NEW AUTOMATED METHOD FOR CELL REGISTRATION AND SIGNAL QUANTIFICATION

Claus S. Kristensen csk at maaloe.lm.dth.dk
Mon Mar 14 15:46:29 EST 1994


A NEW AUTOMATED METHOD FOR CELL REGISTRATION AND SIGNAL
QUANTIFICATION.

A new computer program for automatic cell detection and signal
quantification has been developed in a cooperation between The
Department of Microbiology and Institute of Mathematical Modelling
at the Danish Technical University.

The biological basis of the system is the ribosomal probing
technique using flourescence-labelled oligonucleotide probes.
The signal emitted from whole cells hybridized with a probe
targeting the ribosomal RNA is proportional to the number of
ribosomes in the cell. Using digital cameras (CCD cameras)
mounted on microscopes, the flourescent signal from such cells
can be captured and analyzed on a computer. 

Traditionally, the quantification of the light is done manually
using a pointing device (mouse) to circumscribe the cells and
read out the corresponding light-values from the encircled
pixels in the digital image. This method, however, is error-
prone since it is dependent on the operator, the quality
(contrast) of the displayed image and background noise in the
particular image.

We have developed a new, unix-based computer program which
automatically detects cells, determines the cell volume and
calculates the corresponding average light per pixel. When
comparing to chemical determinations this program provides
precise information on ribosomal contents of the depicted
cells, and further limits the variation compared to the manual
method. One 1024 X 1024 pixel image can be automatically
quantified in approx. 1.5 minutes (dependent of hardware) and
the batch processing of series of images can be carried out
e.g. over night. Reference images are generated for control of
the detection. Output is either pure ASCII or CSV (comma-separated
files, ready for use in spreadsheets).

As we in the future would consider to make this program
available to educational institutions for a symbolic [non-profit] 
fee (to cover development costs and future enhancements), we would 
like to inquire who would possibly be interested in the program. 

Supported platforms: SUN, IBM AIX, HP, Silicon Graphics, Linux
- and possibly more on inquiry.

If you think this could be of your interest, please send us a
note. If you have specific whishes, please indicate so, and we
will consider to include such features. 

Contact adresses can be found below (E-mail preferred). Please do not 
post replies to the news groups, as we may not see it.

Claus Kristensen

-------------------------------------------------------------------
*  Claus Sternberg Kristensen, PhD                                *
*  Department of Microbiology   Danish Technical University /DTU/ *
*  Building 221                 DK 2800 Lyngby Denmark            *
*  Phone (Den:) 45 93 12 22     ext. 2522 or 2515                 *
*  Phone (Int:) +45 45 93 34 22 FAX +45 45 93 28 09               *
*  E-mail: csk at lm.dth.dk                                          *
-------------------------------------------------------------------



More information about the Bionews mailing list