Back to results
US 12,198,300 B2Granted
Computational refocusing-assisted deep learning
AssigneesCalifornia Institute of Technology, Washington University
InventorsChanghuei Yang, Cheng Shen, Richard J. Cote, Siddarth Rawal
Life of this patent
Estimating the term…
Abstract
Computational refocusing-assisted deep learning methods, apparatus, and systems are described. In certain pathology examples, a representative image is generated using a machine learning model trained with uniformly focused training images generated by a Fourier ptychographic digital refocusing procedure and abnormalities are automatedly identified and/or enumerated based on the representative image.
Claims31
A method of computational refocusing-assisted deep learning, the method comprising:
- (a) generating a segmented image of an analysis image of a specimen using a machine learning model trained by a first training dataset comprising one or more substantially uniformly focused images generated using a Fourier ptychography digital refocusing procedure, the one or more substantially uniformly focused images of the first training dataset annotated to demarcate one or more boundaries of known portions of interest;
- and
- (b) automatedly identifying one or more portions of interest in the analysis image based on demarcated one or more boundaries of the one or more portions of interest on the segmented image.
A method for analyzing a cytology specimen, the method comprising:
- obtaining an all-in-focus analysis image of the cytology specimen using a Fourier ptychographic digital refocusing procedure;
- generating a segmented image of the all-in-focus analysis image based on a machine learning model;
- and
- automatedly identifying one or more points of interest in the cytology specimen based on one or more demarcation lines on the segmented image, the one or more demarcation lines corresponding to the one or more points of interest;
- wherein the machine learning model is trained by at least:
- one or more all-in-focus training images generated by the Fourier ptychographic digital refocusing procedure;
- and
- at least one training segmented image indicative of positions of points of interest in the one or more all-in-focus training images.
5 more elements
A method for identifying points of interest in a specimen, the method comprising:
- obtaining an analysis image of the specimen;
- generating a segmented image of the analysis image obtained based on a machine learning model;
- and
- automatedly identifying one or more points of interest in the specimen based on one or more demarcation boundaries of the one or more points of interest on the segmented image;
- wherein the machine learning model comprises an encoder-decoder network trained by at least:
- receiving at least one substantially uniformly focused training image determined based on digitally refocused images at different lateral positions;
- and
- generating at least one training segmented image indicative of positions of points of interest in the at least one substantially uniformly focused training image.
5 more elements
A method of training a machine learning model for identifying abnormalities in a specimen, the method comprising:
- using a Fourier ptychographic digital refocusing procedure to determine at least one substantially uniformly focused training image;
- generating, using an encoder-decoder convolutional network, at least one training segmented image indicative of positions of abnormalities in the at least one substantially uniformly focused training image;
- and
- generating a trained machine learning model based at least on the at least one training segmented image, the trained machine learning model configured to, during inference:
- receive an analysis image of the specimen generated using the Fourier ptychographic digital refocusing procedure;
- generate a segmented image comprising one or more image segments corresponding to one or more categories of tissue;
- and
- automatedly identify one or more abnormalities in the received analysis image based at least on a probability of one or more pixels in the segmented image exceeding a threshold.
5 more elements
An apparatus for identifying abnormalities in a specimen, the apparatus comprising:
- a machine learning model;
- one or more processor apparatus configured to operate the machine learning model;
- and
- a non-transitory computer-readable apparatus coupled to the one or more processor apparatus and comprising a storage medium, the storage medium comprising a plurality of instructions configured to, when executed by the one or more processor apparatus, cause the apparatus to:
- obtain an analysis image of the specimen;
- generate a segmented image of the analysis image obtained of the specimen using the machine learning model, the machine learning model trained by (i) generation of one or more convolutional representations of at least one substantially uniformly focused image obtained using a Fourier ptychography digital refocusing procedure, and (ii) generation of at least one training segmented image based on the one or more convolutional representations of the at least one substantially uniformly focused training image;
- and
- based on one or more demarcated boundaries, in the segmented image, of one or more image segments determined to correspond to one or more abnormalities, automatedly identify one or more abnormalities in the specimen.
5 more elements
Dependencies are inferred from the claim text.