IP Library › Granted Patent US 12,488,603
Granted Patent B2
US 12,488,603 · App. 17/720,678 · Granted Dec 2, 2025

Systems and methods for automatically identifying features of a cytology specimen

Inventors: Hamid Reza Tizhoosh (Waterloo, CA); Seyed Rohollah Moosavitayebi (Oshawa, CA); Clinton James Vaughan Campbell (Oakville, CA)
Assignees: HAMID REZA TIZHOOSH; CLINTON JAMES VAUGHAN CAMPBELL
G06V20/69G06T7/0012G06V10/25G06V10/761G06V10/82G06T2207/20084G06T2207/30024
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,488,603
App. No.
17/720,678
Granted
Dec 2, 2025
Kind
B2
Abstract

Computer-implemented methods and systems are provided for automatically identifying features of a cytology specimen. An example method can involve dividing a whole slide image of the cytology specimen into a plurality of image portions; applying a first neural network to each image portion to identify one or more relevant image portions; and applying a second neural network to a first relevant image portion and a second relevant image portion to generate a respective first and second cell data. Each relevant image portion can be an image portion containing a region of interest for the cytology specimen. The method can further involve comparing the first and second cell data to determine whether a similarity threshold is satisfied; and in response to determining the similarity threshold is not satisfied, continuing to apply the second neural network to a subsequent relevant image portion until the similarity threshold is satisfied.

Claims (62)

1 . A method of improving a feature identification process for a cytology specimen, the method comprising:

dividing a digital whole slide image of the cytology specimen into a plurality of image portions;

evaluating each image portion of the plurality of image portions with a feature detection neural network trained to identify one or more relevant image portions from the plurality of image portions that contain a region of interest for the cytology specimen, whereby reducing a number of image portions from being further processed when absent the region of interest for the cytology specimen;

automatically selecting a base set of relevant image portions from the one or more relevant image portions, the base set of relevant image portions containing an initial set of relevant image portions for identifying features for the cytology specimen;

evaluating the base set of relevant image portions with a feature identification neural network trained to generate a base cell data comprising a predicted feature type for each feature identified within that image portion and an accuracy likelihood associated with the predicted feature type;

evaluating one or more relevant image portions outside the base set of relevant image portions with the feature identification neural network to generate a supplemental cell data comprising the predicted feature type for each feature within the one or more relevant image portions and the associated accuracy likelihood for the predicted feature type;

determining whether the supplemental cell data and the base cell data satisfy a similarity threshold indicative that the base cell data is sufficiently representative of features present in the cytology specimen;

in response to determining the similarity threshold is not satisfied:

updating the base set to include the supplemental cell data, and generating the updated base cell data with the updated base set with the feature identification neural network; and

continuing to apply the feature identification neural network to another one or more relevant image portions to generate the supplemental cell data, and determining whether the supplemental cell data and the updated base cell data satisfy the similarity threshold until the similarity threshold is satisfied; and

otherwise, providing the feature types portion identified for the cytology specimen.

2 . The method of claim 1 , wherein determining whether the supplemental cell data and the base cell data satisfy the similarity threshold comprises:

applying a statistical model to the supplemental cell data and the base cell data to determine whether the similarity threshold is satisfied.

3 . The method of claim 2 , wherein applying the statistical model comprises:

determining a chi-square distance between the supplemental cell data and the base cell data.

4 . The method of claim 1 further comprises generating a histogram to represent each cell data.

5 . The method of claim 1 , wherein the predicted feature type comprises a predicted cell type.

6 . The method of claim 1 , wherein the predicted feature type comprises a predicted non-cellular type.

7 . The method of claim 1 , wherein the feature detection neural network is trained to identify the regions of interest suitable for identifying features of the cytology specimen.

8 . The method of claim 1 , wherein the feature identification neural network is trained to detect and identify the features within the cytology specimen.

