ARTIFICIAL INTELLIGENCE FOR DIGITAL PATHOLOGY. Kyunghyun Paeng, Co-founder and Research Scientist, Lunit Inc.

Similar documents
arxiv: v1 [cs.cv] 30 May 2018

Y-Net: Joint Segmentation and Classification for Diagnosis of Breast Biopsy Images

Highly Accurate Brain Stroke Diagnostic System and Generative Lesion Model. Junghwan Cho, Ph.D. CAIDE Systems, Inc. Deep Learning R&D Team

Digital Pathology - moving on after implementation. Catarina Eloy, MD, PhD

Combination of life science entrepreneurs, software development and machine vision experts & recognized scientists.

Methods for Segmentation and Classification of Digital Microscopy Tissue Images

ITERATIVELY TRAINING CLASSIFIERS FOR CIRCULATING TUMOR CELL DETECTION

Session: Imaging for Clinical Decision Support

arxiv: v1 [cs.cv] 1 Oct 2018

Deep learning approaches to medical applications

Deep learning and non-negative matrix factorization in recognition of mammograms

Classification of breast cancer histology images using transfer learning

Differentiating Tumor and Edema in Brain Magnetic Resonance Images Using a Convolutional Neural Network

Deep-Learning Based Semantic Labeling for 2D Mammography & Comparison of Complexity for Machine Learning Tasks

Visual interpretation in pathology

Skin cancer reorganization and classification with deep neural network

Background Information

3D Deep Learning for Multi-modal Imaging-Guided Survival Time Prediction of Brain Tumor Patients

POC Brain Tumor Segmentation. vlife Use Case

A Multi-Resolution Deep Learning Framework for Lung Adenocarcinoma Growth Pattern Classification

Is digital pathology ready for use in routine histopathology? Dr David Snead University Hospitals of Coventry and Warwickshire NHS Trust Coventry UK

Artificial Intelligence in Breast Imaging

Two-Stage Convolutional Neural Network for Breast Cancer Histology Image Classification

Computer based delineation and follow-up multisite abdominal tumors in longitudinal CT studies

CS231n Project: Prediction of Head and Neck Cancer Submolecular Types from Patholoy Images

Image Captioning using Reinforcement Learning. Presentation by: Samarth Gupta

Improving Network Accuracy with Augmented Imagery Training Data

Lessons learned in the use of digital imaging at Memorial Sloan Kettering Cancer Center

MRI Image Processing Operations for Brain Tumor Detection

arxiv: v1 [cs.cv] 26 Feb 2018

Healthcare Research You

arxiv: v1 [cs.cv] 21 Jul 2017

Virtual Microscopy: Express Surgical Pathology Consultation. Mercè Jordà, University of Miami, Florida

Weak Supervision. Vincent Chen and Nish Khandwala

CHAPTER 2 MAMMOGRAMS AND COMPUTER AIDED DETECTION

Tumor Cellularity Assessment. Rene Bidart

A DEEP LEARNING METHOD FOR DETECTING AND CLASSIFYING BREAST CANCER METASTASES IN LYMPH NODES ON HISTOPATHOLOGICAL IMAGES

arxiv: v1 [stat.ml] 23 Jan 2017

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS, VOL. X, NO. Y, MONTH YEAR. 1

Supplemental Information

A REVIEW ON CLASSIFICATION OF BREAST CANCER DETECTION USING COMBINATION OF THE FEATURE EXTRACTION MODELS. Aeronautical Engineering. Hyderabad. India.

Automatic Classification of Breast Masses for Diagnosis of Breast Cancer in Digital Mammograms using Neural Network

CSE Introduction to High-Perfomance Deep Learning ImageNet & VGG. Jihyung Kil

Lung Region Segmentation using Artificial Neural Network Hopfield Model for Cancer Diagnosis in Thorax CT Images

Automated Blood Vessel Extraction Based on High-Order Local Autocorrelation Features on Retinal Images

HHS Public Access Author manuscript Med Image Comput Comput Assist Interv. Author manuscript; available in PMC 2018 January 04.

