I'm sorry but it is not my field

I have developed a variety of objective metrics of motor control (some links below to papers in open access format and if you need others in traditional journals I can send you pdfs but these are a good starting point).

These biometrics iinclude determining for each individual the signatures of variability (speed and acceleration dependent) from motion trajectories to goals (goal directed deliberate components) and also from spontaneous movements that do not pursue a goal or are not instructed in any explicit way. 

You can use wearable sensors such as inertial measurement units that register gyroscopic motions, acceleration, temperature related signals, electrodermal activity, heartbeat, etc. Basically any continuous time series of a fluctuating parameter will give you a sense for the stochastic signatures of the person and their rate of change over time. This information is gold.

even cell phones accelerometers will give you some useful biometric

there is a lot of noise in those signals though, so I recommend some serious research grain wearables with high sampling resolution, good battery life and reliable wireless capability. Then you can apply these biometrics

I have never done CP but the principles are the same. Obtain the signatures of motion output variability where motion is meant in a very broad sense as the rate of change of a parameter over time. Take its first rate of change for a velocity dependent signal (very useful) or the second rate of change for an accel dependent signal (very useful but careful with the noise at that point) Check out signal processing sites to deal with various filters, etc.

hope that this helps 

http://journal.frontiersin.org/article/10.3389/fnhum.2015.00128/abstract

http://journal.frontiersin.org/article/10.3389/fnhum.2014.00823/abstract

http://journal.frontiersin.org/article/10.3389/fnint.2013.00032/abstract

http://journal.frontiersin.org/article/10.3389/fnint.2013.00046/abstract

http://journal.frontiersin.org/article/10.3389/fnint.2013.00050/abstract

What type of CP are you investigating? Are you asking about upper or lower extremity? Are you looking at adults or children? I don't know if moving a joystick would be a functional task for most with cerebral palsy.

For UE, We use the Assisting Hand Assessment (AHA), but you need to be certified. 

Hi Nicholas, 

I have to say I am biased as I have spent quite a bit of time developing upper extremity assessments. From all our work it the field it I came to the conclusion you can break the field of assessing motor control into two streams. One is to look at the movement, and the other is to look a functional outcomes. If your interest is Upper Extremity motor function I would advise you to consider functional outcomes.  In CP I have seen various people try to use validated standardized stroke assessments (like the Fugl Meyer Assessment) http://www.archives-pmr.org/article/S0003-9993%2809%2900484-5/abstract ,

In Dr. Prochazka's Lab we were first to automate a workstation to assess hand function, this followed by a number of interesting developments in the field such as QUEST on the robotic side developed by the Hogan group, or the AUTOCite by Peter Lum and Dr. Taub .

• http://www.ncbi.nlm.nih.gov/pubmed/18936558
• http://www.ncbi.nlm.nih.gov/pubmed/15543442

My personal input into this field is that I am not a huge proponent of biometric assessments (administration of kinematic data is not easy to standardize across sites, can take quite a bit of setup time per patient, as well as clinical uptake seems very slow ).  Dr. Torres above makes very good suggestions but these types of assessments are typically poorly related to functional outcomes in stroke and CP (just because a stroke patient can move significantly faster, does not mean they can perform functional tasks better – this is clear in upper extremity not so in the lower extremity).

We have worked on  upper extremity functional assessments and developed the RAHFT (Rejoyce Automated Hand Function Test). It is a validated upper extremity function test in Stoke and SCI.

• https://www.researchgate.net/publication/270656431_A_Fully_Automated_Quantitative_Test_of_Upper_Limb_Function
• https://www.researchgate.net/publication/272476724_Tele-Supervised_FES-Assisted_Exercise_for_Hemiplegic_Upper_Limb
• https://www.researchgate.net/publication/221757904_Fully-automated_test_of_upper-extremity_function?ev=prf_pub
• https://www.researchgate.net/publication/50270975_In-Home_Tele-Rehabilitation_Improves_Tetraplegic_Hand_Function

The RAHFT is entirely assessed by the software on a unique joystick specifically designed to replicate ADL's ( Activities of Daily Life) and as a result it does not suffer from typical subjective clinical assessments concerns (it is fully objective). These types of tests are very meaningful for assessing ADLs. What they lack however is very high specificity (Lets take an example of how well someone can write the letter a no tests exists to assess such a specific task). In all cases the biggest tradeoff for these types of tests is how much can be assessed in the time provided (a big concern is that patients get tired during the course of the test so they are more prone to fatigue playing a role – this is a serious concerns if your patients have spasticity). Many of these tests are designed to fit into a quick clinical assessment that can give you a snapshot into the users abilities, unlike wearable sensors that might track the persons movements over a long time.

In the end I think you will need to ask yourself if you are more interested in understanding how functional a person is or how normal their movements are.

