ISBI 2019 C-NMC Challenge: Classification in Cancer Cell Imaging (eBook)
X, 147 Seiten
Springer Singapore (Verlag)
978-981-15-0798-4 (ISBN)
This book comprises select peer-reviewed proceedings of the medical challenge - C-NMC challenge: Classification of normal versus malignant cells in B-ALL white blood cancer microscopic images. The challenge was run as part of the IEEE International Symposium on Biomedical Imaging (IEEE ISBI) 2019 held at Venice, Italy in April 2019. Cell classification via image processing has recently gained interest from the point of view of building computer-assisted diagnostic tools for blood disorders such as leukaemia. In order to arrive at a conclusive decision on disease diagnosis and degree of progression, it is very important to identify malignant cells with high accuracy. Computer-assisted tools can be very helpful in automating the process of cell segmentation and identification because morphologically both cell types appear similar. This particular challenge was run on a curated data set of more than 14000 cell images of very high quality. More than 200 international teams participated in the challenge. This book covers various solutions using machine learning and deep learning approaches. The book will prove useful for academics, researchers, and professionals interested in building low-cost automated diagnostic tools for cancer diagnosis and treatment.
Dr. Anubha Gupta received her PhD in Electrical Engineering from Indian Institute of Technology (IIT) Delhi in 2006. She completed her second Master's as a full time student from the University of Maryland, College Park, USA from 2008-2010 in education with concentration: Higher Education Leadership and Policy Studies. She worked as Assistant Director with the Ministry of Information and Broadcasting, Government of India (through Indian Engineering Services) from 1993 to 1999 and as faculty for about 10 years before joining IIIT Delhi in December 2013, where she is currently working as Professor. She has author/co-authored over 80 technical papers in scientific journals and conferences. She has also filed 4 patents. She has published research papers in both engineering and education. Her research interests include biomedical signal and image processing including fMRI, MRI, DTI, EEG, ECG signal processing, genomics signal processing in cancer research, wavelets in deep learning, and signal processing for communication engineering. Dr. Gupta is a senior member of IEEE Signal processing Society, a member of IEEE Women in Engineering Society, and is Associate Editor of IEEE Access journal.
Dr. Ritu Gupta is Professor of Laboratory Oncology at the All India Institute of Medical Sciences (AIIMS), New Delhi. She is currently spearheading the cancer laboratories at Dr. B.R. Ambedkar IRCH, AIIMS and is actively engaged in establishing diagnostic and research laboratories at the National Cancer Institute (NCI), Jhajjar, India. Dr. Gupta and her research group have established Unit of Excellence on Multiple Myeloma at AIIMS. Her lab is investigating the genomic and epigenomic alterations responsible for disease progression and treatment response in chronic lymphocytic leukemia and the molecular basis of disease heterogeneity in multiple myeloma. She is currently evaluating the prognostic and therapeutic implications of leukemic stem cells in acute myeloid leukemia. As a hematopathologist, she has a keen interest in digital processing of tumor cells and is working on image processing based software solutions for clinical applications. She has published more than 80 papers in peer reviewed journals and is actively engaged in academic activities at the national level for training of medical fraternity on advanced laboratory diagnostics including multi-parametric flow cytometry and molecular assays for clinical diagnostics and research.This book comprises select peer-reviewed proceedings of the medical challenge - C-NMC challenge: Classification of normal versus malignant cells in B-ALL white blood cancer microscopic images. The challenge was run as part of the IEEE International Symposium on Biomedical Imaging (IEEE ISBI) 2019 held at Venice, Italy in April 2019. Cell classification via image processing has recently gained interest from the point of view of building computer-assisted diagnostic tools for blood disorders such as leukaemia. In order to arrive at a conclusive decision on disease diagnosis and degree of progression, it is very important to identify malignant cells with high accuracy. Computer-assisted tools can be very helpful in automating the process of cell segmentation and identification because morphologically both cell types appear similar. This particular challenge was run on a curated data set of more than 14000 cell images of very high quality. More than 200 international teams participated in the challenge. This book covers various solutions using machine learning and deep learning approaches. The book will prove useful for academics, researchers, and professionals interested in building low-cost automated diagnostic tools for cancer diagnosis and treatment.