9 . A system for improving a feature identification process for a cytology specimen, the system comprising a processor operable to:

divide a digital whole slide image of the cytology specimen into a plurality of image portions;

evaluate each image portion of the plurality of image portions with a feature detection neural network trained to identify one or more relevant image portions from the plurality of image portions that contain a region of interest for the cytology specimen, whereby reducing a number of image portions from being further processed when absent the region of interest for the cytology specimen;

automatically select a base set of relevant image portions from the one or more relevant image portions, the base set of relevant image portions containing an initial set of relevant image portions for identifying features for the cytology specimen;

evaluate the base set of relevant image portions with a feature identification neural network trained to generate a base cell data comprising a predicted feature type for each feature identified within that image portion and an accuracy likelihood associated with the predicted feature type;

evaluate one or more relevant image portions outside the base set of relevant image portions with the feature identification neural network to generate a supplemental cell data comprising the predicted feature type for each feature within the one or more relevant image portions and the associated accuracy likelihood for the predicted feature type;

determine whether the supplemental cell data and the base cell data satisfy a similarity threshold indicative that the base cell data is sufficiently representative of features present in the cytology specimen;

in response to determining the similarity threshold is not satisfied:

update the base set to include supplemental cell data, and generate the updated base cell data with the updated base set with the feature identification neural network; and

continue to apply the feature identification neural network to another one or more relevant image portions to generate the supplemental cell data, and determine whether the supplemental cell data and the updated base cell data satisfy the similarity threshold until the similarity threshold is satisfied; and

otherwise, provide the feature types identified for the cytology specimen.

10 . The system of claim 9 , wherein the processor is operable to:

apply a statistical model to the supplemental cell data and the base cell data to determine whether the similarity threshold is satisfied.

11 . The system of claim 10 , wherein the processor is operable to determine a chi-square distance between the supplemental cell data and the base cell data.

12 . The system of claim 9 , wherein the processor is operable to generate a histogram to represent each cell data.

13 . The system of claim 9 , wherein the predicted feature type comprises a predicted cell type.

14 . The system of claim 9 , wherein the predicted feature type comprises a predicted non-cellular type.

15 . The system of claim 9 , wherein the first feature detection neural network is trained to identify the regions of interest suitable for identifying features of the cytology specimen.

16 . The system of claim 9 , wherein the feature identification neural network is trained to detect and identify the features within the cytology specimen.

17 . A method of improving a feature identification process for a cytology specimen, the method comprising:

dividing a digital whole slide image of the cytology specimen into a plurality of image portions;

evaluating each image portion of the plurality of image portions with a feature detection neural network trained to identify one or more relevant image portions from the plurality of image portions that contain a region of interest for the cytology specimen, whereby reducing a number of image portions from being further processed when absent the region of interest for the cytology specimen;

evaluating a first relevant image portion and a second relevant image portion with a feature identification neural network trained to generate a respective first and second cell data comprising a predicted feature type for each feature identified within that image portion and an accuracy likelihood associated with the predicted feature type;

determining whether the first cell data and the second cell data satisfy a similarity threshold indicative that the cell data is sufficiently representative of features present in the cytology specimen;

in response to determining the similarity threshold is not satisfied;

updating the first cell data to include the second cell data;

continuing to apply the feature identification neural network to a subsequent relevant image portion of the one or more relevant image portions until the similarity threshold is satisfied; and

otherwise, providing the feature types identified for the cytology specimen.

18 . The method of claim 17 , wherein determining whether the first cell data and the second cell data satisfy the similarity threshold comprises:

applying a statistical model to the first cell data and the second cell data to determine whether the similarity threshold is satisfied.

19 . The method of claim 18 , wherein applying the statistical model comprises determining a chi-square distance between the first cell data and the second cell data.

20 . A system for improving a feature identified process for a cytology specimen, the system comprising a processor operable to:

divide a digital whole slide image of the cytology specimen into a plurality of image portions;

evaluate each image portion of the plurality of image portions with a feature detection neural network trained to identify one or more relevant image portions from the plurality of image portions that contain a region of interest for the cytology specimen, whereby reducing a number of image portions from being further processed when absent the region of interest for the cytology specimen;

evaluate a first relevant image portion and a second relevant image portion with a feature identification neural network trained to generate a respective first and second cell data comprising a predicted feature type for each feature identified within that image portion and an accuracy likelihood for associated with the predicted feature type;

determine whether the first cell data and the second cell data satisfy a similarity threshold indicative that the cell data is sufficiently representative of features present in the cytology specimen;

in response to determining the similarity threshold is not satisfied:

update the first cell data to include the second cell data;

continue to apply the feature identification neural network to a subsequent relevant image portion of the one or more relevant image portions until the similarity threshold is met; and

otherwise, provide the feature types identified for the cytology specimen.

