Leica Biosystems Imaging, Inc.

Aperio Technologies, Incorporated

Allen Olson

Greg J Crandall

Yunlu Zou

Dirk G Soenksen

Kiran Saligrama

Ole Eichhorn

Peyman Najmabadi

Leng-Chun Chen

Gregory J Crandall

Walter Georgescu

Bharat Annaldas

Nicholas Newberg

Aaron Alan Stearrett

Ruby Chen

Kevin Shields

Mark Gregson

Darragh Lawler

Greg Crandall

Ole Eichorn

Ruben Campos Zurriaga

Girish Mallya Udupi

Bruno Oliveira

Reiner Jung

Kiran Saligramal

Gregory Crandall

A CNN is applied to a histological image to identify areas of interest. The CNN classifies pixels according to relevance classes including one or more classes indicating levels of interest and at least one class indicating lack of interest. The CNN is trained on a training data set including data which has recorded how pathologists have interacted with visualizations of histological images. In the trained CNN, the in-terest-based pixel classification is used to generate a segmentation mask that defines areas of interest. The mask can be used to indicate where in an image clinically relevant features may be located. Further, it can be used to guide variable data compression of the histological image. Moreover, it can be used to control loading of image data in either a client-server model or within a memory cache policy. Furthermore, a histological image of a tissue sample of a tissue type that has been treated with a test compound is image processed in order to detect areas where toxic reactions to the test compound may have occurred. An autoencoder is trained with a training data set comprising histological images of tissue samples which are of the given tissue type, but which have not been treated with the test compound. The trained autoencoder is applied to detect tissue areas by their deviation from the normal variation seen in that tissue type as learnt by the training process, and so build up a toxicity map of the image. The toxicity map can then be used to direct a toxicological pathologist to examine the areas identified by the autoencoder as lying outside the normal range of heterogeneity for the tissue type. This makes the pathologist's review quicker and more reliable. The toxicity map can also be overlayed with the segmentation mask indicating areas of interest. When an area of interest and an area identified as lying outside the normal range of heterogeneity for the tissue type, and increased confidence score is applied to the overlapping area.

Neural network based identification of areas of interest in digital pathology images

Real-time autofocus. In an embodiment, a scanning apparatus includes an imaging sensor, a focusing sensor, an objective lens, and processor(s) configured to analyze image data captured by the imaging and focusing sensors, and move the objective lens. Real-time autofocus during scanning of a sample is achieved by determining a true-Z value for the objective lens for a point on a sample and for each of a plurality of regions on the sample. The true-Z values and/or surfaces calculated therefrom are used to determine a predicted-Z value for an unscanned region of the sample. The objective lens is adjusted to the predicted-Z value at the beginning of the unscanned region. After scanning the region, a true-Z value is determined for the region and compared to the predicted-Z value. A rescan of the region is initiated if the comparison exceeds a predetermined threshold.

Real-time autofocus scanning

Systems and methods for capturing a digital image of a slide using an imaging line sensor and a focusing line sensor. In an embodiment, a beam-splitter is optically coupled to an objective lens and configured to receive one or more images of a portion of a sample through the objective lens. The beam-splitter simultaneously provides a first portion of the one or more images to the focusing sensor and a second portion of the one or more images to the imaging sensor. A processor controls the stage and/or objective lens such that each portion of the one or more images is received by the focusing sensor prior to it being received by the imaging sensor. In this manner, a focus of the objective lens can be controlled using data received from the focusing sensor prior to capturing an image of a portion of the sample using the imaging sensor.

Real-time focusing in line scan imaging

Real-time focusing in a slide-scanning system. In an embodiment, focus points are added to an initialized focus map while acquiring a plurality of image stripes of a sample on a glass slide. For each image stripe, a plurality of frames, collectively representing the image stripe, may be acquired using both an imaging line-scan camera and a tilted focusing line-scan camera. Focus points, representing positions of best focus for trusted frames, are added to the focus map. Outlying focus points are removed from the focus map. In some cases, one or more image stripes may be reacquired. Finally, the image stripes are assembled into a composite image of the sample.

Real-time focusing in a slide-scanning system

Image viewing in digital pathology using eye-tracking. In an embodiment, a position of a user's gaze on a graphical user interface, comprising at least a portion of a digital slide image within a macro view, is repeatedly detected based on an output from an eye-tracking device. After detecting a change of the user's gaze from a first position to a second position on the graphical user interface, a view of the digital slide image within the macro view is automatically panned based on the second position, so as to move a position on the digital slide image that corresponds to the second position on the graphical user interface toward a center of the macro view.

