Mechanical testing problems

The test is awkward to apply

The mouse is typically housed in a  rectangular enclosure with a mesh floor, with bars of about 1mm diameter and spaced about 5 mm apart. The mouse is free to move in all directions while the tester pokes the von Frey filament, or transducer probe, up between the bars and onto the plantar surface of the hind paw. As mice tend to grip the bars it can be awkward to access the same part of the paw each time (usually the middle of the pad where the “thumb” joins) and it is easy to accidentally hit the grid with the filament or probe. Also, if the cage is not at the right height, the operator must crouch or stretch to see and perform the test. There is often no support for the arms, and we’ve noticed that many people tend to hook their fingers over the edge of the mouse enclosure to steady themselves.
None of this is ideal for delicate, highly repetitive testing.

The “touch-on” force

The threshold forces are tiny; the threshold for healthy mice is generally 1-4gf but they will also sometimes respond to the very first touch of a probe or von Frey filament. We call this the “touch-on” force and we see it in mechanical testing of many species; data falls into two groups, those where the subject responded to the first touch and those where it ignored (or did not feel) it and allowed the force to rise to a true mechanical threshold before reacting.
We have found that, with von Frey filaments, the mouse might react sometimes at 0.4gf or even 0.16gf but then not again until 2g or 4g.
A reaction to “touch-on” is more likely if the force rises quickly, or if there is any sideways movement of the probe, leading it to slide or scratch across the plantar surface. Von Frey filaments avoid both these problems, being unable to apply more than their buckling force and being inherently “bendy” from side to side, thus adsorbing any hand tremor.

The effect of hand tremor

Unlike a von Frey filament, a probe mounted on an electronic force transducer tends to be stiff, or non-bendy, from side-to-side.  This means that, with the slightest amount of hand tremor, it’s more likely to slide  across the plantar surface and to elicit a touch-on response than a filament is.
A force transducer also tends to be stiff in the vertical direction. This means that any up-down hand tremor produces an instant and relatively large variation in the applied force. This doesn’t happen with von Frey filaments; once they have buckled, the force produced is pretty constant and so if your hand tremors during the 2-3 seconds that the filament is held in contact, the force remains the same.
Finally, it appears that the electronic von frey systems currently available are based around force transducers with a much higher force capacity than is required for mice (and sometimes for rats). Force ranges of 500gf are not uncommon, but not ideally suited for a measurement that rarely exceeds 5gf.

So why not stick to Von Frey filaments?

They are simple, relatively cheap and certainly well proven. Their main drawback is the number of measurements required to achieve a threshold reading. The up-down method requires 6 readings “around the threshold force” and, if starting with filaments that are not expected to produce a reaction, it is common to start counting at the first positive response (going down one filament and then taking four more measurements after that). If the first response is a spurious one (perhaps a touch-on) then the subsequent row of Os while you wait for the next (higher) positive can reduce the validity of the data.
There is also no doubt that mice become aware of the von Frey filaments and, if tested too much or too often, will avoid the filament by positioning themselves unhelpfully (perhaps with the relevant paw up the side of the enclosure) or just by reacting continually to touch-on. A test method which reduces the number of measurements is, therefore, desirable.
Finally, the stimulus applied by a von Frey filament is actually a rather complex one; although a flat, circular end is presented during the application of the force, this becomes an edge contact at the point at which the filament buckles. The effect of this is not yet understood (although readers might be interested in the poster abstract presented recently by Dr Michael Dixon at the WCVA conference in Capetown, concerning the effect of probe area on threshold force in conventional algometry).

(For information about our in-house validation system for both filament and EvF measurements, click on MouseCal. )


Topcat Metrology Ltd is a small, highly innovative company run by Dr Polly Taylor, an internationally known veterinary anaethestist and Dr Michael Dixon, an engineer with thirty years experience of difficult and unusual  measurement problems. We designed Mousemet from first principles, attempting to retain the many positive aspects of traditional von Frey filaments while removing the need for repetitive measurements and post test analysis. The result is a unique “soft” force transducer of an appropriate range for mice, software which allows quick assessment of the ramp rate and profile and a cage system that is ergonomically sound and maximises throughput. Click here to read more about it.