Phidget to analyse movement

[IanBloomfield]

Hi all,

I am looking to perform a study analyzing movement of surgeons hands whilst performing laparoscopic surgical skills. I thought the accelerometer sensor would be useful for this?

I really want to compare how different skilled operators move when they perform the same task - I thought a sensor attached to back of each hand as I also want to demonstrate the difference between the dominant and non dominant hand.

Would this be something the phidget device is capable of? how complex is it for me to acquire this data I have android phone or PC? I don't have much experience with this technology but I can learn very quickly if anyone is able to help me.

THanks
Ian Bloomfield

[mparadis]

The tricky thing about using Phidgets in this application is that they must always be connected to a computer via USB cable. I'd be worried that having an SBC and two Spatial Phidgets strapped to the surgeon's body might get in the way of the surgery. If you know someone who's into electronics, they might be able to build you a more compact wearable solution using accelerometer chips and a portable battery.

Have you considered using a high-resolution camera and color tracking software to analyze the surgeon's hand movements (feasible if they use the blue-colored gloves)? It would probably require more sophisticated programming in order to get meaningful results, but it would also be completely non-intrusive for the surgeon.

[IanBloomfield]

For this initial experiment we won't actually be performing surgery, merely specific tasks within a simulator. I need to have a few different surgeons perform the same task so as to have an idea of the movement waveform. My hope is to establish a reproducible method to determine that a training surgeon has acquired the necessary dexterity and prescision of movement to justify learning more complicated tasks.
If taking this to the operating room, I had seen Bluetooth accelerometers - but it takes a lot of ethical approval and testing to get a device into the operating field, hence I'm not fussed about the wires, I merely want something that will work and not too costly as it's all self funded. Presumeably I would need to acquire some software to collect and interpret the data or are there programs out there that will collect / calibrate this?

[IanBloomfield]

Would the accelerometer be the ideal sensor or the spatial phidget? I have seen a study where someone used the accelerometers in their Iphone, literally strapped to the wrist. My plan was to reproduce this but with smaller less imposing sensors plus incorporating the head movement?

I don't know between the sensors nor what a VINT hub is?

[mparadis]

If the cables aren't a problem then you could use Phidgets with the following parts:

● MOT1101_0 - Spatial Phidget x2 or x3
● HUB0000_0 - VINT Hub Phidget x1
● Phidget Cable x2 or x3
● Mini-USB Cable x1

You would attach one MOT1101 to each hand, or whatever other points you would want to monitor the movement of. You could take the board out of the enclosure if you wanted it to be less bulky on the hands. Then the Phidget cables would run through the sleeves and out the back, where they would connect to the VINT Hub, which is the board that gathers and sends all of the data to your computer. Since this is a test environment, you could probably have a laptop computer on a nearby table to plug the VINT hub into. Then you would start your program and begin collecting data.

You'll need to know at least a little bit of programming to use Phidgets- or know someone who can write you a quick program. Probably all you'd need to do is modify our examples to output the data to a spreadsheet where you can process it later.

The difference between our Accelerometer Phidget and our Spatial Phidget is that the former has a 3-axis accelerometer, and the latter has a 3-axis accelerometer, gyroscope, and magnetometer. If you have the proper math, these extra two sensors will help paint a better picture of what movement is happening.

[IanBloomfield]

Thanks,

thats really helpful advice, I plan to do a test run with the sensors in my phone then will look at progressing on with the devices. The spatial's are cheaper than the 3axis accelerometer? I can't get my head round that since they record more data? are the sensors sensitive for small movements like hand rotation and sliding in and out?

kind of like rotation for driving a screwdriver or something?
I have a whole year to do this project so can either figure the programming myself - it's the sort of challenge I like, else we have a uni up road I'm sure I could find an IT student willing to help out.

Ian

[mparadis]

The reason some of the spatials are less expensive than some of the accelerometers is because our newer spatials and accelerometers (the ones that connect to a VINT hub) are more cost effective to produce than the older ones.

They are sensitive enough to detect small movements. It may be difficult to make any use of the data (for example, translating a small rotation and tilt into a specific hand movement), but the data will be there.