Hans Netten

Autofocusing in microscopy based on the OTF and sampling

Frank R. Boddeke, Lucas J. van Vliet, Hans Netten and Ian T. YoungPattern Recognition Group of the Faculty of Applied PhysicsDelft University of TechnologyLorentzweg 1, 2628 CJ Delft, The NetherlandsAbstractIn the literature many autofocus algorithms have been proposed and compared (Groen et al. 1985;Firestone at al. 1991; Yeo et al. 1993; Price and Gough 1994) for use in optical microscopy (brightfield and fluorescence microscopy). Most of the focus criteria measure the high frequency contents ofa recorded image as a measure of focus. In this paper we show that a focus criteria should measure thesignal power of the middle frequency, since defocusing mainly reduces the frequencies around halfthe cut-off frequency of the optical system. The filter that provides the required band–pass filteringdepends strongly on the sampling density of the camera. There are two practical combinations ofsampling density and one-dimensional digital band-pass filter:• Sampling at the Nyquist frequency and the {1,0,–1} filter;• Sampling at half the Nyquist frequency and the {1,–1} filter.The latter is to be preferred due to noise considerations and the fact that it uses four times fewersample points. Calculation speed can also be increased by further reducing the sampling densityperpendicular to the filter (on chip or in software) down to 1/8 of the Nyquist frequency. We havedesigned a three-phase autofocus algorithm that works well in fluorescence and bright fieldmicroscopy. The phases are:• Coarse, find the region near focus (step size of typically a few microns);• Fine, find a quadratic region around focus (step size around one micron);• Refine, use a quadratic fit on samples around the peak to find the in-focus position.We found that the final focus error is smaller than the mechanical reproducibility of our z-axis (50nm) for light levels down to 400 photo-electrons per pixel (sampling at the Nyquist frequency using acooled CCD camera with pixels of 6.8

Automation of spot counting in interphase cytogenetics using brightfield microscopy

In situ hybridization techniques allow the enumeration of chromosomal abnormalities and form a great potential for many clinical applications. Although the use of fluorescent labels is preferable regarding sensitivity and colormultiplicity, chromogenic labels can provide an excellent alternative in relatively simple situations, e.g., where it is sufficient to use a centromere specific probe to detect abnormalities of one specific chromosome. When the frequency of chromosomal aberrations is low, several hundreds or even thousands of cells have to be evaluated to achieve sufficient statistical confidence. Since manual counting is tedious, fatiguing, and time consuming, automation can assist to process the slides more efficiently. Therefore, a system has been developed for automated spot counting using brightfield microscopy. This paper addresses both the hardware system aspects and the software image analysis algorithms for nuclei and spot detection. As a result of the automated slide analysis the system provides the frequency spot distribution of the selected cells. The automatic classification can, however, be overruled by human interaction, since each individual cell is stored in a gallery and can be relocated for visual inspection. With this system a thousand cells can be automatically analyzed in approximately 10 min, while an extra 5-10 min is necessary for visual evaluation. The performance of the system was analyzed using a model system for trisomy consisting of a mixture of male and female lymphocytes hybridized with probes for chromosomes 7 and Y. The sensitivity for trisomy detection in the seeding experiment was such that a frequency of 3% trisomic cells could be picked up automatically as being abnormal according to the multiple proportion test, while trisomy as low as 1.5% could be detected after interaction.

Quantitative evaluation of light microscopes based on image processing techniques

In this note we will present methods based on image processing techniques to evaluate the performance of light microscopes. These procedures are applied to three different ‘high-end’ light microscopes. Tests are carried out to measure the homogeneity of the illumination system. From these tests it follows that Kohler illuminated images can have an exceedingly high amount of shading. Another result found from the illumination calibration is that traditional lamp housings are not designed to make fine-tuning easy. Next, the (automated) stage is considered. Several tests are performed to measure the stage motion in the xy-plane and in the axial direction to address accuracy, precision, and hysteresis effects. Finally, the entire electro-optical system is characterized by measuring the optical transfer function (OTF) at wavelengths 400 nm, 500 nm, 600 nm, and 700 nm. The results of these experiments show that there is a consistent deviation from the theoretical OTF at wavelengths around 400 nm. The final conclusion is that modern light microscopes perform better than their five-to-ten-year-old predecessors.

Automation of fluorescent dot counting in cell nuclei

We have developed a completely automated fluorescence microscope system that can examine 500 cells in approximately 20 minutes to determine the number of labeled chromosomes (seen as dots) in each cell nucleus. This system works with two fluorescent dyes-one for the DNA hybridization dots (e.g. FITC) and one for the cell nucleus (e.g. DAPI). After the stage has moved to a new field the image is automatically focused, acquired by a Photometrics KAF 1400 camera, and then analyzed on a Macintosh Quadra 840AV computer. After the required number of cells has been analyzed, the user may interact to correct the computer by working with a gallery of the cell images. The machine accuracies are equal to panels of human experts (manual) and limited by the overlapping of dots in the 3D cell as seen through the 2D projection.

A fast scanner for fluorescence microscopy using a 2-D CCD and time delayed integration

We have developed an imaging system for high speed image acquisition in fluorescence microscopy. The use of a two- dimensional CCD array in a special operation mode called TDI (Time Delayed Integration) permits a significant increase in photon integration time compared to 1D scanners (higher signal-to-noise ratio) without compromising the total data throughput rate. Instead of a start–stop system we use continuous stage motion in the CCDs parallel shift direction. Synchronizing the parallel clock and the stage velocity guarantees a one-to-one relationship between a moving cell and its image onto the CCD. Compared to start-stop system, TDI scanning offers a speed improvement, negligible blurring in the scanning direction and a complete suppression of pixel variability boosting the SNR more than 10 dB.