Virtual Patient Simulation at U.S. and Canadian Medical Schools

Grace Huang, Robby Reynolds, Chris Candler
Academic Medicine Vol 82 No 5 May 2007- Educational Strategies

Interesting state of Virtual Patients (VPs) paper. They reviewed the "Virtual Patient" simulators of 108 schools. It seems most of them are doing an online flash/video approach. The point out that the cases do not generally exhibit ethnic diversity, so that may still be relatively unique to our system. The goal of the paper was to promote sharing of existing cases and future collaboration. Some of them present a longitudinal care of a patient, something our system is still far from providing. They also expose students to diseases and conditions they would otherwise miss, similarly to our future plans, but generally in a less immersive way. Since it was such a broad set of VPs they were nonspecific about what the systems actually consisted of, some of them may be as immersive as ours, though there were only 4 that were workstation based, they didn't mention VR at all. More than that, only 5 offered different outcomes based on user decisions. Though you can't kill our patient with bad decisions, you may misdiagnose.

Quality: 5
Relevance: 5

Since the primary point of the paper was collaboration, I was thinking, how could we promote collaboration with our system? It seems to me, creating a symbiotic online system to go with our VR system could lead to significant benefits. The primary one being, many more cases. I'm picturing a php or ASP.net based script creator. People could interact with the scripts using text only input, or even Windows based speech recognition. We'd basically need to reproduce simpy's text functions online, which would be a large work, but not infeasible. We could run a study on how the scripting tools are used, and, left to their own devices, what scripts the doctors really want. With the new feedback system for what questions are actually recognized, we could get a decent view of the quality of the scripts. All in all, it would give our system a much larger audience and that large distribution could give it a better chance of becoming the Xerox of Virtual Patients. Just an idea, though I'd like to hear feedback.

Words That Matter

Medical Writings
2001 American College of Physicians

Empathy is different than sympathy. Sympathy means that because someone feels an emotion, the sympathetic person feels it as well or feels a corresponding emotion, empathy is much more general. Empathy means the empathetic person understands the emotion of the other, and helps the other to know they are being understood. Empathy is very important for doctors to display and has been shown to lead to patient health benefits. One of the main ways for doctors to display empathy is to restate what the person has said, "Let me see if I have this right" or "I want to be sure I understand what you mean."

Our system needs to be able to catch these and similar phrases and give appropriate responses. It also talks about doctors learning patience, "don't just do something, stand there," meaning our patient should tell at least one long story. For the breast mass scene, I think when they first get around to the sister having breast cancer, she should tell the whole story, at least 30 seconds, if not longer. Then when the doctor says, "that must have been scary", she needs an appropriate response.

There are several other phrases indicated in the paper that really need to be added to the system. Right now, the scripts and the scripting system are a group effort. I think we will eventually need someone who's job it is to really make a great scripting system. Someone who is getting their PhD on NLP. But for now, we just need to add a bunch more empathy related responses and get it to recognize those responses correctly.

They also mention talking to someone with a different culture than the Doctor's own. This could be a very interesting scenario. I might like to investigate that in the future.

Quality: 5
Relevance: 5

Manual for the Defensive Verbal Behavior Ratings Scale (Version 3.0)

This is a manual on coding defensive verbal behavior, 25 pgs.

Paper can be found here: http://www2.bc.edu/~barretli/tools.shtml

"According to Psychotherapy of Neurotic Character (Shapiro, 1989), speech is an action that is motivated to serve some purpose." Defense means they are using the speech primarily to influence him/herself, instead of as a means to communicate with others. The paper is interesting, but does not reflect the kind of emotion from voice for which I am searching.

Quality: 4/5
Relevance: 1/5

Extracting Emotion from Voice

Authors: S. Fukuda, V. Kostov
Tokyo Metro. Institute of Technology

Emotions are gleaned mostly from voice, facial expressions, and body language. Humans have a fairly accurate ability to determine emotions from just voice. Humans distinguish four basic characteristics of sound: intensity, pitch, color, and direction. By analyzing Wavelets/Cepstrum the authors appear to have been able to get emotions from wav file inputs. The paper talks about the method they used, but isn't that specific about the accuracy of their method. They found that there is a difference between Japanese and English inputs, but weren't that specific about what the differences were. The paper concludes with saying they have taken a first step towards "Emotional Sensors", but that the field has a long way to go.

Quality: 3/5
Relevance: 5/5