医疗影像分析指南：方法与深度学习算法

下载需积分: 9 | PDF格式 | 17.76MB | 更新于2024-07-18 | 45 浏览量 | 举报

"《医学影像分析指南：方法与算法》第二版详细介绍了2010-2011年医学影像分析领域的核心方法，那时这些技术被认为是该行业的基石。然而，随着时间的推移，计算机科学领域特别是医学影像分析方面的发展日新月异，以下几点是本书所涵盖的关键更新： 1. 手术室图像引导：传统的手术室图像指导已不再仅仅局限于术前规划图像的展示，而是发展成为实时操作中的辅助工具。为了帮助医生在手术过程中灵活调整操作技术，需要可靠且直观的分析方法，这要求方法必须具有高效性和易用性。 2. 图像分割和标注：现在的处理策略主要基于优化问题，不论是离散域（第8章）还是连续域（第9章）。虽然启发式方法如第6章所述仍在非商业和商业软件产品中存在，但采用优化假设寻找最优解的方法更加可靠，能够提供更一致的结果。 3. 深度学习的应用：深度学习极大地推动了医学影像分析的革新，它结合了数据特征学习和多层感知器的抽象能力。这意味着在标注任务中，可以直接对像素进行学习，从而开发出无需针对特定问题设计的通用分析方法，只需要通过训练数据进行学习。本书不仅回顾了基础的计算机视觉和模式识别技术，还深入探讨了如何将这些技术与医学实践相结合，尤其是在医疗图像分析中，强调了从数据驱动的智能决策到实际临床应用的转化过程。作者Klaus D. Toennies，作为Otto-von-Guericke-Universität Magdeburg计算机科学系的教授，提供了全面而实用的指南，使读者能够跟上这一快速发展的领域。对于那些希望在医疗影像分析领域探索最新进展和技术的人来说，此书是一份不可或缺的参考资料。"

