High accuracy distance/range/displacement sensor, short range

Hi tech-fans!

My research group and I are developing a machine to diagnose early stage knee osteoarthritis. It is essentially a inclined leg-press and we need to know the displacement of the plate you push. For this we need very high accuracy (preferrably <5 mm) over approximately 1 meter travel. Budget is around 150 £, but is flexible.

I have done a lot of searching for cheap common arduino sensors, but they all seem to have too low accuracy (Best i found was +- 2.5 cm, which is not nearly enough. I think it was the lidar).

I have also looked at industrial-grade laser sensors, which provide me the accuracy i require (ex. keyence IL-600), but i have a limited knowledge on the connective possibilities with these types and the arduino. Is is possible to use this kind of sensor with an arduino - or is there any simpler way to record output from this? Best case scenario would be to use matlab, if possible!

Any other ideas and poosibilities are most welcome!

Best Regards
Alex Dalsgaard

use parallax effect?

hav a laser close to the knee and shine from the side and create a "shadow of the knee on the wall
Move the laser with a linear motor, and the shadow will move larger distances than the laser.

Get the idea?

Where is your research group located?

Cheap incremental rotary encoder, add a belt/sprockets if needed to convert linear to rotary motion?

Perhaps a CNC aftermarket DRO electronic scale would be usable (more expensive, but
already designed for linear motion).

(Of course some people would regard "<5mm" as 4 to 6 orders of magnitude short of 'high accuracy'
when it comes to motion sensing!)

robtillaart:
hav a laser close to the knee and shine from the side and create a "shadow of the knee on the wall
Move the laser with a linear motor, and the shadow will move larger distances than the laser.

Where is your research group located?

Grretings Thank you for the answer! We are not measuring the knee itself but the machine, i a motor would just overcomplicate things - we just need the distance as input for a model. We are at Aalborg University in Denmark

MarkT:
Cheap incremental rotary encoder, add a belt/sprockets if needed to convert linear to rotary motion?

Perhaps a CNC aftermarket DRO electronic scale would be usable (more expensive, but
already designed for linear motion).

(Of course some people would regard "<5mm" as 4 to 6 orders of magnitude short of 'high accuracy'
when it comes to motion sensing!)

Thanks - i have no experience with these systems. Would that be a stand-alone, fully functioning system you are describing? As for the accuracy - around 1 mm would be optimal, lower than that would not make any difference when taking into consideration other errors such as tissue compression and such. However i would like to know more of the systems you are describing!

Best regards

Alex Dalsgaard