21
Efficient Sparse Shape Composition with its Applications in Biomedical Image Analysis: An Overview Shaoting Zhang, Yiqiang Zhan, Yan Zhou, and Dimitris N. Metaxas

Shaoting Zhang, Yiqiang Zhan, Yan Zhou, and Dimitris N. Metaxas

Embed Size (px)

DESCRIPTION

Efficient Sparse Shape Composition with its Applications in Biomedical Image Analysis: An Overview. Shaoting Zhang, Yiqiang Zhan, Yan Zhou, and Dimitris N. Metaxas. Outline. Introduction Our methods Experiments Future work. Introduction Motivations. Chest X-ray. - PowerPoint PPT Presentation

Citation preview

Page 1: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Efficient Sparse Shape Composition with its Applications in Biomedical Image Analysis:

An Overview

Shaoting Zhang, Yiqiang Zhan, Yan Zhou, andDimitris N. Metaxas

Page 2: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Outline• Introduction• Our methods• Experiments• Future work

2

Page 3: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

IntroductionMotivations

• Segmentation (i.e., finding 2D/3D ROI) is a fundamental problem in medical image analysis.

• We use deformable models and shape prior. Shape representation is based on landmarks.

Chest X-ray

Lung CADLiver in low-dose whole-body CT

Rat brain structure in MR Microscopy

3

Page 4: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

IntroductionMotivations

• Shape prior is important:– Automatic, accuracy,

efficiency, robustness.– Handle weak or misleading

appearance cues from image information.

– Discover or preserve complex shape details.

Segmentation system

4

Page 5: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

IntroductionChallenges – shape prior modeling

• Good shape prior method needs to:– Handle gross errors of the input data.– Model complex shape variations.– Preserve local shape details.

5

Page 6: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

IntroductionRelevant work – shape prior methods

• Handling gross errors. – RANSAC + ASM [M. Rogers, ECCV’02]– Robust Point Matching [J. Nahed, MICCAI’06]

• Complex shape variation (multimodal distribution).– Mixture of Gaussians [T. F. Cootes, IVC’97]– Patient-specific shape [Y. Zhu, MICCAI’09]

• Recovering local details.– Sparse PCA [K. Sjostrand, TMI’07]– Hierarchical ASM [D. Shen, TMI’03]

6

Page 7: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Our ApproachShape prior using sparse shape composition

• Our method is based on two observations:– An input shape can be approximately represented by a

sparse linear combination of training shapes.– The given shape information may contain gross errors, but

such errors are often sparse....

7

Page 8: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Our ApproachShape prior using sparse shape composition

• Formulation:–

• Sparse linear combination:–

...≈

...≈

Dense xSparse xAligned data matrix D

Weight x

Input y

Global transformationparameter

Number of nonzero elements

8

Global transformationoperator

Page 9: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Our ApproachShape prior using sparse shape composition

• Non-Gaussian errors:–

9

Page 10: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Our ApproachShape prior using sparse shape composition

• Why it works?– Robust: Explicitly modeling “e” with L0 norm constraint.

Thus it can detect gross (sparse) errors.– General: No assumption of a parametric distribution

model (e.g., a unimodal distribution assumption in ASM). Thus it can model complex shape statistics.

– Lossless: It uses all training shapes. Thus it is able to recover detail information even if the detail is not statistically significant in training data.

Zhang, Metaxas, et.al.: MedIA’11

10

Page 11: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Experiments2D lung localization in X-ray

• Setting:– Manually select landmarks for

training purpose.– 200 training and 167 testing.– To locate the lung, detect

landmarks, then predict a shape to fit them.

– Sensitivity (P), Specificity (Q), Dice Similarity Coefficient.

11

Page 12: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Experiments 2D lung localization in X-ray

• Handling gross errors

Detection PA ASM RASM NN TPS Sparse1 Sparse2

Procrustes analysisActive Shape ModelRobust ASMNearest NeighborsThin-plate-splineWithout modeling “e”

12

Proposed method

Page 13: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Experiments2D lung localization in X-ray

• Multimodal distribution

Detection PA ASM/RASM NN TPS Sparse1 Sparse2

13

Page 14: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Experiments2D lung localization in X-ray

• Recover local detail information

Detection PA ASM/RASM NN TPS Sparse1 Sparse2

14

Page 15: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Experiments2D lung localization in X-ray

• Sparse shape components

• ASM modes:

15

0.5760

+ +

0.2156 0.0982

Page 16: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Experiments2D lung localization in X-ray

• Mean values (µ) and standard deviations (σ).

Left lung

Right lung

1)PA, 2)ASM, 3)RASM, 4)NN, 5)TPS, 6)Sparse1, 7)Sparse2

µσ

16

Page 17: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Future WorkDictionary Learning

17

... ≈... ...

Dictionary size = 128 #Coefficients = 800 #Input samples = 800

Page 18: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

18

DInitialize D

Sparse CodingUse MP or BP

Dictionary UpdateColumn-by-Column by SVD computation

Use K-SVD to learn a compact dictionary Aharon, Elad, & Bruckstein (‘04)

YT

* The slides are from http://www.cs.technion.ac.il/~elad/talks/

Future WorkDictionary Learning

Page 19: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Future WorkOnline Update Dictionary

19

...≈

...

Page 20: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Future WorkMesh Partitioning for Local Shape Priors

20

Segmentation system

Lung Heart

Rib

Colon

Abdomen

Page 21: Shaoting Zhang,  Yiqiang  Zhan, Yan Zhou, and Dimitris  N. Metaxas

Thanks!Questions and comments

21