HEALTHCARE AI DEVELOPMENT CYCLE

Automatic Prostate Cancer Classification using Deep Learning. Ida Arvidsson Centre for Mathematical Sciences, Lund University, Sweden

CHS-NET: A Cascaded Neural Network with Semi-Focal Loss for Mitosis Detection

Critical reading of diagnostic imaging studies. Lecture Goals. Constantine Gatsonis, PhD. Brown University

A Deep Learning Approach for Breast Cancer Mass Detection

Roadmap for Developing and Validating Therapeutically Relevant Genomic Classifiers. Richard Simon, J Clin Oncol 23:

Machine Learning in Precision Medicine Coronary Health Prediction - Cardiac Events (Atherosclerosis) - Heart Transplant (Vasculopathy)

Final Report: Automated Semantic Segmentation of Volumetric Cardiovascular Features and Disease Assessment

Final Project Report Sean Fischer CS229 Introduction

Automatic Extraction of Synoptic Data. George Cernile Artificial Intelligence in Medicine AIM

Supplementary Online Content

An Artificial Neural Network Architecture Based on Context Transformations in Cortical Minicolumns

Abstract. Background. Objective

Automated diagnosis of pneumothorax using an ensemble of convolutional neural networks with multi-sized chest radiography images

arxiv: v1 [cs.lg] 4 Feb 2019

EARLY STAGE DIAGNOSIS OF LUNG CANCER USING CT-SCAN IMAGES BASED ON CELLULAR LEARNING AUTOMATE

INTRODUCTION TO MACHINE LEARNING. Decision tree learning

Outline (1) Outline (2) Concepts in Prostate Pathology. Peculiarities of Prostate Cancer. Peculiarities of Prostate Cancer

Beyond the DVH: Voxelbased Automated Planning and Quality Assurance

Multi-task Learning of Dish Detection and Calorie Estimation

Convolutional Neural Networks (CNN)

Comparison of Two Approaches for Direct Food Calorie Estimation

arxiv: v2 [cs.cv] 19 Dec 2017

Developing ML Models for semantic segmentation of medical images

Why did the network make this prediction?

Convolutional Neural Networks for Estimating Left Ventricular Volume

Automated detection of masses on whole breast volume ultrasound scanner: false positive reduction using deep convolutional neural network

Reviewer's report. Version: 1 Date: 24 May Reviewer: Cathy Moelans. Reviewer's report:

MR-Radiomics in Neuro-Oncology

arxiv: v2 [cs.cv] 8 Mar 2018

Inferring Clinical Correlations from EEG Reports with Deep Neural Learning

Automated Estimation of mts Score in Hand Joint X-Ray Image Using Machine Learning

Cancer Cells Detection using OTSU Threshold Algorithm

NMF-Density: NMF-Based Breast Density Classifier

Flat Epithelial Atypia

MITOS & ATYPIA Detection of Mitosis and Evaluation of Nuclear Atypia Score in Breast Cancer Histological Images

The Human Behaviour-Change Project

Patch-based Head and Neck Cancer Subtype Classification

ANALYSIS AND DETECTION OF BRAIN TUMOUR USING IMAGE PROCESSING TECHNIQUES

Early Detection of Lung Cancer

Efficient Deep Model Selection

FAQs for UK Pathology Departments

Computational Pathology

AN ALGORITHM FOR EARLY BREAST CANCER DETECTION IN MAMMOGRAMS

AN EFFICIENT DIGITAL SUPPORT SYSTEM FOR DIAGNOSING BRAIN TUMOR

Segmentation of Tumor Region from Brain Mri Images Using Fuzzy C-Means Clustering And Seeded Region Growing

Mammography limitations. Clinical performance of digital breast tomosynthesis compared to digital mammography: blinded multi-reader study

Contributions to Brain MRI Processing and Analysis

Convolutional Neural Networks for Text Classification

TMIST A Bridge to Personalized Screening. Canadian Society of Breast Imaging April 26, 2018

Clustering of MRI Images of Brain for the Detection of Brain Tumor Using Pixel Density Self Organizing Map (SOM)

Transcription:

ARTIFICIAL INTELLIGENCE FOR DIGITAL PATHOLOGY Kyunghyun Paeng, Co-founder and Research Scientist, Lunit Inc.