7 Segmentation in Feature Space ............................. 249

7.1 Segmentation by Classiﬁcation in Feature Space

........... 251

7.1.1 Computing the Likelihood Function .............. 252

7.1.2 Multi-dimensional Feature Vectors

............... 256

7.1.3 Computing a Priori Probability

.................. 258

7.1.4 Extension to More Than Two Classes

............. 260

7.2 Clustering in Feature Space

........................... 260

7.2.1 Partitional Clustering and K-means Clustering

....... 261

7.2.2 Mean Shift Clustering

......................... 264

7.2.3 Kohonen’s Self-organizing Maps

................. 267

7.3 Concluding Remarks

................................. 270

7.4 Exercises

.......................................... 271

References

............................................... 272

8 Segmentation as a Graph Problem

........................... 275

8.1 Graph Cuts

........................................ 277

8.1.1 Graph Cuts for Computing a Segmentation

......... 278

8.1.2 Graph Cuts to Approximate

a Bayesian Segmentation

....................... 284

8.1.3 Problem-Speciﬁc Adaptation of t-Link Weights

...... 289

8.1.4 Adding Generic Domain Knowledge

.............. 291

8.1.5 Normalized Graph Cuts ........................ 293

8.2 Segmentation as a Path Problem

........................ 296

8.2.1 Fuzzy Connectedness.......................... 296

8.2.2 The Image Foresting Transform

.................. 298

8.2.3 Random Walks

.............................. 300

8.2.4 Other Methods Based on Seed Placements

......... 304

8.3 Similarities and Integrations of Graph-Based Strategies

...... 305

8.4 Concluding Remarks

................................. 306

8.5 Exercises

.......................................... 306

References

............................................... 307

9 Active Contours and Active Surfaces

......................... 311

9.1 Explicit Active Contours and Surfaces

................... 313

9.1.1 Deriving the Model

........................... 313

9.1.2 The Use of Additional Constraints

................ 316

9.1.3 T-Snakes and T-Surfaces

....................... 319

9.2 The Level Set Model

................................ 321

9.2.1 Level Sets

.................................. 322

9.2.2 Level Sets and Wave Propagation ................ 324

9.2.3 Schemes for Computing Level Set Evolution

....... 329

9.2.4 Computing Stationary Level Set Evolution

......... 333

9.2.5 Computing Dynamic Level Set Evolution

.......... 336

9.2.6 Segmentation and Speed Functions

............... 338

9.2.7 Geodesic Active Contours

...................... 341

xvi Contents

9.2.8 Level Sets and the Mumford–Shah Functional ...... 343

9.2.9 Avoiding Reinitialization

....................... 351

9.2.10 Noise Reduction and Boundary Enhancement ....... 352

9.2.11 Initial Level Sets and Initial Regions of Interest

..... 353

9.2.12 Topologically Constrained Level Sets

............. 354

9.3 Concluding Remarks

................................. 355

9.4 Exercises

.......................................... 356

References

............................................... 357

10 Registration and Normalization

............................. 361

10.1 Feature Space and Correspondence Criterion

.............. 363

10.1.1 Correspondence Criteria for Extrinsic

and Intrinsic Markers

.......................... 364

10.1.2 Requirements for Extrinsic and Intrinsic Markers

.... 367

10.1.3 Similarity for Image-Based Features

.............. 370

10.2 Rigid Registration

................................... 375

10.2.1 Registration Using Paired Landmarks

............. 375

10.2.2 Registration with Unpaired Landmarks

............ 378

10.2.3 Outlier Detection

............................. 379

10.2.4 Registration in the Frequency Domain

............. 380

10.2.5 The Head-Hat Registration

..................... 381

10.2.6 The ICP Algorithm ........................... 382

10.3 Registration of Projection Images to 3d Data

.............. 383

10.3.1 Registration Using Digitally

Reconstructed Radiographs

..................... 384

10.3.2 Optimization of the Similarity Criterion

............ 386

10.4 Search Space and Optimization in Non-rigid Registration

..... 387

10.4.1 Elastic Registration

........................... 388

10.4.2 Using a Viscous Fluid Model

................... 388

10.4.3 Free-Form Deformations

....................... 390

10.4.4 Neighborhood Similarity

....................... 390

10.4.5 Multi-scale Registration

........................ 391

10.5 Normalization

...................................... 392

10.6 Joint Segmentation and Registration

..................... 395

10.7 Non-rigid Discontinuous Registration

.................... 397

10.8 Concluding Remarks

................................. 400

10.9 Exercises

.......................................... 400

References

............................................... 401

11 Shape, Appearance and Spatial Relationships

.................. 405

11.1 Shape Model s

...................................... 407

11.2 Simple Models

..................................... 411

11.2.1 Template Matching

........................... 411

11.2.2 Hough Transform

............................ 413

11.3 Implicit Models

..................................... 415

Contents xvii

11.4 The Medial Axis Representation ........................ 417

11.4.1 Computation of the Medial Axis Transform

........ 417

11.4.2 Shape Representation by Medial Axes............. 419

11.5 Active Shape and Active Appearance Models

............. 420

11.5.1 Creating an ASM

............................. 421

11.5.2 Using ASMs for Segmentation

