Imaging solutions for H&E nuclear counts

<< Previous Message | Next Message >>
From:George McNamara <>
Date:Mon, 28 Jun 1999 10:38:12 -0700
Content-Type:text/plain; charset="us-ascii"

Dear Ted,

  You can try ImagePath ( or Bacus laboratories Inc. ( the product is called B.L.I.S.S.) for pathology
oriented imaging systems. These will probably have an advantage (compared to
the 'generic" packages below) for you in terms of knowing your applications. I
am particularly impressed by how knowledgeable Dr. Alton Floyd of ImagePath is
(for example, see his technical notes such as Blue Feulgen stain, and his Automated Microscopy review, ). Bacus currently seems more interested in
developing a telepathology application than in doing more with the object
counts software (maybe becuse they've done all that's needed for object

  If you don't mind less knowledgeable (or no) tech support, try:

NIH Image (, free ... hint: you get what you
pay for)
Scanalytics IPLab (, the Microsoft Windows based product is
pretty simple to use for object counts.
Media Cybernetics Image Pro+ (, the Windows version is more
powerful than IPLab but somewhat more complicated to use.
Imaging Research, Inc., ( see See for "evaluating image analysis

In the USA many local microscope companies carry Image Pro+ and you can get a
demo CD from their web site (you still need to buy the right camera, etc.). 

If you have a Zeiss microscope (and like your local Zeiss rep's support), they
have an interesting new imaging software product, "Carl Zeiss Vision",  (however, I have not
used it for object counts).

An example of software you can download and try out for free (and then pay for
it if you find it useful) is:
SigmaScan image analysis software from
(12.5 Mb download).

  You should also be aware that ANY counterstain (including eosin) adds some
complexity to counting nuclei. For H&E nuclei quantitation you would typically
use a monochrome CCD camera with a color filter that transmits only 600 nm
light, so that only hematoxylin absoprtion contributes to the image used for
quantitation. Use of an ~520 nm filter would capture the eosin (plus
hematoxylin) image.
  An alternative is to use a color camera and nuclei identification with "HSI"
color space (both IPLab/Windows and ImagePro+ may be able to do this).

  A good question to your salesperson(s) is to ask them their preferred system
for counting nuclei in standard H&E stained slides (and don't let them cheat by
suggesting light eosin counterstaining). Salespeople with poor explanations
would not be invited in for demo's.

Good luck,


>Date: 24 Jun 1999 08:44:51 -0500 
>Subject: Image Analysis/nuclei counts
>I am considering purchase of image analysis software, and I would enjoy 
> recommendations on software that can count nuclei in digital images of 
>standard H&E or ToluidineBlue-stained sections. We work predominantly with
bone and 
>cartilage samples. For the software, ease of use and accessibility of
>support are important. Does anyone have extensive (positive) experience with 
>such software, or are there websites that show comparisons of different 
>package's capabilities?
>- -Ted

George McNamara, Ph.D.
Applied Spectral Imaging, Inc.
2120 Las Palmas Drive, Suite D
Carlsbad, CA 92009
voice   760-929-2840 ext 17
fax     760-929-2842
Pager 888-858-7243 pin 124797 (use area code phone ext)
Electronic Mail   
Company Web Site
Personal Web Page
Imaging Applications for Definitive InCyte:
   SKY - Spectral Karyotyping, 24 color human or 21 color mouse chromosomes
   SPY - Spectral Pathology, fluorescence or bright-field molecular diagnostics
   SIM - Spectral Imaging Microscopy (call or email to for what this can do for

<< Previous Message | Next Message >>