System for and method of intelligently directed segmentation analysis for automated microscope systems

Abstract

The present disclosure includes systems and techniques relating to intelligently directed segmentation analysis for automated microscope systems. In general, in one implementation, the technique includes obtaining an image of at least a portion of a scan region including a biological specimen, partitioning the obtained image into zelles, determining one or more parameters of the zelles, performing a cluster analysis on the one or more parameters of the zelles, differentiating tissue of greater interest from tissue of lesser interest in the obtained image based on the cluster analysis and based on a test being performed for the biological specimen, and storing more information for the tissue of greater interest than information for the tissue of lesser interest. The cluster analysis can be a multivariate statistical cluster analysis, and the zelles can be test-dependent zelles (e.g., having dimensions defined according to the test being performed for the biological specimen).

1 . An article comprising a machine-readable medium storing instructions operable to cause one or more machines to perform operations comprising: 
 processing an image of at least a portion of a scan region including a biological specimen, said processing including differentiating tissue of greater interest from tissue of lesser interest in the image based on a test being performed for the biological specimen and based on a cluster analysis of data from the image; and    storing information for the tissue of greater interest, which falls in the scan region.    
     
     
         2 . The article of  claim 1 , wherein the differentiating comprises: 
 defining subimages in the image based on the test being performed for the biological specimen;    determining one or more parameters of the subimages;    performing the cluster analysis on the one or more parameters of the subimages; and    identifying one or more areas in the subimages based on results of the cluster analysis, the one or more areas including the tissue of greater interest.    
     
     
         3 . The article of  claim 2 , wherein the defining subimages comprises specifying subimage dimensions based on the test being performed for the biological specimen.  
     
     
         4 . The article of  claim 2 , wherein the determining comprises determining multiple parameters of the subimages, and the performing comprises performing a multivariate statistical cluster analysis on the multiple parameters.  
     
     
         5 . The article of  claim 2 , wherein the identifying comprises selecting a proper subset of resulting clusters based on operating parameters set for the test being performed for the biological specimen.  
     
     
         6 . The article of  claim 2 , wherein the operations further comprise obtaining one or more higher magnification images of the biological specimen in the one or more areas of the subimages, and the information storing comprises saving the one or more higher magnification images.  
     
     
         7 . The article of  claim 2 , wherein the operations further comprise obtaining one or more higher magnification image samples of a cluster until a predefined number of samples meeting a specified criteria have been obtained for the cluster.  
     
     
         8 . The article of  claim 2 , wherein the operations further comprise obtaining a higher magnification image sample of a cluster, the information storing comprises saving the higher magnification image sample along with a lower magnification image of the cluster and information linking the higher magnification image sample with the lower magnification image, such that the higher magnification image sample is returned in response to a request for a high power image of a region in the lower magnification image, wherein the region does not overlap with the higher magnification image sample but is statistically similar to the higher magnification image sample according to the cluster.  
     
     
         9 . The article of  claim 8 , wherein the obtaining the higher magnification image sample comprises obtaining multiple samples covering representative members of the cluster, and the information storing comprises saving the samples, the lower magnification image and the linking information in a single file for distribution.  
     
     
         10 . The article of  claim 1 , the operations further comprising obtaining the image by performing a silhouette scan.  
     
     
         11 . The article of  claim 1 , wherein the information storing comprises retaining high resolution data for the tissue of greater interest, and discarding high resolution data for the tissue of lesser interest.  
     
     
         12 . A method comprising: 
 obtaining an image of at least a portion of a scan region including a biological specimen;    subdividing the obtained image into a plurality of subimages, wherein the subdividing is based on a test being performed for the biological specimen;    generating a derivative image wherein image units of the derivative image are derived from respective ones the subimages;    performing an automated analysis of the derivative image to identify one or more areas of interest for the test; and    storing information for the one or more areas of interest, which fall in the at least a portion of the scan region.    
     
     
         13 . The method of  claim 12 , wherein the performing the automated analysis comprises: 
 performing a multivariate statistical cluster analysis; and    grouping quadrants of the obtained image based on results of the multivariate statistical cluster analysis and the test being performed for the biological specimen.    
     
     
         14 . The method of  claim 13 , wherein the subdividing the obtained image comprises specifying subimage dimensions based on the test being performed for the biological specimen.  
     
     
         15 . The method of  claim 13 , wherein the grouping forms groups of quadrants, and the performing the automated analysis further comprises selecting a proper subset of the groups based on the test being performed for the biological specimen to identify the one or more areas of interest.  
     
     
         16 . The method of  claim 13 , wherein the grouping forms groups of quadrants, the performing the automated analysis further comprises determining a number of sample locations covering representative members of the groups based on the test, and the information storing comprises saving lower magnification image data for the groups and higher magnification image data for the sample locations.  
     
