Evaluation of confocal microscopy system performance.

BACKGROUND: The confocal laser scanning microscope (CLSM) has been used by scientists to visualize three-dimensional (3D) biological samples. Although this system involves lasers, electronics, optics, and microscopes, there are few published tests that can be used to assess the performance of this equipment. Usually the CLSM is assessed by subjectively evaluating a biological/histological test slide for image quality. Although there is a use for the test slide, there are many other components in the CLSM that need to be assessed. It would be useful if tests existed that produced reference values for machine performance. The aim of this research was to develop quality assurance tests to ensure that the CLSM was stable while delivering reproducible intensity measurements with excellent image quality.
METHODS: Our ultimate research objective was to quantify fluorescence using a CLSM. To achieve this goal, it is essential that the CLSM be stable while delivering known parameters of performance. Using Leica TCS-SP1 and TCS-4D systems, a number of tests have been devised to evaluate equipment performance. Tests measuring dichroic reflectivity, field illumination, lens performance, laser power output, spectral registration, axial resolution, laser stability, photomultiplier tube (PMT) reliability, and system noise were either incorporated from the literature or derived in our laboratory to measure performance. These tests are also applicable to other manufacturer's systems with minor modifications.
RESULTS: A preliminary report from our laboratory has addressed a number of the QA issues necessary to achieve CLSM performance. This report extends our initial work on the evaluation of CLSM system performance. Tests that were described previously have been modified and new tests involved in laser stability and sensitivity are described. The QA tests on the CLSM measured laser power, PMT function, dichroic reflection, spectral registration, axial registration, system noise and sensitivity, lens performance, and laser stability. Laser power stability varied between 3% and 30% due to various factors, which may include incompatibility of the fiber-optic polarization with laser polarization, thermal instability of the acoustical optical transmission filter (AOTF), and laser noise. The sensitivity of the system was measured using a 10-microm Spherotech bead and the PMTs were assessed with the CV concept (image noise). The maximum sensitivity obtainable on our TCS-SP1 system measured on the 10-microm Spherotech beads was approximately 4% for 488 nm, 2.5% for 568 nm, 20% for 647 nm, and 19% for 365 nm laser light. The values serve as a comparison to test machine sensitivity from the same or different manufacturers.
CONCLUSIONS: QA tests are described on the CLSM to assess performance and ensure that reproducing data are obtained. It is suggested strongly that these tests be used in place of a biological/histological sample to evaluate system performance. The tests are more specific and can recognize instrument functionality and problems better than a biological/histological sample. Utilization of this testing approach will eliminate the subjective assessment of the CLSM and may allow the data from different machines to be compared. These tests are essential if one is interested in making intensity measurements on experimental samples as well as obtaining the best signal detection and image resolution from a CLSM. Published 2001 Wiley-Liss, Inc.