A belt around the leg should be the best choice. In my opinion you should look at the different materials used for similar belts to attach small devices to users arms or legs.

One example I remember is Advanced Bionics Neptune (cochlear implant speech processor for under water usage). It's belt for attaching the processor to the users arm seems to work good.

http://www.advancedbionics.com/content/dam/ab/Global/en_ce/documents/recipient/AB_Neptune_brochure.pdf

Some of the sportive band for mp3 players also look promising.

Question: 1. Why exactly 20 cm below the hip (= suprailiac?) What is the rationale for this ?

2. Longitudinal accelerometer means antero-posterior axis?

1. Indeed belowe the sias and in anterior position. This is the position how the accelerometer works.

2. Sorry longitudinal will be the design I was to fast with my question.

I agree essentially with the reply of Daniel Polterauer...which I missed initially.

You can use the (attachement) sport stretch shorts with little pockets. I'm a sprinter (200m). I use to train and run (even with 36km/h) listening my mp3 sliped in it's pocket.

May I ask - How many active axes does your accelerometer have? And what kind of hip movements are you interested in?

Before properly answering your question, you should first explain why do you want to use an accelerometer and what do you want to measure ? Only then, could be discussed the place where to fix it on the body and how to do that.

activPAL use a hydrogel patch to attach their accelerometer based sensor to the thigh. You could also use the patches used to hold IV drip needles in place, there are multiple brand available, these can last around 3 day during normal everyday activities with a light sensor.

The accelerometer is a 3D accelerometer. The postures and motions categories (lying,sitting, standing, walking,cycling and running) are mainly determined by the angular position of the unit- the system is based on a position that is to some extent parallel to the front side of the trouser/thig and by the variability of the singnal which depends on the intensity of movement.

I want to measure the different postures and montions during the day. But the accelerometer had to by parallel tot the front side of the trouser/thig. So the place on the body is prescribed but the method how to wear not.

It sounds like the methods adopted by activPAL would suit you well http://www.paltech.plus.com/products.htm see also these papers attached for activity monitoring algorithms using thigh mounted sensors.

Sensors were located at the chest, thigh and shin.

An additional system with sensors at the chest and thigh.

Roderick, thank you for these details on your experimentations.

As far as you want to measure angular positions of a segment (e.g. thigh) in particular postures (e.g. lying, sitting, standing), I agree that a 3-D accelerometer will be useful, because you will use it as an inclinometer (i.e. you will use the measurement of gravity acceleration to compute different angles).

However, as soon as the subject will move (e.g. walking, cycling, running), the accelerometer will also measure the acceleration of the point where it is fixed on. As it integrates both a tangential and a centripetal components, this 3-D acceleration will thus depend on how far from the hip joint center is the accelerometer fixed. That is, you will measure different acceleration components for a same movement whether the 3-D accelerometer is fixed near the hip or near the knee.

But the main drawback for your purpose is that, in all cases where the subject is moving, it is physically and mathematically impossible to substract the tangential and centripetal components from the 3-D components of an acceleration that is only measured by a 3-D accelerometer. As a consequence, it is impossible to compute the angular position of the segment on which it is fixed.

Nevertheless, I wish you good luck for your experimentations!

In cases where we were not sure about the durability of the 'stickies' or double-adhesive tapes that are suggested for the ActviPal we decided to use OpSite Flexigrid surgical tape/film that is really thin and flexible and can even be used to seal the device against water (e.g. if you want to leave it on for taking a shower).

The 'belt-around-the-leg' approach did not really work well for us because of the conical shape of the thigh: The device either does not stay in place but slips down or you have to make the belt/strap so tight that it is uncomfortable or even reduce blood circulatiopn (both not really practical for long-term recording).

Thanks a lot for the respons!