Eye-tracking image viewer for digital pathology

Color calibration for digital pathology is provided. A standard glass slide is prepared with a specimen having zero or more stains. The specimen is scanned a first time using a hyperspectral imaging system to produce a first digital image having XYZ color values. The specimen is scanned a second time using a digital pathology imaging system to produce a second digital image having RGB color values. The first and second digital images are then registered against each other to align the digital image data. Individual pixels of the first and second images may be combined in the registration process so that the first and second digital images have substantially similar pixel sizes. A lookup table is generated to associate XYZ color values to RGB color values. Once the lookup table has been generated, it can be used to present RGB color on a display using the corresponding XYZ color.

Digital pathology color calibration and validation

A digital scanning apparatus is provided that includes imaging and focusing sensors and a processor to analyze the image data captured by the imaging and focusing sensors and adjust the focus of the scanning apparatus in real time during a scanning operation. The individual pixels of the imaging sensor are all in the same image plane with respect to the optical path of the digital scanning apparatus. The individual pixels of the focusing sensor are each in a different image plane with respect to the optical path, and one pixel of the focusing sensor is on the same image plane as the image sensor. The processor analyzes image data from the imaging sensor and the focusing sensor and determines a distance and direction to adjust the relative position of an objective lens and a stage of the digital scanning apparatus to achieve optimal focus during the scanning operation.

Real-time autofocus focusing algorithm

A convolutional neural network (CNN) is applied to identifying tumors in a histological image. The CNN has one channel assigned to each of a plurality of tissue classes that are to be identified, there being at least one class for each of non-tumorous and tumorous tissue types. Multi-stage convolution is performed on image patches extracted from the histological image followed by multi-stage transpose convolution to recover a layer matched in size to the input image patch. The output image patch thus has a one-to-one pixel-to-pixel correspondence with the input image patch such that each pixel in the output image patch has assigned to it one of the multiple available classes. The output image patches are then assembled into a probability map that can be co-rendered with the histological image either alongside it or over it as an overlay. The probability map can then be stored linked to the histological image.

Processing of histology images with a convolutional neural network to identify tumors

System for acquiring a digital image of a sample on a microscope slide. In an embodiment, the system comprises a stage configured to support a sample, an objective lens having a single optical axis that is orthogonal to the stage, an imaging sensor, and a focusing sensor. The system further comprises at least one beam splitter optically coupled to the objective lens and configured to receive a field of view corresponding to the optical axis of the objective lens, and simultaneously provide at least a first portion of the field of view to the imaging sensor and at least a second portion of the field of view to the focusing sensor. The focusing sensor may simultaneously acquire image(s) at a plurality of different focal distances and/or simultaneously acquire a pair of mirrored images, each comprising pixels acquired at a plurality of different focal distances.

Automated stain finding. In an embodiment, an image of a sample comprising one or more stains is received. For each of a plurality of pixels in the image, an optical density vector for the pixel is determined. The optical density vector comprises a value for each of the one or more stains, and represents a point in an optical density space that has a number of dimensions equal to a number of the one or more stains. The optical density vectors are transformed from the optical density space into a representation in a lower dimensional space. The lower dimensional space has a number of dimensions equal to one less than the number of dimensions of the optical density space. An optical density vector corresponding to each of the one or more stains is identified based on the representation.

Automated stain finding in pathology bright-field images

Two-pass capture of a macro image. In an embodiment, a scanning apparatus comprises a stage, a high-resolution camera, and a lens that provides a field of view, substantially equal in width to a slide width, to the high-resolution camera. The apparatus also comprises a first illumination system for transmission-mode illumination, and a second illumination system for reflection-mode illumination. Processor(s) move the stage in a first direction to capture a first macro image of a specimen during a single pass while the field of view is illuminated by the first illumination system, and move the stage in a second direction to capture a second macro image of the specimen during a single pass while the field of view is illuminated by the second illumination system. The processor(s) identify artifacts in the second macro image, and, based on those artifacts, correct the first macro image to generate a modified first macro image.

Two pass macro image

Described herein are systems and methods that place a known color monitor (known by unique serial number or SKU) into a desired state for displaying digital pathology image data. Using an application programming interface, any color monitor that implements MCCS can be calibrated and characterized immediately before each display of digital pathology image data and can also be periodically reset (if needed) during display of digital pathology image data.

Growing community of inventors

San Diego, CA, United States of America

Allen Olson