21 . The system of claim 20 , wherein the processor is operable to apply a statistical model to the first cell data and the second cell data to determine whether the similarity threshold is satisfied.

22 . The system of claim 21 , wherein the processor is operable to determine a chi-quare distance between the first cell data and the second cell data.

Assignments (3)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Mar 9, 2023
From: PARAPIXEL DIAGNOSTICS INC.
To: TIZHOOSH, HAMID REZA; CAMPBELL, CLINTON JAMES VAUGHAN
Reel/Frame 062935/0452 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 15, 2022
From: TIZHOOSH, HAMID REZA; MOOSAVITAYEBI, SEYED ROHOLLAH; CAMPBELL, CLINTON JAMES VAUGHAN
To: TIZHOOSH, HAMID REZA; CAMPBELL, CLINTON JAMES VAUGHAN
Reel/Frame 061106/0205 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Sep 15, 2022
From: TIZHOOSH, HAMID REZA; CAMPBELL, CLINTON JAMES VAUGHAN
To: PARAPIXEL DIAGNOSTICS INC.
Reel/Frame 061106/0257 →
Continuity (2)
Provisional Application 63175819 · Apr 16, 2021
Related Publication 20220335736A1 · Oct 20, 2022
References Cited (86)
US 20220108123A1 · Bhargava · 2022 [cited by examiner]
Gupta, M., Das, C., Roy, A., Gupta, P., Pillai, G. R., & Patole, K. (2020, April). Region of interest identification for cervical cancer images. In 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI) (pp… [cited by examiner]
Golden, J. A. Deep Learning Algorithms for Detection of Lymph Node Metastases From Breast Cancer. Jama 318:22, 2184-2186 (2017). [cited by applicant]
Bejnordi, B. E. et al. Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer. Jama 318:22, 2199-2210 (2017). [cited by applicant]
McKinney, S. M. et al. International evaluation of an AI system for breast cancer screening. Nature 577, 89-94 (2020). [cited by applicant]
Xu, Y. et al. Large scale tissue histopathology image classification, segmentation, and visualization via deep convolutional activation features. BMC Bioinformatics (2017) 18:281. [cited by applicant]
Sirinukunwattana, K. et al. Gland segmentation in colon histology images: The glas challenge contest. Med Image Anal 35, 489-502 (2017). [cited by applicant]
Korbar, B. et al. Deep learning for classification of colorectal polyps on whole-slide images. J Pathology Informatics 1:30 (2017). [cited by applicant]
Esteva, A. et al. Dermatologist-level classification of skin cancer with deep neural networks. Nature 542(7639), 115-118 (2017). [cited by applicant]
Bulten, W. et al. Unsupervised Prostate Cancer Detection on H&E using Convolutional Adversarial Autoencoders. Arxiv (2018). [cited by applicant]
Nguyen, A. N. et al. Symbolic rule-based classification of lung cancer stages from free-text pathology reports. Joumal of the American Medical Informatics Association 17:440-445, (2010). [cited by applicant]
Faust, K. et al. Intelligent feature engineering and ontological mapping of brain tumour histomorphologies by deep learning. Nat Mach Intell 1, 316-321 (2019). [cited by applicant]
Barker, J. et al. Automated classification of brain tumor type in whole-slide digital pathology images using local representative tiles. Medical Image Analysis 30, 60-71 (2016). [cited by applicant]
Nagpal, K. et al. Development and validation of a deep learning algorithm for improving Gleason scoring of prostate cancer. Npj Digital Medicine 2:48 (2019). [cited by applicant]
Masmoudi, M. et al. Object Detection Learning Techniques for Autonomous Vehicle Applications. 2019 IEEE Int Conf Veh Electron Saf (ICVES) 00, 1-5 (2019). [cited by applicant]
Masita, K. L. et al. Deep Learning in Object Detection: a Review. 2020 Int Conf Artif Intell Big Data Comput Data Commun Syst Icabcd 00, 1-11 (2020). [cited by applicant]
Girshick, R. et al. Rich feature hierarchies for accurate object detection and semantic segmentation. Arxiv (2013). [cited by applicant]
