Image analysis – Applications – Biomedical applications
Reexamination Certificate
2000-12-04
2002-08-20
Johns, Andrew W. (Department: 2621)
Image analysis
Applications
Biomedical applications
Reexamination Certificate
active
06438261
ABSTRACT:
FIELD AND BACKGROUND OF THE INVENTION
The present invention relates to methods of imaging and analysis of particles and, in particular, to a method for in-situ focus-fusion multi-layer spectral imaging and analysis of particulate samples.
When a sample featuring, for example, a particle, an aggregate of particles, or a dispersion of particles, has large layer or depth variations relative to changes in the distance from which it is viewed, an image of the sample exhibits a layer dependent or spatially varying degree of sharpness. This is referred to as a defocused image of the sample or scene, where some of the objects of the scene are in focus, while other objects of the scene are out of focus. Defocused images contain information potentially useful for scene analysis. The analysis of scenes from defocused images is of general interest in machine vision applications, for example, in active vision or robot vision where a camera actively explores a scene by continuously changing its position or field of view, relative to scene features. Applying scene analysis to defocused images is highly useful for accurately interpreting and understanding images of pharmaceutical, biomedical, biological, environmental, and microscopy samples, where layer or depth variations of imaged samples of powders, frozen suspensions of powders, biological specimens, air pollution particulates, or other multi-layered particulate samples are typically large compared to imaging distances. Scene analysis of defocused images is of particular applicability to depth dependent particulate samples, where, for instance, one or more layers of bacterial or fungal growth, exhibiting fluorescent emission properties in addition to the fluorescent emission properties of the particles themselves, is present on the particles, and there is a need for separation of imaging and analysis of the bacterial or fungal growth from that of the particles. Additionally, scene analysis is particularly applicable to depth dependent particulate samples of aerosols containing polycyclic aromatic hydrocarbons (PAHs) and other fluorescent particulate contaminants.
In conventional scene analysis using methods and systems for imaging particles, for example, for each scene, there is auto-focusing, where a best focal position is determined for use in analyzing or classifying particle properties. For some scenes, this is possible, and a focused image may be obtained in an automatic manner. Typically, an auto-focus module is coupled with a computer controlled mechanism that automatically changes the focal position, by moving along an axis parallel to the optical axis of the imaging or focusing sensor, thereby enabling identification of a good focal position. For other scenes, a good focal position is not guaranteed to exist and further image processing based on focus-fusion methodology is required.
When a focused image of a spatially varying or depth dependent scene can not be generated by using such electromechanical microscopy means, such that a single focal position can not be identified, a focused representation of the scene can be constructed by combining or fusing several defocused images of the same scene. This process is referred to as focus-fusion imaging, and the resulting image of such processing is referred to as a focus-fusion image. Defocused images, for example, those acquired during auto-focusing, are fused together such that each target in a given scene is in correct focus. Scene targets are detected by analyzing either the focused image, if it exists, or the focus-fusion image.
A current technique of imaging particles is based on spectral imaging. In spectral imaging, a particulate sample is affected in a way, for example, excitation by incident ultraviolet light upon the sample, which causes the sample to emit light featuring an emission spectra. Emitted light is recorded by an instrument such as a scanning interferometer that generates a set of interferogram images, which in turn are used to produce a spectral image, also referred to as a cube image, of the sample. Each cube (spectral) image is a three dimensional data set of voxels (volume of pixels) in which two dimensions are spatial coordinates or position, (x, y), in the sample and the third dimension is the wavelength, (&lgr;), of the imaged (emitted) light of the sample, such that coordinates of each voxel in a spectral image or cube image may be represented as (x, y, &lgr;). Any particular wavelength, (&lgr;), of imaged light of the sample is associated with a set of cube images or spectral fingerprints of the sample in two dimensions, for example, along the x and y directions, whereby voxels having that value of wavelength constitute the pixels of a monochromatic image of the sample at that wavelength. Each cube image, featuring a range of wavelengths of imaged light of the sample is analyzed to produce a two dimensional map of the chemical composition, or of some other physicochemical property of the sample, for example, particle size distribution.
An example of a method and system for real-time, on-line chemical analysis of particulate samples, for example, polycyclic aromatic hydrocarbon (PAH) particles in aerosols, in which the PAH sample is excited to emit light, for example fluorescence, is that of U.S. Pat. No. 5,880,830, issued to Schechter, and manufactured by GreenVision Systems Ltd. of Tel Aviv, Israel, and is incorporated by reference for all purposes as if fully set forth herein. In the disclosed method, spectral imaging techniques are implemented to acquire an image and analyze the properties of fixed position PAH particles. As part of this method, air is sampled by means of a high volume pump sucking a large volume of air featuring aerosol contaminated with PAH particles onto a substrate, followed by on-line imaging and scene analysis of the stationary particles.
A method of calibration and real-time analysis of particles is described in U.S. Pat. No. 6,091,843, to Moshe et al., and is incorporated by reference for all purposes as if fully set forth herein. The method described, is based on using essentially the same system of U.S. Pat. No. 5,880,830, for acquiring spectral images of static particles on a filter. Targets are identified in static particle images and are classified according to morphology type and spectrum type. Each target is assigned a value of an extensive property. A descriptor vector is formed, where each element of the descriptor vector is the sum of the extensive property values for one target class. The descriptor vector is transformed, for example, to a vector of mass concentrations of chemical species of interest, or of number concentrations of biological species of interest, using a relationship determined in the calibration procedure. In the calibration procedure, spectral images of calibration samples of static particles having known composition are acquired, and empirical morphology types and spectrum types are inferred from the spectral images. Targets are identified in the calibration spectral images, classified according to morphology type and spectrum type, and assigned values of an extensive property. For each calibration sample, a calibration descriptor vector and a calibration concentration vector is formed. A collective relationship between the calibration descriptor vectors and the calibration concentration vectors is found using chemometric methods.
In the method of U.S. Pat. No. 6,091,843, standard spectra are determined empirically in the calibration procedure. In such analytical procedures, empirical calibration is quite important for leading to highly accurate results based on image analysis and classification, because spectra of adsorbed chemical species in general, and, of PAHs in particular, are known to be altered by the surfaces on which they are adsorbed, and by the presence of contaminants during sample preparation and image acquisition. Moreover, in the described method, the relationship between the descriptor vector and the concentration vector accounts explicitly and simultaneously for both morphol
Khazanski Michael
Moshe Danny S.
G.E. Ehrlich Ltd.
Green Vision Systems Ltd.
Johns Andrew W.
LandOfFree
Method of in-situ focus-fusion multi-layer spectral imaging... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of in-situ focus-fusion multi-layer spectral imaging..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of in-situ focus-fusion multi-layer spectral imaging... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2895768