Microscopica acta. Supplement最新文献

A prescreening system for the early detection of uterine cancer was developed. For the diagnosis of each input specimen (PAP-stained), a figure of malignancy has to be calculated automatically. Methods, problems and results of the system are summarized in this paper. Some thousand microscopic subfields of a specimen are successively scanned by an optimized TV-camera with high spatial resolution. The automated specimen analysis can be described by a two step procedure: single cell classification and evaluation of cell population. For the single cell classifier only features derived from the nucleus are used. The pattern recognition procedures are based on a processor-oriented strategy, and can be adapted to other cytological specimen. The algorithms have been tested with 3 . 10(5) cell images of about 300 specimens.

In the recent past symptoms of a reorganization of laboratory medicine became apparent. They were brought about by methodological progress in analytical biochemistry, chemistry and also by the rapidly increasing efficiency of automation in the clinical laboratory as well as in medical practice. A somewhat parallel change also comes into existence in cytology and also, within certain limits, the histology. Since the early seventies the German Government, in accord with these developments, has supported some projects purporting to design better, more reproducible and more simple procedures in clinical diagnosis. With regard to cytological detection of malignancy this meant the development of methods to render cytological diagnosis a more potent tool for the recognition of cancer at an earlier stage on pre-selected material and with a high degree of sensitivity. Requirements on the quality of these methods with regard to reliability and safety will be discussed together with those marginal conditions which are expected to have a bearing on decisions aiming at the introduction of those methods into cytological routine. Consideration of those problems will be based on qualification requirements on prescreening procedures in general and manual cytological prescreening in particular. Discussion will also be extended on procedural sources of error.

Analytical and quantitative cytology is based upon photometric and morphometric measurements of single cells in stained cytological specimens. It analyzes the features of single cells and cell populations with respect to their cytodiagnostic relevance. One way to do this is the establishment of learning sets of visually selected cells, and the intercomparison of unknown cells with these data sets. Difficulties arise with the visual classification of single cells from PAP-stained gynecological smears and the pooling into classes due to the considerable photometric variation from specimen to specimen. It is demonstrated, that neither for a cancer prescreening apparatus nor an interactive diagnostic machine, highly differentiated learning sets can be dispensed with. The build-up of the TUDAB learning sets and results of single cell classification, visual reclassification, as well as population analysis and specimen classification are shown.

The prototype of an automated screening system for cytological preparations is under development. The conditions concerning the handling of such system require the development of a special microscope with certain adjustment functions to be automated. The main specifications of this microscope are: --1 Splitting the light path into two synchronously usable channels for low and high magnification; whereby both images are projected onto two tv-cameras --Additional light path for visual observation --Automated focusing via movement of the microscope objective --Automated supply of immersion oil --Positioning and loading unit for slides --Microprocessor control --Closed and dust-protected design. Besides the mentioned specifications different microscopical observation methods, i. e. transmitted light bright field, incident fluorescence, are possible. The system can be used for all samples which are prepared on standard slides.

Registration of slide-movements in conventional gynecological smears - its relevance for automated cytology - The information contained in conventional cytologic smears can only be obtained by means of a very scrutinizing scanning of the slide. This is one of the reasons, why people are interested in automated cytology. In the present study slide movements during 216 visual screenings of conventional gynecologic smears were graphically and electronically recorded. The recordings show alternating movements and stops. The time of stops shows a high variation with a mean value of about 0,3 seconds. The moving times are mostly below 0,2 seconds. The relation of total stop times to total movement times is in most cases between two and three. The percentage of total stop times, the total screening time, the relation of stop times to movement times and the number of microscopic fields evaluated while standing rise in difficult specimens. There are obvious differences between the screening patterns of different cytotechnologists. An analysis of the recordings concerning the number of stops shows that even in smears which are screened very carefully a considerable percentage of the smear is not seen during the stops. The pattern of mechanical slide movements allows conclusions about the "data acquisition" by the human cytoscreener.

Measurement and quantification are the prerequisites of automated cell diagnosis. The goal of the analysis is to recognize quantitative differences for the classification of benign and malignant cells. Absorption and fluorescence dyes can be bound to definite cell components proportional to their amounts and can be determined photometrically. In static photometry the cells are spread out in a single layer on a glass slide. By examining the cell point by point a scanning image can be obtained. These high resolution photometric measurements provide many individual data on relatively few cells. In flow photometry the cells are suspended in a fluid. The fluid flows past the lens of the microscope. Because each cell remains in front of the lens for only a brief period, it is possible to detect only one or few parameters per cell. These zero or low resolution flow photometric procedures provide few individual data on a great number of cells. Hybrid procedures combine flow photometry with subsequent high resolution photometry of the individual cell. To make this possible suspicious cells are sorted out from the flowing fluid and deposited on a slide. In individual cases the efficiency of visual diagnosis is almost achieved by machines.

Monolayer preparations used in automated cytology are generated from gynecologic cell suspensions. The suspension-derived preparations (s.p.) show some peculiarities compared with conventional cytologic smears. One important feature of the s.p. is a distribution of cells mainly by chance. Therefore it seems to be possible to determine a cell sample size in minimum needed for cytologic diagnosis. In a blind test the needed cell sample size was determined by counting the cells in the microscopic fields of vision (ocular 12,5 x, objective 10 x, phi of the field of vision 2 mm) step by step until it was possible to classify the specimen (diagnosis 1). To be on the safe side (personal safety requirement) some further cells were assessed (diagnosis 2). We used s.p. from 50 non-suspicious women and 50 women with invasive squamous cell carcinoma of the uterine cervix and its precursors. To recognize strong positive cases (severe dysplasia to invasive carcinoma) only about 600 cells are needed (mean: 150/147 cells). The most cells (about 1200 cells) are required in negative cases and cases of mild to moderate dysplasia (mean: 293/236 cells). The highest personal safety requirement was found in cases of mild to moderate dysplasia (311% compared with a mean sample size of 236 or 100%). The results support such approaches in automated cytology which analyze only some hundreds or about 1000 cells with high resolution.