High-resolution cytometry : Hardware approaches, image analysis techniques and applications

KOZUBEK, Michal. High-resolution cytometry : Hardware approaches, image analysis techniques and applications. Brno: Masaryk University, 1998, 122 pp.

Basic information

• Original name: High-resolution cytometry : Hardware approaches, image analysis techniques and applications
• Authors: KOZUBEK, Michal.
• Edition: Brno, 122 pp. 1998.
• Publisher: Masaryk University

Other information

• Type of outcome: Special-purpose publication
• Confidentiality degree: is not subject to a state or trade secret
• Organization unit: Faculty of Informatics
• Keywords in English: image analysis; high resolution cytometry; fluorescence in situ hybridization; automated microscopy
• Tags: automated microscopy, fluorescence in situ hybridization, high resolution cytometry, Image analysis

Abstract

This thesis presents a new automated optical microscopy system developed in Brno and called Brno high-resolution cytometer (HRCM). HRCM enables high-resolution 2-D and 3-D topology analysis of small cell components stained with various fluorescent dyes which is performed using automated fluorescence microscopy, automated image acquisition and appropriate image analysis software developed by the author. HRCM is currently used for gene topology studies using fluorescence in situ hybridization (FISH) technique. The main applications are human genome studies and diagnostics of deleterious diseases such as leukemia. The system is unique and measurements performed using HRCM cannot be performed using any other instrument at comparable speed and comparable quantities of processed cells. At present time, the HRCM system is based on Leica DMRXA fully motorized computer-driven epi-fluorescence microscope, Quantix high-resolution cooled digital CCD camera and a high-performance two-processor PC computer which performs both the acquisition and related on-line image analysis. The images of different dyes are acquired sequentially using highly-specific filters and superimposed in computer memory. For each cell nucleus and each hybridization dot, user-selected attributes (such as position, size, intensity, etc.) are computed off-line using a spare processor or another computer connected with a network. Using HRCM, it is possible to analyze multi-color preparations including UV-excited dyes as well as repeatedly hybridized preparations re-acquiring individual cell nuclei. The speed of the acquisition and analysis is up to about 100 nuclei per minute (dependent on the density of nuclei on the slide); the precision of the lateral and axial measurements is approximately 100 nm. Thus, using overnight acquisition, quantities comparable to those of flow-cytometry or laser scanning cytometry measurements can be analyzed with an accuracy comparable to confocal microscopy. The whole software system was created by the author in the C programming language and the user interface currently runs in Microsoft Windows NT environment. The software has been developed, tested and used in experiments for about four years (Kozubek M. et al. 1995a, 1995b, 1998a, 1998b and 1998c, Kozubek S. et al. 1997 and 1998, Lukasova et al. 1997 and 1998). Both known algorithms and newly developed ones were used. Newly developed were the algorithms of local thresholding of cell nuclei, determination of unambiguous nuclear boundary chain, nuclear boundary smoothing, hybridization dot segmentation and precise determination of 3-D dot positions using conventional microscopy. At present, a new version of the HRCM system is being built based on a motorized version of Zeiss Axiovert 100 fluorescence microscope equipped with CARV confocal unit. This version should provide even more information about the observed cell content. Recently, a new confocal technique was developed, namely scanning correlation aperture microscopy (Juskaitis et al. 1996, Wilson et al. 1996 and 1997), which enables non-laser confocal observations with high light budget. This type of microscopy is also described in this thesis and is planned to be used in future confocal modifications of HRCM.

Abstract (in English)

This thesis presents a new automated optical microscopy system developed in Brno and called Brno high-resolution cytometer (HRCM). HRCM enables high-resolution 2-D and 3-D topology analysis of small cell components stained with various fluorescent dyes which is performed using automated fluorescence microscopy, automated image acquisition and appropriate image analysis software developed by the author. HRCM is currently used for gene topology studies using fluorescence in situ hybridization (FISH) technique. The main applications are human genome studies and diagnostics of deleterious diseases such as leukemia. The system is unique and measurements performed using HRCM cannot be performed using any other instrument at comparable speed and comparable quantities of processed cells. At present time, the HRCM system is based on Leica DMRXA fully motorized computer-driven epi-fluorescence microscope, Quantix high-resolution cooled digital CCD camera and a high-performance two-processor PC computer which performs both the acquisition and related on-line image analysis. The images of different dyes are acquired sequentially using highly-specific filters and superimposed in computer memory. For each cell nucleus and each hybridization dot, user-selected attributes (such as position, size, intensity, etc.) are computed off-line using a spare processor or another computer connected with a network. Using HRCM, it is possible to analyze multi-color preparations including UV-excited dyes as well as repeatedly hybridized preparations re-acquiring individual cell nuclei. The speed of the acquisition and analysis is up to about 100 nuclei per minute (dependent on the density of nuclei on the slide); the precision of the lateral and axial measurements is approximately 100 nm. Thus, using overnight acquisition, quantities comparable to those of flow-cytometry or laser scanning cytometry measurements can be analyzed with an accuracy comparable to confocal microscopy. The whole software system was created by the author in the C programming language and the user interface currently runs in Microsoft Windows NT environment. The software has been developed, tested and used in experiments for about four years (Kozubek M. et al. 1995a, 1995b, 1998a, 1998b and 1998c, Kozubek S. et al. 1997 and 1998, Lukasova et al. 1997 and 1998). Both known algorithms and newly developed ones were used. Newly developed were the algorithms of local thresholding of cell nuclei, determination of unambiguous nuclear boundary chain, nuclear boundary smoothing, hybridization dot segmentation and precise determination of 3-D dot positions using conventional microscopy. At present, a new version of the HRCM system is being built based on a motorized version of Zeiss Axiovert 100 fluorescence microscope equipped with CARV confocal unit. This version should provide even more information about the observed cell content. Recently, a new confocal technique was developed, namely scanning correlation aperture microscopy (Juskaitis et al. 1996, Wilson et al. 1996 and 1997), which enables non-laser confocal observations with high light budget. This type of microscopy is also described in this thesis and is planned to be used in future confocal modifications of HRCM.