     
         17 . The method of  claim 13 , further comprising returning, in response to a request for a high power image of a first region in the lower magnification image data, at least a portion of the higher magnification image data corresponding to a second region with similar characteristics to the first region according to the multivariate statistical cluster analysis.  
     
     
         18 . The method of  claim 17 , further comprising updating a knowledge-base according to user input provided with respect to the at least a portion of the higher magnification image data returned, wherein the updated knowledge-base affects future applications of the multivariate statistical cluster analysis for the test.  
     
     
         19 . An automated imaging system comprising: 
 a microscope;    a controller coupled with the microscope; and    a display device coupled with the controller;    wherein the controller is configured to operate the microscope autonomously, to present image data on the display device, and to perform operations including:    obtaining an image of at least a portion of a scan region including a biological specimen;    partitioning the obtained image into zelles;    determining one or more parameters of the zelles;    performing a cluster analysis on the one or more parameters of the zelles;    differentiating tissue of greater interest from tissue of lesser interest in the obtained image based on the cluster analysis and based on a test being performed for the biological specimen; and    storing more information for the tissue of greater interest than information for the tissue of lesser interest.    
     
     
         20 . The system of  claim 19 , wherein the cluster analysis comprises a multivariate statistical cluster analysis, and the zelles comprise test-dependent zelles.  
     
     
         21 . The system of  claim 19 , wherein the operations further include segmenting the zelles into clusters exhibiting similar characteristics among a portion of the one or more parameters determined to best cluster the zelles according to the cluster analysis.  
     
     
         22 . The system of  claim 21 , wherein the operations further include determining which clusters contain zelles that most likely contain content that is valuable to a pathologist in making a diagnostic evaluation.  
     
     
         23 . The system of  claim 21 , wherein the operations further include determining how many high-power images of zelles to retain for each cluster based on a knowledge-base codifying previous test experience.  
     
     
         24 . The system of  claim 23 , wherein the operations further include capturing the high-power images of zelles, analyzing the captured high-power images to determine if they meet specified criteria, and terminating the capturing once a sufficient number of high-power images have been acquired for the test according to the determining how many high-power images of zelles to retain.  
     
     
         25 . The system of  claim 23 , wherein the operations further include presenting on the display device, in response to a request for a high power image of a first region, at least a portion of one or more of the high-power images, the at least a portion corresponding to a second region with similar characteristics to the first region according to the cluster analysis.  
     
     
         26 . The system of  claim 25 , wherein the operations further include updating the knowledge-base according to user input, wherein the updated knowledge-base affects future applications of the cluster analysis for the test.  
     
     
         27 . An apparatus comprising: 
 an interface configured to connect with a microscope; and    a controller configured to send signals through the interface to operate the microscope and to perform operations including:    obtaining an image of at least a portion of a scan region including a biological specimen;    partitioning the obtained image into zelles;    determining one or more parameters of the zelles;    performing a cluster analysis on the one or more parameters of the zelles;    differentiating tissue of greater interest from tissue of lesser interest in the obtained image based on the cluster analysis and based on a test being performed for the biological specimen; and    storing more information for the tissue of greater interest than information for the tissue of lesser interest.    
     
     
         28 . The apparatus of  claim 27 , wherein the cluster analysis comprises a multivariate statistical cluster analysis, and the zelles comprise test-dependent zelles.  
     
     
         29 . The apparatus of  claim 27 , wherein the operations further include segmenting the zelles into clusters exhibiting similar characteristics among a portion of the one or more parameters determined to best cluster the zelles according to the cluster analysis.  
     
     
         30 . The apparatus of  claim 29 , wherein the operations further include determining which clusters contain zelles that most likely contain content that is valuable to a pathologist in making a diagnostic evaluation.  
     
     
         31 . The apparatus of  claim 29 , wherein the operations further include determining how many high-power images of zelles to retain for each cluster based on a knowledge-base codifying previous test experience.  
     
     
         32 . The apparatus of  claim 31 , wherein the operations further include capturing the high-power images of zelles, analyzing the captured high-power images to determine if they meet specified criteria, and terminating the capturing once a sufficient number of high-power images have been acquired for the test according to the determining how many high-power images of zelles to retain.  
     
     
         33 . The apparatus of  claim 31 , wherein the operations further include presenting on the display device, in response to a request for a high power image of a first region, at least a portion of one or more of the high-power images, the at least a portion corresponding to a second region with similar characteristics to the first region according to the cluster analysis.  
     
     
         34 . The apparatus of  claim 33 , wherein the operations further include updating the knowledge-base according to user input, wherein the updated knowledge-base affects future applications of the cluster analysis for the test.