Dear Biomch-L readers

This is a summary of the replies I received regarding INTERACTIVE SPLINES.
Many thanks to those who replied, and sincere appologies about the delay in
posting. The original posting read

>For data smoothing, differentiation and interpolation purposes, (cubic-)
>splines are often fitted to raw data. On a number of occasions I have found
>reference to INTERACTIVE spline routines (for example: McCaw and DeVita,
>1995). In the near future I plan to down load and utilise the 'generalised
>cross validated spline' routines made available by Herman Woltring.
> ARE WOLTRIN'S SPLINE ROUTINES INTERACTIVE?

>If not:

> ARE THERE ANY ADVANTAGE IN EMPLOYING INTERACTIVE SPLINE
> ROUTINES?

>and if there are any advantages:

> DOES ANYONE KNOW WHERE I COULD GET 'FREE' SOFTWARE FOR AN
> INTERACTIVE (CUBIC-) SPLINE ROUTINE.

>As always I will post a summary of all responses.

>........................................
>McCaw, S. and DeVita, (1995) Errors in the alignment of centre of pressure
>and foot coordinates affect the predicted lower extremity torques. J.
>Biomechanics, V28(8): 985-988
>.......................................

Reply 1:
Kieran:

I would advise you very strongly to stay away from CUBIC splines.
Usually (always?) they force the second derivative to be zero at the
beginning and end of the data set. When that is not the case in the
activity being analyzed (i.e., almost always), the result is very distorted
data in the early and late parts of the trial (and even the middle parts can
get messed up too).


Quintic spline does not have this problem. (I believe that the
FOURTH derivative is what is forced to be zero when you use quintic spline,
and this does not pose a problem at all.)


The differences between cubic and quintic spline was shown very
clearly in Kit Vaughan's Ph.D. dissertation (University of Iowa, ca. 1981)
when he fit cubic and quintic splines to the vertical location data of a
free-falling ball. The quintic gave a beautiful fit, with acceleration very
near 9.81 m/s2 throughout the entire airborne period. The cubic forced
acceleration to be zero at the beginning and end of the data set, the curve
then overshot the 9.81 m/s2 value in the middle part of the trial; the whole
curve was distorted, so that acceleration was not near 9.81 m/s2 hardly
anywhere at all.


I have been using splines for about twenty years now, and my
experience tells me that quintic spline is excellent, but you don't want to
touch cubic spline with a ten foot pole!

I see that you are aware of Woltring's program. You may also want
to check Les Jennings' quintic spline program -- this is the one that we
routinely use. Jennings' email address at Western Australia is:
les@maths.uwa.oz.au

Unfortunately, I don't know who has "interactive" quintic spline
subroutines. (in fact I'm not sure what the hell they are!)

---
Jesus Dapena
Department of Kinesiology
Indiana University
Bloomington, IN 47405, USA
1-812-855-8407 (office phone)
dapena@valeri.hper.indiana.edu (email)
.................................................. ................


Reply 2

Dear Kieran,
the advantage of using splines or any other kind of algorithms to smooth
your data is that you suspect your data. In human movement we are very lucky
(comparing with other areas/domains) because our knowledge of anatomy
give us several constraints that we can take into account to avoid unrealistic
situations. If we will add to these constraints the mechanical (Newtonian
laws, etc.) plus the calculus derivatives tests, we have several other
information that we can "trust" and use. Since you are really into these
stuff I suggest you the following readings:
Wood, G.A. (1982) Exercise and Sport Science Reviews, 10: 308-362.
Hatze, H. (1983) Computerized....developments, J of Sport Sciences, 1:3-12
Woltring, H.J. (1985) Human Mov Science, 4: 229-245.
Jackson, K.M. (1979). IEEE trans, on Biomedical Engineering, 26: 122-124.
Feel free to write me back with any questions.

Nick Stergiou, Ph.D.
Biomechanics Lab
Univ. of Oregon
.................................................. ..................


Reply 3

I read your message on the Biomech-l concerning interactive
splines. The term interactive spline simply means that you
interact with the program to set the spline windows based on what
your data looks like as opposed to an automated technique that
has the computer set the windows based on a predetermined
criterion. I have the routines that Steve McCaw and Paul Devita
use to spline data because I translated them and gave Steve the
routine while we were at Oregon together. The Sub routine is
fairly straight forward and robust, and run in Quick Basic. If
you want it I will try to figure out a way to get it to you even
if it means sending it via the postal service. Let me know.


Steve Ingram (_)
KRUG Life Science ,JSC, NASA +
SINGRAM@sdpcmail.jsc.nasa.gov / | \/
\ |___
/ /
/ /__ ____________________
____________________________ /_ _______/
/ \/
__________________________/__/_______
/____________________
................End of message...........End of message.................

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% KIERAN MORAN
%
% EMAIL: KA.MORAN@ULST.AC.UK
% TELEPHONE: 0044 1232 368997
% FAX: 0044 1232 366863
%
% BIOMECHANICS SECTION
% NORTHERN IRELAND BIOENGINEERING CENTRE
% UNIVERSITY OF ULSTER
% NEWTOWNABBEY
% BT37 OQB
% NORTHERN IRELAND
%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%