lung nodule classification using deep features in ct
play

Lung Nodule Classification Using Deep Features in CT Images - PowerPoint PPT Presentation

Sep 23, 2023 •271 likes •463 views

Lung Nodule Classification Using Deep Features in CT Images Devinder Kumar, Alexander Wong, and David A. Clausi June 5 th , 2015 Vision and Image Processing Research Group, UWaterloo CRV conference, 2015 Outline Why? Motivation What?

  1. Lung Nodule Classification Using Deep Features in CT Images Devinder Kumar, Alexander Wong, and David A. Clausi June 5 th , 2015 Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  2. Outline Why? Motivation What? Proposed Approach How? Exp. Setup So, What? Future Work Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  3. Challenges and Motivation Why? Lung cancer results in 17% of total cancer related deaths. Early diagnosis required as it is harder to contain in later stages. Burden on doctors for early diagnosis. Untapped data is now available to build effective computer aided diagnosis (CAD) systems. Goal: second opinion ! Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  4. Proposed Approach Build an effective CAD system to classify annotated nodules as malignant or benign using deep features extracted from autoencoder and binary decision tree as classifier. Figure : Proposed system flow diagram Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  5. CAD system Design : Dataset LIDC-IDRI dataset Thoracic CT images of 1010 patients Diagnostic data for 157 patients avialable (ground truth) Ratings: 0-Unknown, 1-benign, 2-Primary malignant, 3-metastatic Annotations provided! Nodule size: 3 mm to 30 mm (A) (C) (B) (D) Figure : Annotations provided by four different radiologists Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  6. CAD system Design : Autoencoder Design: Encoder Decoder Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  7. Autoencoder Let input be f ( x i ) ǫ [0 , 1] d latent space y ǫ [0 , 1] d φ be non linear function y = φ ( Wf ( x i ) + b ) (1) Reconstruction: f ( x i ) ′ = φ ( W ′ y + b ′ ) (2) Error minimization: n � f ( x i ) ′ − f ( x i ) � 2 � min (3) W , b i =1 Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  8. Stacked Autoencoder Figure : Stacked autoencoder formation Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  9. Stacked Autoencoder Figure : Stacked autoencoder formation Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  10. Stacked Autoencoder Figure : Stacked autoencoder formation Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  11. Stacked Autoencoder Figure : Stacked autoencoder formation Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  12. Stacked Autoencoder Our Design 3 Hidden layers layer size 200,100,200 Iteration set: 30 Batch size: 400 Feature extraction at 3rd hidden layer Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  13. Experimental Setup Data: 4303 Instances (4323 nodules) Obtained from diagnostic data all provided annotation considered Rating: 1: benign & 0,2,3: malignant Feature extraction: features are extracted from 4th layer (3rd hidden layer) 200 dim. vector Training: 90% of 4303 Instances 10-fold cross validation Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  14. Results 10 fold cross validation avg. : Accuracy: 75.01 Sensitivity: 83.35 FP/patient: 0.39 Belief Decision Trees 1 Deep Features Accuracy 75.01% 68.66% 1 D. Zinovev et al., Probabilistic lung nodule classification with belief decision trees in EMBC, 2011 Annual International Conference of the IEEE. IEEE, 2011, pp. 44934498 Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  15. Results:Dificult cases Benign (a) (b) (c) malignant (d) (e) (f) Figure : significant visual similarities between the annotated nodules in (a,d), (b,e) and (c,f), making it very difficult to differentiate between such nodules during the classification process. Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  16. So,What? Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  17. Future Work different deep architectures (e.g. CNN) & more hidden layers i.e. very deep networks (16-19 layers) combination of features STAPLE SPIE lung nodule classification challenge Automatic nodule detection Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

  18. Thank you for listening! Contact: Devinder Kumar Email: devinder.kumar@uwaterloo.ca Vision and Image Processing Research Group, UWaterloo CRV conference, 2015

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Lung Nodule Detection based on Convolutional Neural Networks Julio Mendoza Bobadilla Advisor:

Lung Nodule Detection based on Convolutional Neural Networks Julio Mendoza Bobadilla Advisor:

Lung Nodule Detection based on Convolutional Neural Networks Julio Mendoza Bobadilla Advisor: Prof. Helio Pedrini Masters Defense Institute of Computing University of Campinas December 18, 2017 Julio Mendoza Bobadilla IC/UNICAMP Lung

456 views • 43 slides
Pulmonary Nodule Malignancy Classification Using Temporal Evolution with Two-Stream 3D

Pulmonary Nodule Malignancy Classification Using Temporal Evolution with Two-Stream 3D

Pulmonary Nodule Malignancy Classification Using Temporal Evolution with Two-Stream 3D Convolutional Neural Networks X. Rafael-Palou 1,2 , A. Aubanell 3 , I. Bonavita 2 , M. Ceresa 1 , G. Piella 1 , V. Ribas 2 , M. Gonzlez Ballester 2,4 1 BCN

262 views • 7 slides
Update on Lung Cancer Classification International Multidisciplinary Classification of

Update on Lung Cancer Classification International Multidisciplinary Classification of

