Automatic analysis of blood cells.
About: This article is published in Scientific American.The article was published on 1970-11-01. It has received 120 citations till now.
Citations
More filters
TL;DR: This survey summarizes some of the proposed segmentation techniques in the area of biomedical image segmentation, which fall into the categories of characteristic feature thresholding or clustering and edge detection.
1,160 citations
TL;DR: It is shown that traditional methods based on measurements of image power-give the best results when tested on one set of real images and two sets of synthetic images, while the worst methods are those based on the image probability density function or histogram.
Abstract: Traditional autofocus methods were designed for microscopes driven by single processor computers. As computers are developed that exploit massive parallelism when acquiring and analyzing images, parallel cellular logic techniques became available to focus automatically. This paper introduces the reader to both cellular logic techniques for autofocus and a new spectral moment autofocus measure. It then compares these methods with more traditional autofocus methods. It is shown that traditional methods based on measurements of image power-give the best results when tested on one set of real images and two sets of synthetic images. The next best methods are the cellular logic and spectral moment techniques, while the worst are those based on the image probability density function or histogram.
378 citations
TL;DR: This paper selectively surveys contributions to major topics in pattern recognition since 1968, including contributions to error estimation and the experimental design of pattern classifiers.
Abstract: This paper selectively surveys contributions to major topics in pattern recognition since 1968. Representative books and surveys pattern recognition published during this period are listed. Theoretical models for automatic pattern recognition are contrasted with practical,, design methodology. Research contributions to statistical and structural pattern recognition are selectively discussed, including contributions to error estimation and the experimental design of pattern classifiers. The survey concludes with a representative set of applications of pattern recognition technology.
297 citations
TL;DR: Histograms of cell silhouette areas indicate that rapid and accurate estimates of bacterial biovolume and biomass will be possible with this image analysis system, which has been satisfactorily tested at sea.
Abstract: Epifluorescence microscopy is now being widely used to characterize planktonic procaryote populations. The tedium and subjectivity of visual enumeration and sizing have been largely alleviated by our use of an image analysis system consisting of a modified Artek 810 image analyzer and an Olympus BHT-F epifluorescence microscope. This system digitizes the video image of autofluorescing or fluorochrome-stained cells in a microscope field. The digitized image can then be stored, edited, and analyzed for total count or individual cell size and shape parameters. Results can be printed as raw data, statistical summaries, or histograms. By using a stain concentration of 5 micrograms of 4'6-diamidino-2-phenylindole per ml of sample and the optimal sensitivity level and mode, counts by image analysis of natural bacterial populations from a variety of habitats were found to be statistically equal to standard visual counts. Although the time required to prepare slides, focus, and change fields is the same for visual and image analysis methods, the time and effort required for counting is eliminated since image analysis is instantaneous. The system has been satisfactorily tested at sea. Histograms of cell silhouette areas indicate that rapid and accurate estimates of bacterial biovolume and biomass will be possible with this system.
205 citations
01 May 1979
TL;DR: Cellular logic computers have become a commercial product in biomedical image processing where they are used in clinical instruments whose purpose is to classify white blood cell images at rates of several thousand per hour.
Abstract: Cellular logic operations (CLO's) are performed digitally to transform an array of data P(I,J) into a new data array P'(I,J). The value of each element in the new array is determined by its value in the original array and the orginal values of its nearest neighbors. The neighborhood configuration (tessellation) is usually called the "cell"; whence the term "cellular logic." CLO's may be categorized acording to the tessellation in which they are embedded and according to the type or types of CLO sequences: sequences which are carried out in a single step; those which iterate the same CLO for many steps; those which repetitively alternate subsequences of CLO strings. The effect of the CLO sequence on the contents of the data array is frequently one of boundary modification. Depending on the CLO sequence(s) utilized, a boundary may be expanded to form the convex hull, or reduced so as to form the convex kernel, skeleton, or residue. As of 1977, cellular logic computers have become a commercial product in biomedical image processing where they are used in clinical instruments whose purpose is to classify white blood cell images at rates of several thousand per hour. Many other applications are foreseen and, as further examples, preliminary results in automatic X-ray image analysis and tissue image analysis are presented.
198 citations