Abstract:
Systems and methods for image pattern recognition comprise digital image capture and encoding using vector quantization (nullVQnull) of the image. A vocabulary of vectors is built by segmenting images into kernels and creating vectors corresponding to each kernel. Images are encoded by creating a vector index file having indices that point to the vectors stored in the vocabulary. The vector index file can be used to reconstruct an image by looking up vectors stored in the vocabulary. Pattern recognition of candidate regions of images can be accomplished by correlating image vectors to a pre-trained vocabulary of vector sets comprising vectors that correlate with particular image characteristics. In virtual microscopy, the systems and methods are suitable for rare-event finding, such as detection of micrometastasis clusters, tissue identification, such as locating regions of analysis for immunohistochemical assays, and rapid screening of tissue samples, such as histology sections arranged as tissue microarrays (TMAs).
Abstract:
Methods and apparatus are provided for computing focus information during scanning digital microscope slide data with a line scan camera. The systems and methods include a dynamically interleaved procedure that works by moving the specimen relative to the objective lens while the height of the objective lens is adjusted relative to the stage. Imagery data is acquired at a plurality of objective lens heights the image data from the objective lens height having maximum contrast is stored and combined into a composite digital image of at least a portion of the specimen.
Abstract:
Systems and methods for image pattern recognition comprise digital image capture and encoding using vector quantization (“VQ”) of the image. A vocabulary of vectors is built by segmenting images into kernels and creating vectors corresponding to each kernel. Images are encoded by creating a vector index file having indices that point to the vectors stored in the vocabulary. The vector index file can be used to reconstruct an image by looking up vectors stored in the vocabulary. Pattern recognition of candidate regions of images can be accomplished by correlating image vectors to a pre-trained vocabulary of vector sets comprising vectors that correlate with particular image characteristics. In virtual microscopy, the systems and methods are suitable for rare-event finding, such as detection of micrometastasis clusters, tissue identification, such as locating regions of analysis for immunohistochemical assays, and rapid screening of tissue samples, such as histology sections arranged as tissue microarrays (“TMAs”).
Abstract:
A system and method for processing and analyzing virtual microscopy digital images (nullvirtual slidesnull) is provided. The system comprises an algorithm server that maintains or has access to a plurality of image processing and analysis routines. The algorithm server additionally has access to a plurality of virtual slides. The algorithm server executes a selected routine on an identified virtual slide and provides the resulting data. The virtual slide can be accessed locally or remotely across a network. Similarly, the image processing routines can be obtained from local storage or across a network, or both. Advantageously, certain common sub-routines may be stored locally for inclusion in other local or remotely obtained routines. Access to image processing and analysis may be restricted through a monitor process that authenticates requests to process or view virtual slides. Variations in restrictions to images provide a rich diversity in access levels that allow sharing of virtual slides and demonstrations of image processing algorithms.
Abstract:
A system and method for processing and analyzing virtual microscopy digital images ('virtual slides') is provided. The system comprises an algorithm server (20) that maintains or has access to a plurality of image processing and analysis routines. The algorithm server (20) additionally has access to a plurality of virtual slides. The algorithm server (20) executes a selected routine on an identified virtual slide and provides the resulting data. The virtual slide can be accessed locally or remotely across a network (80). Similarly, the image processing routines can be obtained from local storage (40) or across a network (80), or both. Advantageously, certain common sub-routines may be stored locally for inclusion in other local or remotely obtained routines. Access to image processing and analysis may be restricted through a monitor process that authenticates requests to process or view virtual slides. Variations in restrictions to images provide a rich diversity in access levels that allow sharing of virtual slides and demonstrations of image processing algorithms.
Abstract:
Systems and methods are provided to facilitate consultations between a referral source (e.g., labs, pathologists and patients) and a consultant (e.g., pathologist, radiologist, or other digital image analyst). Links between the various referral sources and consultants are established through a scanning center via a data communication network such as the Internet. The referral source sends a slide to the scanning center where the corresponding digital slide is posted for review and analysis by the consultant. Upon completion of the analysis and report, a digital slide conference is conducted through the scanning center that provides a venue for direct communication regarding the consultation. The scanning center may also facilitate payment from the referral source to the consultant.
Abstract:
Systems and methods for microscope slide scanning using multiple sensor arrays that receive imagery data from a single optical axis are provided. A single, high quality, easily obtained microscope objective lens is used to project an image onto two or more sensor arrays. The sensor arrays can be linear or two dimensional and imaging takes place along a single optical axis. Simultaneous sensor acquisition and parallel data processing reduce the image acquisition time by a factor of N, where N represents the number of sensors employed.