5/25/2013 Overview Update on Lung Cancer Classification International Multidisciplinary Classification of Adenocarcinoma Categorization Grading Invasion Molecular Testing Guidelines CAP/IASLC/AMP Kirk D. Jones, MD

363 views • 21 slides
Challenges in the Diagnosis of Interstitial Lung Disease New Classification of IIP Prior

Challenges in the Diagnosis of Interstitial Lung Disease New Classification of IIP Prior

5/24/2014 Overview Challenges in the Diagnosis of Interstitial Lung Disease New Classification of IIP Prior classification Modifications for new classification Diagnosis of UIP/NSIP Clinical, radiologic, pathologic findings

166 views • 16 slides
Challenges in the Diagnosis of Interstitial Lung Disease New Classification of IIP Prior

Challenges in the Diagnosis of Interstitial Lung Disease New Classification of IIP Prior

5/12/2014 Overview Challenges in the Diagnosis of Interstitial Lung Disease New Classification of IIP Prior classification Modifications for new classification Diagnosis of UIP/NSIP Clinical, radiologic, pathologic findings

190 views • 15 slides
Higher Dimensional Approach for Classification of Lung Cancer Microarray Data Nathan Palmer

Higher Dimensional Approach for Classification of Lung Cancer Microarray Data Nathan Palmer

Higher Dimensional Approach for Classification of Lung Cancer Microarray Data Nathan Palmer Tufts University / MIT (Joint work with Frederick Crimins, Robert Dimitri, Tsvika Klein and Lenore J. Cowen) Outline Classification of Tissue

600 views • 47 slides
Convolutional Neural Networks for Language CS 6956: Deep Learning for NLP Features from text

Convolutional Neural Networks for Language CS 6956: Deep Learning for NLP Features from text

Convolutional Neural Networks for Language CS 6956: Deep Learning for NLP Features from text Example: Sentiment classification The goal: Is the sentiment of a sentence positive, negative or neutral? The film is fun and is host to some truly

1.37k views • 88 slides
Multiclass Classification CS 6956: Deep Learning for NLP 1 So far: Binary Classification We

Multiclass Classification CS 6956: Deep Learning for NLP 1 So far: Binary Classification We

Multiclass Classification CS 6956: Deep Learning for NLP 1 So far: Binary Classification We have seen linear models for binary classification We can write down a loss for binary classification Common losses: Hinge loss and log loss 2

1.25k views • 36 slides
Image Classification with DIGITS NVIDIA Deep Learning Institute 1 DEEP LEARNING INSTITUTE DLI

Image Classification with DIGITS NVIDIA Deep Learning Institute 1 DEEP LEARNING INSTITUTE DLI

Image Classification with DIGITS NVIDIA Deep Learning Institute 1 DEEP LEARNING INSTITUTE DLI Mission Helping people solve challenging problems using AI and deep learning. Developers, data scientists and engineers Self-driving cars,

1.44k views • 72 slides
Classification Based on Missing Features in Deep Convolutional Neural Networks Nemanja Milo

Classification Based on Missing Features in Deep Convolutional Neural Networks Nemanja Milo

Classification Based on Missing Features in Deep Convolutional Neural Networks Nemanja Milo sevi c UNSPMF EuroPython 2019 Nemanja Milo sevi c ( UNSPMF ) EuroPython 2019 1 / 19 Hello! Nemanja Milo sevi c PhD Student,

682 views • 19 slides
Image Classification with Deep Networks Ronan Collobert Facebook AI Research Feb 11, 2015

Image Classification with Deep Networks Ronan Collobert Facebook AI Research Feb 11, 2015

Image Classification with Deep Networks Ronan Collobert Facebook AI Research Feb 11, 2015 Overview Origins of Deep Learning Shallow vs Deep Perceptron Multi Layer Perceptrons Going Deeper Why? Issues (and fix)?

916 views • 65 slides
Deep Learning for Classification CS293S, Yang, 2017 Computational graph for classification w 1 f

Deep Learning for Classification CS293S, Yang, 2017 Computational graph for classification w 1 f

Deep Learning for Classification CS293S, Yang, 2017 Computational graph for classification w 1 f 1 w 2 S f 2 >0? w 3 f 3 Objective: Classification Accuracy m l acc ( w ) = 1 sign( w > f ( x ( i ) )) == y ( i ) X m i =1

509 views • 49 slides
Concise Introduction to Deep Neural Networks Outline: Classification problems Motivating

Concise Introduction to Deep Neural Networks Outline: Classification problems Motivating

Concise Introduction to Deep Neural Networks Outline: Classification problems Motivating Deep (large) Neural Network (DNN) classifiers Neurons and DNN architectures Numerical training of DNNs (supervised deep learning) Spiking

630 views • 33 slides
Deep Neural Nets and Features Sung-Eui Yoon ( ) Course URL:

Deep Neural Nets and Features Sung-Eui Yoon ( ) Course URL:

CS688: Web-Scale Image Search Deep Neural Nets and Features Sung-Eui Yoon ( ) Course URL: http://sgvr.kaist.ac.kr/~sungeui/IR Class Objectives Browse main components of deep neural nets Does not aim for giving in-depth