Hope this helps, 

Jan

Article A Fully Automated, Quantitative Test of Upper Limb Function

Article Tele-Supervised FES-Assisted Exercise for Hemiplegic Upper Limb

Article Fully-automated test of upper-extremity function

Article In-Home Tele-Rehabilitation Improves Tetraplegic Hand Function

here are some examples of a mathematical model to examine upper extremity motions directed to a general goal

and applying those concepts to stroke and PD so as to identify the best source of sensory guidance before attempting therapies that could make things worse for the patient

and we also have examples involving  children in the spectrum of autism who cannot speak as they co-adapt their motions with those of an external video that they spontaneously learn to tigger

In the latter case we sensory-substituted corrupted motor output variability by visual and auditory inputs and were able to characterize each child's plasticity and flexibility to learn the process of self-discovery. This ended up increasing their self awareness and gave them a boost of confidence (those were our functional goals in that case)

Today we are working with horse therapy in an occupational therapy setting that includes children with CP.

We have never worked with children with CP but the OT's that we work with are experts and we are working together to develop the proper outcome measures for insurance coverage

And BTW this work that I posted before has nothing to do with whether the person moves slow or fast

that statement by Jan was actually kind of funny

http://jn.physiology.org/content/jn/88/5/2355.full.pdf

http://www.jneurosci.org/content/31/49/17848.full.pdf+html

Hi Elizabeth, 

I did not mean to imply that the publications you had presented had anything to do with assessing speed, I just wanted to provide a quick example to say that if you are interested in upper extremity functional outcomes (or clinically relevant outcomes) it is very hard to correlate those to biometric measures in stroke or CP. Maybe the example I should have used was a CP patient with a very nice movement trajectory could be much more impaired (as they don't have any functional grasp) than a CP patient with a poor quality movement trajectory (as they might have better control over grasping various objects). That is why I encouraged assessing function on tasks of daily life as those give a much better glimpse of the persons functional impairment. That is not so say I am opposed to looking at kinematic data at all (in fact we do quite a bit of it) , but that will give you a better understanding of the quality of movement in stroke or CP not the functional impairment. Both are important to look at, but it will all depend on the research question being asked. To add a little bit more, most functional tests tend to assess the test's reliability quite a bit (inter-rater, intra rater) , it is a very important aspect of a clinical test for impairment standardization, that is not always the case when assessing movement quality. Hope this clears things up a little.

Sincerely,  

I am an occupational therapist and also work frequently with computer engineers. I agree with Jan that functional outcomes are what is needed from a clinical standpoint. Biometrics, while nice for research, do not translate to children and their families, or payers for that matter. They want to see changes in function to provide more independence and participation.  That is really the bottom line of what we (OT) do and it is what is supported by the World Health Organization.

My methods do not perform the traditional kinematics analyses that you are referring to, nor they correlate X to Y and provide traditional statistical assessments

The word biometrics acquires a new meaning under our new individualized statistical platform but you would have to spend some time reading the papers before dismissing the work

we have spent a great deal of time translating the basic research to the real context of OT

trying to explain the math in this forum would not do justice to the new technology that we have invented and used to include analyses of functional goals among other issues

Angie, several organizations in the OT world are working with my lab to use our biometrics for the assessment of function, effectiveness and risk among other factors

I have given recent talks in the NJ OTA and I am scheduled to give talks at the Pennsylvania meeting in the Fall as well. I am also collaborating with the Sensory Processing Disorder Center in Colorado (directed by Lucy Miller) and with OT's around the country who specialize on pediatrics and sensory-motor integration disorders, including autism which is at present not recognized as such

We did an NSF funded market analyses of the critical needs for our biometrics.

After 117 interviews OT's and PT's across the US identified our statistical platform as filling their critical need for outcome measures to report to insurance companies

(see the links as we made short videos to describe our findings in this market research)

The reason for that is that we provide a way to track changes in real time within the clinical visit time frame and then longitudinally as the person undergoes treatment, including in the home settings

The math behind the new technology is beyond the scope of this forum but the papers are detailed in methods and applications to various populations (autism, PD, stroke, deafferentation and schizophrenia, among others)

If you have an interest in further exploring the new technology, please do not hesitate to ask me, but dismissing it without actually understanding what the new statistical platform does, would be a shame

Since our original publication with a focus on autism we have over 20K views of our papers combined and over 146K views of our special issue related to the work

There is a world wide interest in what we are doing because it is fundamentally different from any other methods that assess kinematics or time series data

The challenge will be to define the functional goals for each set of interventions in each population, but that is a separate issue from having a statistical platform that enables real time monitoring of risk and efficacy with long term gains and generalization.

The definition of functional goals are specific to context and task and we have already covered several areas in basic research. At present we are working with the NJ OTA to adapt the basic research results to the OT gym and open therapeutic areas, including hippo-therapy 

https://www.youtube.com/watch?v=kAYBOlEGL1w

https://www.youtube.com/watch?v=IWG7SxbvtJ4