Dear Image J subscribers,
we are monitoring a clinical study where Image J software is used to determine the % of specifically stained area in pictures of tissue cuts. Software used in clinical studies should be validated. Is there a validation report or other proof of performance available for the Image J software. Any help is extremely welcome. Thank you and best regards -- Dr.Konstantin von Martius Quintiles Consulting Zuccalistr. 19 80639 München Tel: +49 89 176782 Fax: +49 89 176773 Email: [hidden email] *************************************************************************************** Quintiles GmbH, Hugenottenallee 167,D-63263 Neu-Isenburg, Deutschland Geschäftsführer: Dr. Ludger Beckmann, Michael Wilson Registergericht: Offenbach/M., GmbH, HRB 7983 *************************************************************************************** |
Dear Dr. von Martius,
You might check the archives https://list.nih.gov/archives/imagej.html. Search terms "certification" and "validation" produce several (overlapping) hits. Bob Robert P. Dougherty, Ph.D. President, OptiNav, Inc. Phone (425) 990-5912 Fax (425) 467-1119 www.optinav.com > -----Original Message----- > From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of > Konstantin von Martius > Sent: Wednesday, February 13, 2008 6:59 AM > To: [hidden email] > Subject: Validation > > Dear Image J subscribers, > we are monitoring a clinical study where Image J software is used to > determine the % of specifically stained area in pictures of tissue > cuts. Software used in clinical studies should be validated. Is there a > validation report or other proof of performance available for the Image > J software. > Any help is extremely welcome. > Thank you and best regards > > -- > Dr.Konstantin von Martius > Quintiles Consulting > Zuccalistr. 19 > 80639 München > Tel: +49 89 176782 > Fax: +49 89 176773 > Email: [hidden email] > > ************************************************************************** > ************* > Quintiles GmbH, Hugenottenallee 167,D-63263 Neu-Isenburg, Deutschland > Geschäftsführer: Dr. Ludger Beckmann, Michael Wilson > Registergericht: Offenbach/M., GmbH, HRB 7983 > > ************************************************************************** > ************* |
In reply to this post by Konstantin von Martius
Konstantin,
The part of the software industry that need reliable results, e.g. finance, is moving towards extensive use of unit tests, and often test driven development. You don't add new code without new unit tests and you don't check in code without running all unit tests. In my little ImageJ project ShapeLogic, I currently have 320 unit tests. It takes a lot of work to add unit tests, but yo get * Reliability * Speed of making changes and bug fixes * Ability to refactor code easily * Unit tests serves as documentation of codes intentions I could not imagine working without it anymore. The code that is not dependent of ImageJ is tested isolated, that is simpler and faster. This is how I test the plugins that I have written in ImageJ. public ByteProcessor runPluginFilterOnImage(String fileName, PlugInFilter plugInFilter) { Opener opener = new Opener(); ImagePlus image = opener.openImage(fileName); ByteProcessor bp = (ByteProcessor) image.getProcessor(); plugInFilter.setup("", image); plugInFilter.run(bp); return bp; } During development or debugging I run individual unit tests from Eclipse. When I am checking code into Subversion I run all test from command line using Maven 2: mvn test Hope this is helpful, -Sami Badawi http://www.shapelogic.org |
In reply to this post by Konstantin von Martius
Dear Konstantin,
IJ is used in numerous scientific publications. You can try searching in PubMed with a keyword "ImageJ". Hope that this helps, Dimiter Prodanov |
Free forum by Nabble | Edit this page |