PDA

View Full Version : Summary (X-Ray digitization)



Alexandre Templier
08-02-1996, 01:07 AM
Dear Biomch-L colleagues,

Many thanks to those who replied to my request concerning X-ray digitizers and
related Software ...

Original message :
>I'm searching for data on radiographic pictures numerization.
>My purpose is to scan classical X-ray pictures with a sufficient resolution to
>be able to Zoom in and to get accurate 2D coordinates of anatomic landmarks.
>Could anyone indicate me some manufacturers adresses or product references,
>concerning available hardware and software dedicated to Computer-Assisted
>standard radiographic pictures working ?

Here are the numerous responses I got :

__________________________________________________ __________________
__________________________________________________ __________________

1 )from Chris Kirtley

Have you tried NIH-Image? It's freeware at:

http://rsb.info.nih.gov/nih-image/

Best wishes,

Chris

__________________________________________________ __________________
Dr. Chris Kirtley MB ChB, PhD c.kirtley@info.curtin.edu.au
^
Lecturer, Bio-engineering --_ / \
/ \
School of Physiotherapy, Perth #_.---._/
Curtin University of Technology, V
GPO Box U1987,
Perth 6001, Tel +61 9 351 3649
Western Australia. Fax +61 9 351 3636

WWW: http://www.curtin.edu.au/curtin/dept/physio/pt/staff/kirtley
Clinical Gait Analysis: website with clinical cases, data & FAQ:
http://www.curtin.edu.au/curtin/dept/physio/pt/staff/kirtley/cga
Internet Relay Chat: irc.curtin.edu.au (port 6667) "bio-engineering"
__________________________________________________ __________________
__________________________________________________ __________________

2) From Fabio Baruffaldi

Just two of the most famous manufacters of x-ray scanners:

Vidar Systems Corp.
460 Spring Park Place
Herndon, VA 22070
USA
tel +1-703-4717070
fax +1-703-4711165


Lumisys Inc.
PO Box 431
3700 AK Zeist
The Netherlands
tel +31-3404-24721
fax +31-3404-11973


Hope this helps. Have a good research.

------------------------------------------------------------------------------
Fabio Baruffaldi
Laboratorio di Tecnologia dei Materiali tel. 0039-51-6366864
Istituti Ortopedici Rizzoli
via di barbiano 1/10, 40136 - Bologna, Italy fax. 0039-51-6366863

E-mail: fabio@tecno.ior.it
------------------------------------------------------------------------------
__________________________________________________ __________________
__________________________________________________ __________________

3) From R A Hillam

Dear Alexandre,

In reply to your request regarding digitisation of radiographs:

I believe there is an international standard and special high frequency
light boxes are required. The Adobe photoshop software may be adequate but I
have a contact that is very interested in this field with veterianry
radiographs. His details are as following:

Mr Alastair Nelson MRCVS
Ordidge and Pritchard Veterinary Clinic
Rainbow Farm
Old Malton
Malton
N Yorkshire
YO17 OSG
UK

His email address is:
100307.143@compuserve.com

His fax is:
44-(0) 1653-600319

Good luck

Richard Hillam
Dept Anatomy
School of Veterinary Science
University of Bristol
Southwell Street
Bristol BS2 8EJ
UK


__________________________________________________ __________________
__________________________________________________ __________________

4) From David Dean Ph.D.

Dear Dr. Templier

We are doing just that work here. I would call your
attention to the Bolton-Brush Growth Study Web page if
you have access to the WWW:

http://www.cwru.edu/CWRU/Dept/Dent/bbgsc/homepage.html

We are using a Howtek scanner. We have also developed
softward to correct grayscale degredation, i.e., make
old films new again, automatically detect landmark
locations, and compress the images 20:1 with no loss
in accuracy. We have made tests to make sure our
measurements are as good as hand collected data. The
scanned images have 50 micron resolution, so this is
to be expected. One additional step we take is to use
a template to punch holes into the film at specified
distances. These distances insure that the scan was
accurate.

I would be happy to discuss this further.

BTW I have a number of colleagues in Paris and at
INRIA in Nice. I will see my INRIA colleagues when we
present together at Visualization in Biomedical
Computing '96 in Hamburg Sept 22-25.

Good luck.

Sincerely Yours,
David Dean

+---------------------------------------------------------------------+
| David Dean, Ph.D., Assistant Professor |
| Department of Anatomy |
| Case Western Reserve University Voice: (216) 368-1975 |
| 10900 Euclid Avenue FAX: (216) 368-3204 |
| Cleveland, OH 44106-4930 USA E-mail: dean@lucifer.cwru.edu|
| WWW: http://scourge.cwru.edu/morphmet/index.html |
+---------------------------------------------------------------------+
__________________________________________________ __________________
__________________________________________________ __________________

5)From Andrew Milne

Try talking to Jandel Scientific about their program Sigma Scan for
use with a back light flat bed scanner. It has pretty good
resolution and the software is capable of many automated measuring
functions.

European address
Jandel GmbH
40699 Erkrath, Germany

phone: 49 2104 / 9540


Good luck

Andrew D. Milne, P.Eng.
Orthopaedic Research Engineer
HULC Musculoskeletal Research Laboratory
St. Joseph's Health Centre
London, Ontario, Canada
(519) 646-6100 x 5279

__________________________________________________ __________________
__________________________________________________ __________________

6)From Jose Luis Peris Serra

Dear Alexandre,

Greetings from Valencia.

