-
-
I have had a problem with my brain for the past few years now and about 3 years ago had a ‘Brain Pacemaker’ fitted – now without going into the gory details (check my facebook for the pictures J) this involves placing a couple of electrodes into the brain on the left and right sides. When I first had the implant I was the youngest in the world to have it but the 12th person overall. So research was lacking on exactly the right place to put the electrodes. I was convinced for some time they were in the wrong place.
Since having the pacemaker replaced 2 weeks ago the change in the way the stimulator works is dramatic – this time they hit the sweet spot and instead of feeling like someone was sticking needles in my head in the same place at the rate of 5 a second it feels like the whole of my head is being stimulated and this according to the neurologist is a good sign it will work well.
So my theory has been that because it is not an exact science or because people are different that placement of these electrodes needs a way better way of predicting where that sweet spot actually is. The device is not only used for my condition but also for people suffering from Parkinson’s, Tourette’s Syndrome, back pain and even to give women orgasms at the touch of a button (the last one is not available on the NHS)
So I am expanding my knowledge of SQL Server 2008 Spatial beyond the Geography data type – which alone has the potential to change the way we will navigate and visualize data in the upcoming years, to really understanding the Geometry data type and develop a spatial map of the brain. Now I do not actually intend to map the whole thing but there must be data in the public domain that already holds this information in some form of importable / convertable format that could be indexed within SQL Server.
The goal then would be to map the electrode placement information gathered from all patient information and layer that over the map of the brain. Immediate placement pattern will be obvious but to take it a stage further may would look to use data mining to predict best placement.
It’s a fun little project that I aim to develop to demonstrate using spatial data in different ways, so if anyone has data of this type that’s available in the public domain give me a shout. The actual patient data will be fake I just want to demonstrate the principal of using spatial mapping with the Geometry data type