Colony counting hard/software advice needed.

Alan Townsend atown at isnet.is.wfu.edu
Thu Jan 19 17:03:15 EST 1995


	I need a hardware + software solution for counting and sizing
of dark blue stained colonies on clear plastic plates.  The basic
analytical   problem and requirements are:      
1. Must be easily adaptable for rapid & reproducible counting of many
(50 to 500)wells/dishes of variable size/format : e.g. 35, 60, or   100, or
150 mm dishes;  or 6-, 12-, or 24-well plastic tissue culture plates.   
2. The number of colonies and the area of each must be obtained in a
form   that can be exported to a spreadsheet, or stored in a file.      
3. Ability to handle edge refraction lines, by cropping or field
definition.   Colony detection near the edges of the plate.   
4. Image erosion/dilation, image subtraction, etc are desired, but may
not   be essential if the software separates touching colonies .      
	I have seen several approaches to colony counting, all of which  
basically isolate the  circular analysis field and perform "spot
counting" and area determination on binary  images after thresholding,
etc (e.g. NIH   Image). One approach used a video camera to  generate
a live image to allow real-time manual positioning of the wells under a 
stationary but   variable-sized circle superimposed on the monitor,
which defined the analytical field (Artek 880, Image Products Intl.,
Chantilly, VA).  This  system was a  simple video digitizer and gave
only spot count; we also need area, and would like a full  featured
image analysis program if possible.  Another possible approach could
be by  identifying  circular areas (best would be to initialize by mouse
point and click to locate areas), followed by edge elimination or find
center and draw radius   functions, then  analysis of live image or
captured image file.  I have also   heard of scanner - based  methods,
but I believe that this would probably   not be practical here.  I prefer
the  Macintosh platform, but a PC-based   system would be acceptable. 

Please send replies to this newsgroup, or to my email address :
atown at isnet.is.wfu.edu,  or mail to me at : Biochem. Dept.,  BGSM,
Med. Ctr. Blvd, Winston-Salem, N.C. 27157. Thanks, Alan Townsend.  




More information about the Methods mailing list