Ren, S. et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. Arxiv (2015). [cited by applicant]
Redmon, J. et al. You Only Look Once: Unified, Real-Time Object Detection. 2016 IEEE Conf Comput Vis Pattem Recognit Cvpr 779-788 (2016). [cited by applicant]
C, D. H. et al. An Overview of You Only Look Once: Unified, Real-Time Object Detection. Int J Res Appl Sci Eng Technology (IJRASET) 8:VI, 607-609 (2020). [cited by applicant]
Bochkovskiy, A. et al. YOLOv4: Optimal Speed and Accuracy of Object Detection. Arxiv (2020). [cited by applicant]
Laroca, R. et al. A Robust Real-Time Automatic License Plate Recognition Based on the YOLO Detector. 2018 Int Jt Conf Neural Networks Ijcnn 00, 1-10 (2018). [cited by applicant]
Ren, P. et al. A Novel YOLO-Based Real-Time People Counting Approach. 2017 Int Smart Cities Conf Isc2 1-2 (2017). [cited by applicant]
Wang. C-Y. et al. CSPNet: A new backbone that can enhance learning capability of CNN. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshop (CVPR Workshop), 2020. [cited by applicant]
He, K. et al. Spatial pyramid pooling in deep convolutional networks for visual recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 37(9):1904-1916. 2015. [cited by applicant]
Liu, S. et al. Path aggregation network for instance segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8759-8768, 2018. [cited by applicant]
Redmon, J. et al. YOLOv3: An incremental improvement. arXiv preprint arXiv:1804.02767, 2018. [cited by applicant]
Zheng, Z. et al. Distance-loU Loss: Faster and better learning for bounding box regression. In Proceedings of the AAAI Conference on Artificial Intelligence (AAAI-20), 2020. [cited by applicant]
Misra, D. Mish: A self regularized non-monotonic neural activation function. arXiv preprint arXiv:1908.08681, 2019. [cited by applicant]
Ghiasi, G. et al. DropBlock: A regularization method for convolutional networks. In Advances in Neural Information Processing Systems (NIPS), pp. 10727-10737, 2018. [cited by applicant]
Loshchilov, I. et al. SGDR: Stochastic gradient descent with warm restarts. arXiv preprint arXiv:1608.03983, 2016. [cited by applicant]
Yun, S. et al. CutMix: Regularization strategy to train strong classifiers with localizable features. In Proceedings of the IEEE International Conference on Computer Vision (ICCV), pp. 6023-6032, 2019. [cited by applicant]
Huang, G. et al. Densely connected convolutional networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 4700-4708, 2017. [cited by applicant]
Everingham, M. et al. “The PASCAL Visual Object Classes (VOC) Challenge,” Int'l J. Computer Vision, vol. 88, No. 2, pp. 303-338, Jun. 2010. [cited by applicant]
Liu, H. et al. Bone Marrow Cells Detection: A Technique for the Microscopic Image Analysis. J Med Syst 43:82 (2019). [cited by applicant]
Choi, J. W. et al. White blood cell differential count of maturation stages in bone marrow smear using dual-stage convolutional neural networks. PloS One. 12(12):e0189259 (2017). [cited by applicant]
Challen, R. et al. Artificial intelligence, bias and clinical safety. BMJ Qual Saf 28, 231 (2019). [cited by applicant]
Lee, S. et al. ICSH guidelines for the standardization of bone marrow specimens and reports. Int J Lab Hematol 30, 349 364 (2008). [cited by applicant]
Bain, B. J. Bone marrow trephine biopsy. Journal of clinical pathology 54, 737 742 (2001). [cited by applicant]
Bain, B. J. Bone marrow aspiration. Journal of clinical pathology 54, 657 663 (2001),. [cited by applicant]
Ridgeway, J. A. et al. Practical Guide to Bone Marrow Sampling for Suspected Myelodysplastic Syndromes. Journal of the Advanced Practitioner in Oncology 8, (2017). [cited by applicant]
Bain, B. J. The bone marrow aspirate of healthy subjects. Brit J Haematol 94, 206-209 (1996). [cited by applicant]