1. BACKGROUND: DIGITAL PATHOLOGY 2. APPLICATIONS AGENDA BREAST CANCER PROSTATE CANCER 3. DEMONSTRATIONS 4. CONCLUSION 2

BACKGROUND: DIGITAL PATHOLOGY DIAGNOSTIC PROCEDURE Patient Detection (X-ray, CT, MRI,...) Diagnosis (biopsy, resection,...) Treatment Radiology Pathology Oncology 3

BACKGROUND: DIGITAL PATHOLOGY LIMITATIONS OF CONVENTIONAL PATHOLOGY Slide Report Diagnosis (biopsy, resection,...) Pathology (-) Archiving (-) Workflow (-) Analysis 4

BACKGROUND: DIGITAL PATHOLOGY RISE OF DIGITAL PATHOLOGY Diagnosis (biopsy, resection,...) Pathology (+) Archiving (+) Workflow (+) Analysis Digital pathology 5

BACKGROUND: DIGITAL PATHOLOGY WHY DO WE NEED AI IN DIGITAL PATHOLOGY? (+) Reproducibility (+) Accuracy (+) Workload reduction 25% disagreement among pathologists in breast biopsy report. Diagnostic Concordance Among Pathologists Interpreting Breast Biopsy Specimens., JAMA, 2015. 6

BACKGROUND: DIGITAL PATHOLOGY CHALLENGES IN AI FOR DIGITAL PATHOLOGY ~ 100,000 pixels 1. Gigapixel images Grade 1 Grade 2 Grade 3 2. Quality variation 3. Ambiguity in ground-truth definition 3! 4! 3? 4? 7

KEY APPLICATIONS: #1. Tumor proliferation score prediction in breast resection specimen. #2. Gleason score prediction in prostate biopsy specimen. 8

APPLICATION #1: BREAST CANCER WHAT IS TUMOR PROLIFERATION SCORE? Breast resection specimen Proliferation score (in 10 consecutive HPFs) Mitosis Score 1: ~6 mitosis Score 2: 6~10 mitosis Score 3: 10~ mitosis good prognosis bad 9

APPLICATION #1: BREAST CANCER TUMOR PROLIFERATION SCORE PREDICTION Data statistics Tumor Proliferation Assessment Challenge 2016 TUPAC16 MICCAI Grand Challenge Training dataset Test dataset, Proliferation score 500 slides, Proliferation score 321 slides, Mitosis #1 (x,y)... Mitosis #N (x,y) Auxiliary dataset 656 ROIs from 73 slides 10

APPLICATION #1: BREAST CANCER TUMOR PROLIFERATION SCORE PREDICTION System overview Whole slide image... Mitosis Detection Network 1. The number of mitosis 2. The number of cells Tissue region extraction Stain normalization Patch extraction at x40 ROI detection using cell density Phase 1: Handling whole slide images Auxiliary set for mitosis detection Phase 2: Mitosis detection Feature vector based on statistical information Support Vector Machine Proliferation score Phase 3: Score prediction 11

APPLICATION #1: BREAST CANCER TUMOR PROLIFERATION SCORE PREDICTION Phase 1: Handling whole slide images Resizing a whole slide image. Finding a threshold. Morphological operations. Whole slide image... Stain normalization Patch extraction with 10 HPFs size. Cell detection in each patch. Tissue region extraction Patch extraction at x40 ROI detection using cell density 30 ROIs selection. Stain normalization. 12

APPLICATION #1: BREAST CANCER TUMOR PROLIFERATION SCORE PREDICTION Phase 2: Mitosis detection Mitosis Detection Network 128 x 128 conv 1, 3x3, 16 resblock 1.1, 3x3, 32 resblock 1.3, 3x3, 32 resblock 2.1, 3x3, 64 resblock 2.3, 3x3, 64 resblock 3.1, 3x3, 128 resblock 3.3, 3x3, 128 16 mitosis 8 normal Global pooling layer Auxiliary set for mitosis detection Based on Residual Network (ResNet). 9 residual blocks = 21 layers architecture. 2 step training procedure. Cropped global pooling layer. Training step:, Inference step: 13