Preface 6
Contents 8
About the Editors 10
Classification of Normal Versus Malignant Cells in B-ALL White Blood Cancer Microscopic Images 12
1 Introduction 12
2 Previous Works 13
3 Approach 14
3.1 Data 16
3.2 Data Preprocessing 16
3.3 Finetuning 17
4 Observations 18
4.1 Stage 1 Training 18
4.2 Stage 2 Training 18
5 Results 21
6 Conclusions 22
References 22
Classification of Leukemic B-Lymphoblast Cells from Blood Smear Microscopic Images with an Attention-Based Deep Learning Method and Advanced Augmentation Techniques 24
1 Introduction 25
2 Material 26
3 Methods 26
3.1 Data Normalization 27
3.2 Data Augmentation 27
3.3 Data Split 27
3.4 Predicting the Bounding Box 27
3.5 Neuronal Network Model Architecture 28
4 Results 30
5 Discussion 32
References 32
Classification of Normal and Leukemic Blast Cells in B-ALL Cancer Using a Combination of Convolutional and Recurrent Neural Networks 34
1 Introduction 35
2 Related Work 36
3 Methodology 36
3.1 Dataset Description and Preprocessing 36
3.2 Classification Architecture 37
4 Results 39
5 Conclusions 41
References 41
Deep Learning for Classifying of White Blood Cancer 43
1 Introduction 43
2 Related Works 44
3 Approach 45
3.1 Data Preprocessing 45
3.2 Network Architecture 46
3.3 Stacking 47
3.4 Training Details 47
4 Experiments 48
4.1 Precision, Recall and F1 Score 48
4.2 AUC and ROC Curve 48
4.3 Subject- and Class-Level Accuracy 48
5 Discussion 50
References 50
Ensemble Convolutional Neural Networks for Cell Classification in Microscopic Images 52
1 Introduction 52
2 Method 53
2.1 The Selected CNNs: SENet and PNASNet 53
2.2 Data Augmentation 53
2.3 Ensemble CNNs 53
2.4 Grad-CAM for Visualizing 54
3 Related Work 54
4 Experiments 54
4.1 Evaluation of Preliminary Testing 55
4.2 Evaluation of Final Testing 56
4.3 Comparison of Ensemble Voting 56
5 Visualizing by Grad-CAM 57
6 Conclusion 58
References 59
Acute Lymphoblastic Leukemia Classification from Microscopic Images Using Convolutional Neural Networks 61
1 Introduction 61
2 Dataset 62
3 Network Architecture 63
4 Experiments 64
4.1 Training and Testing 64
4.2 Model Selection 65
4.3 Results on the Final Test Set 66
4.4 Ablation Studies 66
5 Related Work 67
6 Conclusion 68
References 68
Toward Automated Classification of B-Acute Lymphoblastic Leukemia 70
1 Introduction 71
1.1 Dataset 71
2 Non-i.i.d. Assumption 71
2.1 Data Splitting 71
2.2 Plain Feedforward Networks 72
2.3 Optimizing Jensen–Shannon Distance 72
2.4 Siamese Networks for Patient Differentiation 73
3 i.i.d. Assumption 73
3.1 Data Preparation and Augmentation 73
3.2 Autoencoder-Based Approach 74
3.3 Ensemble of CNNs and Data Preprocessing 74
4 Phase 2 Training 75
4.1 Class Weights for Cost-Sensitive Training 75
4.2 Augmentation Using GANs 75
5 Results 76
6 Conclusion 77
References 77
Neighborhood-Correction Algorithm for Classification of Normal and Malignant Cells 80
1 Introduction 81
2 Materials 82
3 Method 82
3.1 Fine-Tuning ResNets 83
3.2 Combined ResNet-FV for Cell Representation 83
3.3 Label Correction 84
4 Experiments and Discussions 85
4.1 Cross-Validation of Baseline Method 85
4.2 Performance Gain Caused by Label Correction 85
4.3 Results on Phase-II Dataset 86
4.4 Feedback to BM 87
5 Conclusion 88
References 89
DeepMEN: Multi-model Ensemble Network for B-Lymphoblast Cell Classification 90
1 Introduction 91
2 Related Work 93
3 Methodology 93
3.1 Network Architecture 93
3.2 Pseudo Labeling 95
3.3 Test Time Augmentor 96
3.4 Training Details 96
4 Experiment 96
4.1 Datasets 96
4.2 Image Preprocessing 97
4.3 Result 97
5 Conclusion 98
References 99
Multi-streams and Multi-features for Cell Classification 101
1 Introduction 102
2 Dataset Description 102
2.1 Dataset 102
2.2 Pre-processing 103
3 Network Architecture 103
4 Experiments 105
4.1 Implementation 105
4.2 Evaluation Criterion 105
4.3 Feature Visualization 105
