The big payoff
Yesterday I finally had the chance to get not one, but two new datasets! Those datasets were for a project I’m calling “big idea.” In my mind, it will change the way a lot of things are done, advance science, and I even joked with hospital-PI that I would need to retire after we publish because I’ll never do anything better. I have a lot of hope for big idea, but the first dataset we got was a fail, so the question was, what happened and did we fix it?
The problem with collecting neurological data is that you can very rarely tell if you’re getting something good. When recording EMG, no problem! When you’re doing a lot of averaging with nicely time locked data, forget about it! But when you’re doing very complex things that have no clear start or stop point, well then you have yourself a gift. It’s a gift because until you figure out what you have, you have no idea what you got.
As an aside before we start, time locked data means we have a clear start, usually this is an electrical stimulus or some sort of visual/auditory cue. Basically we know exactly when we sent the command to the person. Normally in our lab we use electrical stimuli and the equipment will put a little flag in our data to let me know exactly when it was delivered. It’s really cool because you can do 1000’s of repetitions and have the exact start point for each of them.
Now yesterday we collected two different datasets, in two different settings, with two very different sets of tasks. There’s still a lot to be done for both of them, but I have some good news, at least for now. It looks (fingers crossed) like we have good data… at least for one of the datasets. Obviously I can’t say too much about the stuff I’m doing, but I can talk about the good and not so good.
The good is that the data look like we expected. These datasets were to verify everything was working (in a roundabout kind of way), the datasets aren’t my first choice for using big idea. These were more “cherry on top” datasets for hospital-PI. As he puts it, we want to verify the stuff is doing like we expect. We have other data to compare our work to before we get started with how I want to use it. The real (actual) good news is that the data match.
Now the bad news. There was more noise than I would’ve liked in the data. There’s a lot of reasons this could be, electrical equipment is everywhere in the hospital. We use the incandescent lights that you find in a lot of buildings and those pump out electrical interference like it’s was designed to screw with electronics. Oh and there’s the not so ideal way we did the work, we had to run the wires on the ground, which was not good.. Running wires on the ground is by far the worse thing you can do when you’re working with sensitive equipment.
All in all, I can do better, I know I can. I just need more time with people and that hasn’t been possible in any of these cases. I am hopeful that now that we have some good data we can arrange for something more long term or at least a longer experiment. Lately we’ve gotten just a few hours if that to do the work and it’s not ideal. So overall I’m very happy with the outcome so far and I’ve only got to scratch the surface of the data.
Small victories. I mean I knew it would work, but getting the equipment working was the tricky part. In fact, we had to set everything up in a way that didn’t exactly make sense to me and I’m not sure what’s going on. Yet another reason why I really need to get someone to work work with for a longer period of time. I need to try a few different setups to nail down which works and to figure out why the setup I want to use doesn’t appear to work.