessed in the simulation with Dr. Anderson and what tasks will be best to test these skills.  We came up with 3-4 tasks to include in the final model that will give surgeons in training a good idea of what is needed for a successful surgery.  These tasks include a diagnostic task, a task involving threading a catheter into a narrow hole, and an object retrieval task . These tasks can be proctored by another doctor to assess how well they are being performed, but we also eventually hope to have them monitored by the model itself so the surgeon can practice and track their progress on their own.

 

Scopes and functionality plus cost

In speaking with Dr. Anderson, we expressed interest in using a real hysteroscope to better understand the limitations of the surgical simulator. He was kind enough to provide two hysteroscopes for us: one diagnostic and one operative. The hysteroscopes operate on a very simple principle: provide a way to look inside and operate on uteruses. The scopes, though, are anything but simple. The scopes’ metal bodies are a series of interconnected rods that can be detached and cleaned between each surgery. An optical scope allows a person to view the inside of the uterus or connect it to an exterior camera.

These scopes have given us a better idea for what the user will experience while using the hysteroscopic surgery simulator. In order to examine every part of the uterus, the end of the scope is angled at either 12 or 30 degrees. While this allows the user to see all parts of the uterus, it can be disorienting for surgical residents. We will have to take this into account when designing tasks for the hysteroscopic surgery simulator and train residents to better understand how this view affects their surgeries. In addition to simply viewing the inside of the uterus, it is important for surgeons to use external tools to remove growths from the inside lining. For this, there is a hollow shaft that runs the length of the optical chasm that allows surgeons to use scissors, cauterizers, and implant devices into the fallopian tubes. These tools will be a major factor in our design as they dictate what kinds of actions we will need the resident to carry out.

Finally, we must consider the cost of the hysteroscopes. Each scope costs around $20,000 and can be costly if they’re only being used for training. For this reason, Dr. Anderson has suggested that we create our own model hysteroscope to include with the hysteroscopic surgery simulator. This will take extensive work with metal machining and could potentially fall beyond the allotted time we have set for the project. We will continue to strive for a model scope but for the time being we will create the uterine model based on the actions carried out by the existing scopes.

 

Touch sensors – positives and negatives

We have considered many different methods for sensing touch within the model uterus. The three we have investigated most include capacitive touch sensors, force sensors, and simple momentary switches. Capacitive touch sensors have a simple binary response for touch, but they only respond to conductive materials. Force sensors respond to pressure from any material, but this response is continuous and has variable sensitivity. Finally, momentary switches inherently have a very well defined open and closed position, but these switches can take a significant amount of force to close.

For our first prototype, we will use momentary switches as targets for a diagnostic test. These switches will attach to our Arduino microcontroller in order to record successful interactions with the hysteroscope. In addition, LEDs will be used as indicators for the operator. If these switches work well, they will provide a cheap and easily reproducible source of targets. However, if they require too much force to hit and record a touch, we will experiment more with capacitive touch sensors.

 

Future plan

Our meeting with Dr. Anderson gave us more insight into the current clinical situation for hysteroscopy training. He stated that physical models, analogous to those readily available for laparoscopic surgery, do not exist for hysteroscopy. Dr. Anderson further identified two essential tasks he envisions for our model. The first task requires simple point touch and feedback. The second task requires sustained touch over a predetermined area of the model. Our immediate progress will seek to address these needs. By next week, using the sensors we have purchased and tested, we will have a point touch task that will assess operator efficiency and accuracy. No such metrics exist for hysteroscopy simulation training so we will work closely with Dr. Anderson to identify the needed skills and assessments. In parallel, we will investigate touch sensitive strips that can track sustained touch and directionality. Dr. Anderson also suggested improvements to our physical model of the uterus, including special paint to simulate the imaging properties of the uterine tissue and anatomical considerations such as the addition of holes for fallopian tubes. We will refine our model along these changes and continue to meet regularly with Dr. Anderson.