609 views • 32 slides
Music Classification Overview and Audio Features Graduate School of Culture Technology, KAIST

Music Classification Overview and Audio Features Graduate School of Culture Technology, KAIST

GCT634: Musical Applications of Machine Learning Music Classification Overview and Audio Features Graduate School of Culture Technology, KAIST Juhan Nam Outlines Definition of Music Classification Tasks Overview of Music Classification

392 views • 37 slides
Pulmonary vasculopathies: 1.2.1 BMPR2 3.3 Other pulmonary diseases with mixed restrictive 1.2.2

Pulmonary vasculopathies: 1.2.1 BMPR2 3.3 Other pulmonary diseases with mixed restrictive 1.2.2

4/21/2018 5TH WORLD SYMPOSIUM CLASSIFICATION OF PH 1. PULMONARY ARTERIAL HYPERTENSION 3. PH DUE TO LUNG DISEASES AND/OR HYPOXIA 1.1 Idiopathic PAH 3.1 Chronic obstructive pulmonary disease 1.2 Heritable PAH 3.2 Interstitial lung disease

636 views • 10 slides
OncoCytes initial focus is on a confirmatory diagnostic solution Confirmatory 7 10M

OncoCytes initial focus is on a confirmatory diagnostic solution Confirmatory 7 10M

OncoCytes initial focus is on a confirmatory diagnostic solution Confirmatory 7 10M Americans 30 Pack-year history Malignant USPSTF Guidelines recommended screening covered by Medicare Biopsy Benign Nodules nodule High-risk

204 views • 5 slides
Thyroid Nodules UCSF Internal Medicine Updates May 22, 2017 Elizabeth Murphy, MD, DPhil No

Thyroid Nodules UCSF Internal Medicine Updates May 22, 2017 Elizabeth Murphy, MD, DPhil No

5/16/2017 Thyroid Nodules UCSF Internal Medicine Updates May 22, 2017 Elizabeth Murphy, MD, DPhil No conflicts Overview Thyroid nodule and cancer review Ultrasound FNA cytology Nodule follow up Putting it all together 1

329 views • 30 slides
Alternative methods for modeling of the cure rate in survival studies Vera Tomazella

Alternative methods for modeling of the cure rate in survival studies Vera Tomazella

Alternative methods for modeling of the cure rate in survival studies Vera Tomazella Uniiversidade Federal de s ao Carlos-S ao Carlos-SP veratomazella@gmail.com VIII Encontro dos Alunos de P os-gradua c ao em Estat stica e

541 views • 39 slides
Management Challenges in NTM Wael ElMaraachli University of California, San diego NTM

Management Challenges in NTM Wael ElMaraachli University of California, San diego NTM

Management Challenges in NTM Wael ElMaraachli University of California, San diego NTM conference; May, 17 th , 2018 Disclosures Paid by Insmed to moderate an Advisory board meeting re: Results of recent trial of ALIS (Amikacin liposome

549 views • 25 slides
Risks Associated with Lead-Free Electronics March 5, 2015 SERDP & ESTCP Webinar Series

Risks Associated with Lead-Free Electronics March 5, 2015 SERDP & ESTCP Webinar Series

SERDP & ESTCP Webinar Series Understanding and Mitigating the Risks Associated with Lead-Free Electronics March 5, 2015 SERDP & ESTCP Webinar Series Welcome and Introductions Rula Deeb, Ph.D. Webinar Coordinator Webinar Agenda

1.34k views • 69 slides
Associate Professor Faculty of Engineering Multimedia University Imaging Informatics

Associate Professor Faculty of Engineering Multimedia University Imaging Informatics

Mohammad Faizal Ahmad Fauzi, Ph.D. Associate Professor Faculty of Engineering Multimedia University Imaging Informatics Imaging basics Imaging modalities PACS and its core functions DICOM 20 mm Nodule 4 mm Nodule 5 mm

672 views • 64 slides
DSM financial cost and modelling Payment Regime Workshop London 1 st & 2 nd December 2016

DSM financial cost and modelling Payment Regime Workshop London 1 st & 2 nd December 2016

DSM financial cost and modelling Payment Regime Workshop London 1 st & 2 nd December 2016 Peter H. Jantzen Capital Power Management Ltd Why get involved in DSM Apart from geopolitical drivers, Deep Seabed Mining is driven by two aspects:

143 views • 11 slides
Learning Tumor Growth via Follow-Up Volume Prediction for Lung Nodules Yamin Li 1 , Jiancheng Yang

Learning Tumor Growth via Follow-Up Volume Prediction for Lung Nodules Yamin Li 1 , Jiancheng Yang

Learning Tumor Growth via Follow-Up Volume Prediction for Lung Nodules Yamin Li 1 , Jiancheng Yang 1 , Yi Xu 1 , Jingwei Xu 1 , Xiaodan Ye 2 , Guangyu Tao 2 , Xueqian Xie 3 , Guixue Liu 3 1 Shanghai Jiao Tong University, 2 Shanghai Chest Hospital,

109 views • 9 slides

More recommend