APPLICATION #1: BREAST CANCER TUMOR PROLIFERATION SCORE PREDICTION Phase 3: Score prediction Converting each WSI to a 21-dim feature vector. 10-fold cross validation from 500 training samples. Feature selection based on cross validation results. 1. The number of mitosis 2. The number of cells Feature vector based on statistical information Support Vector Machine Proliferation score 14

APPLICATION #1: BREAST CANCER TUMOR PROLIFERATION SCORE PREDICTION Results Tumor Proliferation Assessment Challenge 2016 TUPAC16 MICCAI Grand Challenge 15

APPLICATION #2: PROSTATE CANCER WHAT IS GLEASON SCORE? Prostate biopsy specimen Core #1: Core #2: Core #3: Core #4: 5+5 0 3+4 0 Grade 1 Grade 2 Grade 3 Grade 4 Grade 1, 2 Grade 3 Grade 4 Grade 5 Grade 5 16

APPLICATION #2: PROSTATE CANCER GLEASON SCORE PREDICTION Data statistics Training dataset Test dataset, { Grade, Contours } 900 slides, { Grade, Contours } 50 slides The number of patients: 385 The number of slides: 1152 The number of cores: 4907 The number of normal cores: 2872 The number of cancer cores: 2035 Dataset from medical centers 17

APPLICATION #2: PROSTATE CANCER GLEASON SCORE PREDICTION System overview Patch-based classification Normal Grade 3 Grade 4 Grade 5 Gleason score classification network Normal Grade 3 Grade 4 Grade 5 Memory network-based refinement (25 neighbors) 1000 1100 1110 1111 Embedded memory vector... Query vector Embedding Ranking loss with thermometer code Memory network 18 Refined output

APPLICATION #2: PROSTATE CANCER GLEASON SCORE PREDICTION Patch-based classification Normal Grade 3 Baseline settings ResNet 101 architecture. 512x512 patch with 75% overlap. Softmax loss with 4 class classification. Key features for improving performance ~75% Grade 4 Normal patches from only fully normal slides. è +~5% gain Ranking loss with thermometer code. è +2~3% gain Grade 5 Not a classification problem! Ordering problem! 1000 1100 1110 1111 Network decodes from the left-most bit to the right-most bit. 19

APPLICATION #2: PROSTATE CANCER Patch-level outputs (25 neighbors) GLEASON SCORE PREDICTION Memory network-based refinement + ~5% gain 1D-CNN Refined output...... Memory vector (25x4dim) Query vector (1x4dim) Embedding... 25x1024......... 1x1024 Innerproduct... Weighting Attention vector 25x1 Softmax 20

APPLICATION #2: PROSTATE CANCER GLEASON SCORE PREDICTION Results Patch-level performance Baseline: 75% + Data cleansing: 80% + Ranking loss: 82.8% + Memnet refinement: 87.5% Core-level performance Normal or cancer core? AUC: 97.8% Gleason score prediction? Only 1 st major: 83% Both: 76.7% 21

DEMO #1: BREAST CANCER 22

DEMO #2: PROSTATE CANCER 23

Lessons learned CONCLUSION Artificial intelligence for digital pathology Challenge #1. How to handle gigapixel images? (i.e., whole slide images) ü Consider how to sample patches. (patch size, sampling step,...) è with pathologists. ü Consider how to construct whole pipeline from gigapixel images to diagnosis. Challenge #2. How to handle quality variation between slides? ü Design image processing modules carefully. ü Do cross-validation to avoid overfitting. Challenge #3. How to handle ambiguous ground-truth? ü Design task-specific loss. ü Sanitize training dataset as much as possible you can. ü Don t be satisfied with patch-based results. 24

THANK YOU TEAM MEMBERS