.................. 424

11.5.3 The Active Appearance Model

.................. 426

11.5.4 Deﬁciencies of Statistical Shape Model s

........... 427

11.5.5 Changing the Param etric Model

.................. 429

11.5.6 Using a Part-Based Model

...................... 429

11.5.7 Using a Non-parametric Model

.................. 430

11.6 Physically-Based Shape Models

........................ 432

11.6.1 Mass-Spring Models

.......................... 434

11.6.2 Finite Element Models

......................... 439

11.7 Shape, Appearance and Pose Priors for Segmentation

....... 453

11.7.1 Atlas-Based Segmentation

...................... 453

11.7.2 Combining Shape Information with Leve l Set

Segmentation

................................ 455

11.7.3 Solutions Based on Graph Cuts

.................. 460

11.7.4 Graph Cuts on Pre-segmented Images

............. 461

11.8 Using a Graphical Model for Detecting Object Parts

........ 463

11.9 Concluding Remarks

................................. 465

11.10 Exercises

.......................................... 466

References

............................................... 468

12 Classiﬁcation and Clustering

............................... 473

12.1 Features and Feature Space

............................ 474

12.1.1 Linear Decorrelation of Features

................. 475

12.1.2 Linear Discriminant Analysis

.................... 477

12.1.3 Independent Component Analysis

................ 478

12.2 Bayesian Classiﬁer

.................................. 480

12.3 Classiﬁcation Based on Distance to Training Samples

....... 482

12.4 Decision Boundaries

................................. 485

12.4.1 Adaptive Decision Boundaries

................... 486

12.4.2 The Multilayer Perceptron

...................... 488

12.4.3 Support Vector Machines

...................... 495

12.5 Convolutional Neural Networks and Deep Learning

......... 498

12.5.1 Structure of a CNN

........................... 499

12.5.2 What Does a CNN Learn

...................... 502

12.5.3 Applications in Medical Imag e Analysis

........... 504

12.6 Classiﬁcation by Association

.......................... 506

12.7 Clustering Techniques

................................ 508

12.7.1 Agglomerative Clustering

...................... 508

12.7.2 Fuzzy c-Means Clustering

...................... 510

xviii Contents

12.8 Bagging and Boosting................................ 512

12.9 Decision Trees

..................................... 514

12.9.1 Creating a Decision Tree ....................... 514

12.9.2 Random Forest

.............................. 516

12.10 Multiple Instance Learning

............................ 516

12.11 Bag of Visual Words

................................ 518

12.12 Dealing with Imbalanced Data

......................... 521

12.13 Concluding Remarks

................................. 524

12.14 Exercises

.......................................... 524

References

............................................... 526

13 Validation

.............................................. 529

13.1 Measures of Quality

................................. 531

13.1.1 Quality for a Delineation Task

.................. 532

13.1.2 Considering Interaction in a Segmentation Task

..... 536

13.1.3 Quality for a Detection Task

.................... 537

13.1.4 Quality for a Registration Task

.................. 541

13.2 The Ground Truth

................................... 542

13.2.1 Ground Truth from Real Data

................... 543

13.2.2 Ground Truth from Phantoms

................... 546

13.3 Representativeness of Data

............................ 552

13.3.1 Separation Between Training and Test Data ........ 552

13.3.2 Identiﬁcation of Sources of Variation and Outlier

Detection

................................... 553

13.3.3 Robustness with Respect to Parameter Variation

..... 555

13.4 Signiﬁcance of Results

............................... 555

13.5 Concluding Remarks

................................. 559

13.6 Exercises

.......................................... 559

References

............................................... 560

14 Appendix

............................................... 563

14.1 Optimization of Markov Random Fields

.................. 563

14.1.1 Markov Random Fields

........................ 563

14.1.2 Simulated Annealing

.......................... 566

14.1.3 Mean Field Annealing

......................... 568

14.1.4 Iterative Conditional Modes

..................... 569

14.2 Variational Calculus

................................. 569

14.3 Principal Component Analysis

......................... 572

14.3.1 Computing the PCA

.......................... 572

14.3.2 Robust PCA ................................ 575

References

............................................... 577

Erratum to: Guide to Medical Image Analysis

..................... E1

Index

...................................................... 579

Contents xix

剩余608页未读，继续阅读

clarkmant

粉丝: 1

医疗影像分析指南：方法与深度学习算法

探索 dc_qiskit_algorithms Python 库及其功能

解压缩便用Python库：dc_qiskit_algorithms介绍

C语言恐怖作品集：my_own_algorithms

Guide to Medical Image Analysis Methods and Algorithms

Guide.to.Medical.Image.Analysis.Methods.and.Algorithms

Visual Media Processing Using MATLAB Beginner's Guide

Enhancing Image Quality: A Complete Guide to OpenCV Image Enhancement Techniques, from Grayscale ...

Optimization of Medical Data Analysis with MATLAB: Practical Applications of Optimization Algorithms

The Ultimate Guide to Machine Learning Model Selection: 20 Secrets and Tips from Novice to Expert

【Project Practicality】: New Horizons in Image Transformation: A Practical Guide to the Application...

最新资源