Abdulrahman, A. A. et al. Is a 500-Cell Count Necessary for Bone Marrow Differentials? American Journal of Clinical Pathology 150, (2018). [cited by applicant]
Arber, D. A. et al. Initial Diagnostic Workup of Acute Leukemia: Guideline From the College of American Pathologists and the American Society of Hematology. Arch Pathol Lab Med (2017). [cited by applicant]
Döhner, H. et al. Diagnosis and management of AML in adults: 2017 ELN recommendations from an international expert panel. Blood 129, 424 447 (2017). [cited by applicant]
Montalban-Bravo, G. et al. Myelodysplastic syndromes: 2018 update on diagnosis, risk-stratification and management. Am J Hematol 93, 129 147 (2018). [cited by applicant]
Sasada, K. et al. Inter-observer variance and the need for standardization in the morphological classification of myelodysplastic syndrome. Leukemia Res 69, 54 59 (2018). [cited by applicant]
Font, P. et al. Interobserver variance in myelodysplastic syndromes with less than 5% bone marrow blasts: unilineage vs. multilineage dysplasia and reproducibility of the threshold of 2 % blasts. Ann Hematol 94, 565-573… [cited by applicant]
Porta, D. M. et al. Minimal morphological criteria for defining bone marrow dysplasia: a basis for clinical implementation of WHO classification of myelodysplastic syndromes. Leukemia 29, 66 (2015). [cited by applicant]
Goasguen, J. E. et al. Proposal for refining the definition of dysgranulopoiesis in acute myeloid leukemia and myelodysplastic syndromes, Leukemia Res 38, 447-453 (2014). [cited by applicant]
Font, P. et al. Inter-observer variance with the diagnosis of myelodysplastic syndromes (MDS) following the 2008 WHO classification. Ann Hematol 92, 19-24 (2013). [cited by applicant]
Senent, L. et al. Reproducibility of the World Health Organization 2008 criteria for myelodysplastic syndromes. Haematologica 98, 568-575 (2013). [cited by applicant]
Parmentier, S. et al. Assessment of dysplastic hematopoiesis: lessons from healthy bone marrow donors. Haematologica 97, 723-730 (2012). [cited by applicant]
Naqvi, K. et al. Implications of discrepancy in morphologie diagnosis of myelodysplastic syndrome between referral and tertiary care centers. Blood 118, 4690-4693 (2011). [cited by applicant]
Chang, H. et al. Artificial Intelligence in Pathology. J Pathology Transl Medicine 53, 1-12 (2018). [cited by applicant]
Lecun, Y. et al. Deep learning. Nature 521, 436 (2015). [cited by applicant]
Russakovsky, O. et al. ImageNet Large Scale Visual Recognition Challenge. Int J Comput Vision 115, 211-252 (2015). [cited by applicant]
Tizhoosh, H. et al. Artificial intelligence and digital pathology: Challenges and opportunities. J Pathology Informatics 9, 38 (2018). [cited by applicant]
Djuric, U. et al. Precision histology: how deep learning is poised to revitalize histomorphology for personalized cancer care. Npj Precis Oncol 1, 22 (2017). [cited by applicant]
Janowczyk, A. et al., A. Deep learning for digital pathology image analysis: A comprehensive tutorial with selected use cases. J Pathology Informatics 7, 29 (2016). [cited by applicant]
Kutlu, H. et al. White blood cells detection and classification based on regional convolutional neural networks. Med Hypotheses 135 (2020) 109472. [cited by applicant]
Hegde, R. B. et al. Development of a Robust Algorithm for Detection of Nuclei and Classification of White Blood Cells in Peripheral Blood Smear Images. J Med Syst 2018;42(6):110. [cited by applicant]
Neelam, S. et al. Automation of differential blood count. Proc. IEEE Conference on Convergent Technologies for the Asia-Pacific Region, Bangalore, India, Oct. 15-17, 2003, 2, 547-551. [cited by applicant]
Nisha, R. et al. Isolation and two-step classification of normal white blood cells in peripheral blood smears. J Pathol Informatics 2012:3:13. [cited by applicant]