In relation to your request, we have made just a few attempts with
an image analysis system VISILOG (France), trying to detect accurate
object boundaries from classical X-ray pictures and analogical and
digital CT scans.
We have not experience with spine images but probably the problems
could be similar. The main problem is, of course, the quality of the
original image and it depends on the X-ray protocol.
As you know, basically, we use VISILOG software to analyse histomorphometrical
parameters from undecalcified bone samples of fracture callus. But,
sometimes we have had to deal with other image analisys problems.

Do not hesitate to contact with me if you have any doubt about image
analisys although I know that in France you could find expertises on
this subject.


__________________________________________________ __________________
__________________________________________________ __________________

7) From Jim Patton

I've had reasonable success scanning biomechanical illustrations and
digitizing them using MATLAB. I would imagine rdiographs would be similar.
If you already have this program, it is an easy and cheap way to get what
you want. An add on package for MATLAB, the IMAGE PROCESSING TOOLBOX,
provides some excellent tools, but is not necessary. Let me know if you
want more information or my programs.

\\|//
() (o-o)
/\ +---o00-----(_)-----00o--+
Jim Patton_/) .oooO Oooo. | __ /\ \ |
/\ ( ) ( ) | /\ \ /::\ \ |
/ \ \ ( ) / | __ \:\ \ /:/\:\ \ |
_\ _\ \_) (_/ | /\_\ \:\ \ /::\ \:\_\|
|/:/_/ /::\ \ /:/\:\/:/ /|
Doctoral Candidate, 645 N Michigan Av |\:\ \/:/\:\_\:/ /\::/ / |
Biomedical Engineering Suite 1100 | \:\/:/ /\/_// / \/_/ |
Northwestern U. Chicago, IL 60611 | \::/ / \/_/ |
EMAIL: j-patton@nwu.edu (312)908-6785LAB | \/_/ |
http://sulu.smpp.nwu.edu/~jim -0741FAX +------------------------+
COOL TALK: JLP@165.124.9.23 (_) (_)


__________________________________________________ __________________
__________________________________________________ __________________

8) From Andrew Pinder

Alexandre

There are various sonic digitisers available which allow you to point a
stylus at an image / x-ray and get X/Y coordinates to 0.01 mm or so.
The problems would be the time required to digitise the points (depends
how many points and how many images) and the difficulty of identifying
the correct landmarks in the x-rays.

You could scan the x-rays on a flat-bed scanner, display them on a
computer screen and use the mouse to click on landmarks and return the
screen coordinates. Resolution of the scanner and the computer screen /
mouse would be limiting factors. You would also require a powerful
computer and large amounts of disc space to get good accuracy.

Unfortunately, I can't give you any information on available harware or
software.

Andrew Pinder
Ergonomics and Work Psychology Section
Health and Safety Laboratory
Broad Lane
Sheffield S3 7HQ
UK

Phone: + 44 114 289 2594
Fax: + 44 114 289 2526
email: apinder@ewps-hsl.demon.co.uk

__________________________________________________ __________________
__________________________________________________ __________________

9) From Douglas Chang

Hi, try this:

Panjabi M; Chang D; Dvorak J.
An analysis of errors in kinematic parameters associated with in vivo
functional radiographs.
Spine, 1992 Feb, 17(2):200-5.



Good Luck!
doug


__________________________________________________ _________
__________________________________________________ _________

10) From Santosh Zachariah

Alexandre,

we use NIH-Image, a public domain software package for the Apple
Macintosh. It is available from an ftp site. That, and a good
flat-bed scanner will do everything you need.


-- Santosh Zachariah --

Center for Bioengineering [ph:usa (206) 685 3488]
Univ. of Washington, Box 35225 [fax: 543 6124]
Seattle, WA 98195, USA zach@limbs.bioeng.washington.edu

__________________________________________________ _______________
__________________________________________________ _______________

11) From ORL - Morra, Ed

Alexandre,

Contact Ortho-Graphics, Inc. from Salt Lake City, Utah USA (801) 521-4202. They
make stand alone scanners for radiographs (high resolution and 12 bit
grayscale).

- Ed -

|~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~|
| Edward Morra, MSME Mount Sinai Medical Center |
| ed@orl-inc.com Orthopaedic Research Laboratories |
| http://orl-inc.com One Mount Sinai Drive |
| (216) 421-4697 vox Cleveland, Ohio 44106 |
| (216) 421-4843 fax USA |
|_________________________________________________ ____________|

12) From Neil Glossop

Salut Alexandre,
There are cheap and expensive ways to do this. We just used a video
camera pointed at a light box, but we knew what we were interested in
and zoomed in on that. We wrote about it in J hand surgery: I can
get the ref if you want.

Now adays there are several companies that sell high res cameras just
to do this in an effort to digitise the old x-rays. They are
expensive and end up turning your data to DICOM or a strange
radiology standard. Most big companies sell something for this (e.g.
Phillips, GE etc.) I think. Sorry I cant help more.

Neil

Neil Glossop
Houston, Tex. USA

__________________________________________________ ________________
__________________________________________________ ________________

THANKS AGAIN TO EVERYONE ...

Alexandre TEMPLIER

Laboratoire de Biomecanique
Ecole Nationale Superieure d'Arts & Metiers
151 Boulevard de l'Hopital
75013 Paris
France

E-mail : alexandre.templier@paris.ensam.fr
tel : (33).1.44.24.63.65
fax : (33).1.44.24.63.66