4.4 Experimental Results 105
5 Conclusion 107
References 107
Classification of Normal Versus Malignant Cells in B-ALL Microscopic Images Based on a Tiled Convolution Neural Network Approach 109
1 Introduction 109
2 The Proposed Methodology 110
2.1 Training–Validation Splitting 110
2.2 The Machine Learning Pipeline 111
2.3 Pooling the Predictions of the Networks 112
2.4 Evaluation Criterion 114
3 Results 114
4 Discussion and Conclusion 115
References 116
Acute Lymphoblastic Leukemia Cells Image Analysis with Deep Bagging Ensemble Learning 118
1 Introduction 118
2 Methods 120
2.1 Details of Initial Dataset 120
2.2 Images Preprocessing 121
2.3 Training Dataset Generation 121
2.4 Training Strategy and ``Combined'' Model Architecture 122
2.5 Training Settings 123
2.6 Inference 123
3 Results 123
4 Discussions and Conclusion 124
References 125
Leukemic B-Lymphoblast Cell Detection with Monte Carlo Dropout Ensemble Models 127
1 Introduction 127
2 Method 128
2.1 Data and Preprocessing 128
2.2 Monte Carlo Dropout 129
2.3 Ensemble by Uncertainty 130
2.4 Implementation Details 131
3 Results 131
4 Discussion 133
References 133
ISBI Challenge 2019: Convolution Neural Networks for B-ALL Cell Classification 135
1 Introduction 135
2 Data Description 136
3 Method 136
3.1 Data Preprocessing 136
3.2 Transfer Learning 137
3.3 Base Classifier 138
3.4 Base Classifier Variants 138
4 Experimental Settings 138
5 Results and Discussion 141
6 Conclusion 141
References 142
Classification of Cancer Microscopic Images via Convolutional Neural Networks 144
1 Introduction 144
2 Dataset 145
3 Methods 145
3.1 Data Preprocessing 145
3.2 Data Augmentation 146
3.3 Training 146
4 Experimental Results 147
5 Discussion 148
6 Conclusion 149
References 149
| Erscheint lt. Verlag | 28.11.2019 |
|---|---|
| Reihe/Serie | Lecture Notes in Bioengineering |
| Zusatzinfo | X, 147 p. 64 illus., 61 illus. in color. |
| Sprache | englisch |
| Themenwelt | Informatik ► Theorie / Studium ► Künstliche Intelligenz / Robotik |
| Medizin / Pharmazie ► Medizinische Fachgebiete ► Onkologie | |
| Medizin / Pharmazie ► Physiotherapie / Ergotherapie ► Orthopädie | |
| Medizin / Pharmazie ► Studium | |
| Technik ► Elektrotechnik / Energietechnik | |
| Technik ► Medizintechnik | |
| Schlagworte | Biomedical Imaging • Cancer Imaging • computer assisted diagnosis • Deep learning • ISBI 2019 challenges • Signal Processing |
| ISBN-10 | 981-15-0798-8 / 9811507988 |
| ISBN-13 | 978-981-15-0798-4 / 9789811507984 |
| Haben Sie eine Frage zum Produkt? |
PDF (Wasserzeichen)Größe: 5,1 MB
DRM: Digitales Wasserzeichen
Dieses eBook enthält ein digitales Wasserzeichen und ist damit für Sie personalisiert. Bei einer missbräuchlichen Weitergabe des eBooks an Dritte ist eine Rückverfolgung an die Quelle möglich.
Dateiformat: PDF (Portable Document Format)
Mit einem festen Seitenlayout eignet sich die PDF besonders für Fachbücher mit Spalten, Tabellen und Abbildungen. Eine PDF kann auf fast allen Geräten angezeigt werden, ist aber für kleine Displays (Smartphone, eReader) nur eingeschränkt geeignet.
Systemvoraussetzungen:
PC/Mac: Mit einem PC oder Mac können Sie dieses eBook lesen. Sie benötigen dafür einen PDF-Viewer - z.B. den Adobe Reader oder Adobe Digital Editions.
eReader: Dieses eBook kann mit (fast) allen eBook-Readern gelesen werden. Mit dem amazon-Kindle ist es aber nicht kompatibel.
Smartphone/Tablet: Egal ob Apple oder Android, dieses eBook können Sie lesen. Sie benötigen dafür einen PDF-Viewer - z.B. die kostenlose Adobe Digital Editions-App.
Zusätzliches Feature: Online Lesen
Dieses eBook können Sie zusätzlich zum Download auch online im Webbrowser lesen.
Buying eBooks from abroad
For tax law reasons we can sell eBooks just within Germany and Switzerland. Regrettably we cannot fulfill eBook-orders from other countries.
aus dem Bereich