Arslan, S. et al. A color and shape based algorithm for segmentation of white blood cells in peripheral blood and bone marrow images. Cytometry Part A;85:480-90 (2014). [cited by applicant]
Madhloom, H. T. et al. An image processing application for the localization and segmentation of lymphoblast cell using peripheral blood images. J Med Syst (2012) 36:2149-2158. [cited by applicant]
Hiremath, P. S. et al. Automated identification and classification of white blood cells (leukocytes) in digital microscopic images. Int J Comput Appl, Special Issue RTIPPR (2010) 59-63. [cited by applicant]
Sawsan, F.B. et al. Segmentation and classification of white blood cells. Proc IEEE Int Conf Acoust Speech Signal Process (2000) 2259-2261. [cited by applicant]
Nazlibilek, S. et al. Automatic segmentation, counting, size determination and classification of white blood cells. Measurement 55 (2014) 58-65. [cited by applicant]
Rezatofighi, S. H. et al. Automatic recognition of five types of white blood cells in peripheral blood. Comput Med Imaging Graph 35 (2011) 333-43. [cited by applicant]
Mathur, A. et al. Scalable system for classification of white blood cells from Leishman stained blood stain images. J. Pathol. Informatics 2013; 4:S15. [cited by applicant]
Chandradevan, R. et al. Machine-based detection and classification for bone marrow aspirate differential counts: initial development focusing on nonneoplastic cells. Lab Investig. Jan. 2020; 100(1):98-109. [cited by applicant]
Tôtu, B. et al. Canadian Licensure for the Use of Digital Pathology for Routine Diagnoses: One More Step Toward a New Era of Pathology Practice Without Borders. Archives Pathology Amp Laboratory Medicine 138, 302-304 (2… [cited by applicant]
Pantanowitz, L. et al. American Telemedicine Association clinical guidelines for telepathology. J Pathology Informatics 2014;5:39. [cited by applicant]
Hanna, M. G. et al. Overview of contemporary guidelines in digital pathology: what is available in 2015 and what still needs to be addressed? J Clin Pathol 68, 499 (2015). [cited by applicant]
Pantanowitz, L. et al. Validating whole slide imaging for diagnostic purposes in pathology: guideline from the College of American Pathologists Pathology and Laboratory Quality Center, Arch Pathol Lab Med 137, 1710-22 (… [cited by applicant]
Bauer, T. W. et al. Validation of Whole Slide Imaging for Primary Diagnosis in Surgical Pathology. Arch Pathol Lab Med 137, 518-524 (2013). [cited by applicant]
Steensma, D. P. Myelodysplastic syndromes current treatment algorithm 2018. Blood Cancer J (2018) 8:47. [cited by applicant]
Maassen, A. et al. Validation and proposals for a refinement of the WHO 2008 classification of myelodysplastic syndromes without excess of blasts. Leukemia Res 37, 64-70 (2013). [cited by applicant]
Ravandi, F. et al. Evaluating measurable residual disease in acute myeloid leukemia. Blood Adv 2, 1356-1366 (2018). [cited by applicant]
Grimwade, D. et al. Defining minimal residual disease in acute myeloid leukemia: which platforms are ready for “prime time”? Blood 124, 3345-3355 (2014). [cited by applicant]
Radakovich, N. et al. Artificial Intelligence in Hematology: Current Challenges and Opportunities. Curr Hematol Malig Reports (2020) 15:203-210. [cited by applicant]
Jordan, M. I. et al. Machine learning: Trends, perspectives, and prospects. Science 349, 255-260 (2015). [cited by applicant]
Niazi, M. K. K., et al. Digital pathology and artificial intelligence. Lancet Oncol 20, e253-e261 (2019). [cited by applicant]
Araújo, T. et al. Classification of breast cancer histology images using Convolutional Neural Networks. Plos One 12(6), e0177544 (2017). [cited by applicant]
Dong, F. et al. Computational Pathology to Discriminate Benign from Malignant Intraductal Proliferations of the Breast. Plos One 9(12), e114885 (2014). [cited by applicant]