Today was the day the local newspaper and TV outlets (Teleamazonas, Gama TV, El Comercio, La Hora) were coming to Proteus to film our patient using pattern recognition with out myoelectric hand. Considering that the press were coming in at 9am, and we just got pattern recognition working directly on the hand this morning at 6am, we were more than a little anxious. Not to mention that we had been up for about 28 hours at that point. We had yet to get even good myoelectric signals coming from any patients yet due to EMG noise issues in the previous week. We had one shot to get it right when our patient, local Imam Juan Suquillo, came in around 9.
To be sure everything would work, I had hooked myself up to the hand around 8:30am. Luckily, everything was working! I was able to easily get the hand to recognize all 5 states (rest, open, close, three-jaw chuck, and pinch). Then the first interviewer, from the newspaper El Comercio, had come in to interview Mary and I. We were having some issues with communicating with the hand, but I was able to demonstrate on the computer the signals coming from my muscles and how the pattern recognition software we made could distinguish the different states of my hand.
When Juan came, all the media outlets were already present. Unfortunately, this meant that we didn't get a chance to test our device before the media arrived. While we were hooking him up to the system, Dave talked about ROMP and the significance of the project, and Juan told the story of how he lost his left hand. When all the EMG electrodes were placed and plugged into the breadboard, we noticed that the signals were all noise. This wasn't good, especially since we had yet to get the EMG working with a patient. After re-checking and moving things around in the circuit, still nothing was working. At that point, things were looking pretty bleak. Luckily, Dave and Juan kept the press busy while Mary and I troubleshooted the problem. We decided to check the voltages coming from the power supply and we noticed the values were off. It turns out that the batteries were hooked up wrong and were in fact running really hot and became swollen. Fortunately, we had other batteries on us, replaced them, and hooked them up correctly. Then, when we asked Juan to flex and extend his absent left wrist, we were able to see EMG signals appear on the computer! Things were finally starting to look up. At that point, we began to connect our prosthetic hand with the EMG signals coming from Juan's left arm. Given that when I had just tried to control the hand an hour prior it wasn't working, we were a little nervous about getting it to work now. We had Juan go through the training procedure that we had built into the hand. Once the training procedure completes, the hand opens and the user is free to control it. We asked Juan to close his hand in order to make the prosthesis close. It didn't. Something was wrong with the program on board the microcontroller. We decided to connect the microcontroller to the computer so we could reset the device and see the output of the signals going to the microcontroller. As we were doing this, being the local Imam, Juan turned to me and told me to "Say Bismillah," invoking the name of God. After we reset the device and Juan retrained through all the gestures, the hand opened again and was ready to accept his command. I asked Juan to close his hand, and this time, the prosthetic hand closed accordingly! A wave of awe and excitement (and relief for Mary, David, and I) spread throughout the audience. I looked at Juan and thought to myself, "you are a man of God." I asked Juan to open his hand and the hand opened. I then asked Juan to make a three-jaw chuck, though the hand mistakenly made a pinch. Though, since the number and location of the muscles that control the two movements are so similar, these two movements commonly get confused by our classifier. Plus, it looked like one of our EMG channels was dead, so even getting a three-jaw chuck to move as a pinch was a success.
After a couple minutes of controlling the hand, Juan was asked to comment on how he feels. At this point, all the media outlets moved in closer, placing their microphones on the desk in front of him to get his response. Juan was really moved by the ability to control a hand to do more than just open and close. And not just that, but by the possibility of deploying low-cost state-of-the-art prosthetic technology to masses. We unhooked Juan from our hand and then Mary and I were then interviewed. They asked us what we were studying and where we see this technology and our careers going. The reporter from Gama TV even mentioned that this was the coolest thing she has seen in a very long time, and that she's seen a lot of cool things.
After all was said and done, we Mary, Dave, and I were excited, relieved, and above all else, exhausted. Dave had only received 2 hours of sleep the night prior, and Mary and I had been up for around 32 hours at that point. However, we had one more demo to give at 3pm, and that was for the US Ambassador to Ecuador, Adam Namm, at the US Embassy. We grabbed a quick lunch at a Peruvian restaurant before we stopped by Dave's house on our way to the Embassy. We had to go through all the security drills again and set up our system in a conference room. We then got to take pictures with the Ambassador. We all walked to the room, where I had hooked up the electrodes to Adam and he had trained with the device. The control did not working that well, but he was able to open and close the hand. However, he was impressed with what we were able to accomplish, and how well it had worked with the patient. Gama TV had already sent their edited video to the US Embassy, and we all got a chance to watch it before it aired. Finally, after 36 hours of being awake and stressed, Mary and I went back to the hotel, after which we subsequently collapsed into a long-awaited, well-deserved, deep slumber.