Methodology in radiomics : step-by-step guide in radiomics pipeline / edited by Salvatore Claudio Fanni, Lorenzo Faggioni, Isabella Castiglioni, Emanuele Neri.

Format:

Book

Author/Creator:

Neri, E. (Emanuele), author.

Contributor:

Fanni, Salvatore Claudio, editor.

Faggioni, Lorenzo, editor.

Castiglioni, Isabella, editor.

Language:

English

Subjects (All):

Medical radiology--Data processing.

Medical radiology.

Medical radiology--Methodology.

Diagnostic imaging--Data processing.

Diagnostic imaging.

Diagnostic imaging--Methodology.

Diagnostic imaging--Technological innovations.

Radiomics.

Image Processing, Computer-Assisted.

Diagnostic Imaging--methods.

Technology, Radiologic.

Diagnostic Imaging.

Radiology.

Medical Subjects:

Radiomics.

Image Processing, Computer-Assisted.

Diagnostic Imaging--methods.

Technology, Radiologic.

Diagnostic Imaging.

Radiology.

Physical Description:

1 online resource (xiii, 234 pages) : color illustrations

Edition:

1st ed.

Other Title:

Step-by-step guide in radiomics pipeline

Place of Publication:

London : Academic Press, an imprint of Elsevier, 2025.

Summary:

Methodology in Radiomics: Step-by-step Guide in Radiomics Pipeline is an essential resource for anyone interested in the rapidly evolving field of radiomics.This comprehensive guide delves into the history and fundamental principles of radiomics, providing readers with a clear understanding of how to implement each step in the radiomic pipeline.

Contents:

1 Brief history and fundamentals of radiomics

Key terms and definitions

Key learning objectives

1.1 Radiomics: from definition to convolutional neural networks

1.2 Radiomics pipeline

1.2.1 Image acquisition

1.2.2 Region of interest individuation and segmentation task

1.2.3 Features extraction and selection

1.2.4 Radiomics feature classes

1.2.5 Model development

1.3 Conclusions

References

2 Sourcing good data: from image acquisition to segmentation

2.1 Introduction

2.2 Image acquisition: the key to powerful data

2.3 Image preprocessing: the gateway to high-quality analysis

2.4 Image segmentation: cutting through complexity

3 Radiomics features: making order out of chaos

3.1 Introduction

3.2 Features definition

3.2.1 First-order features

3.2.2 Shape-based features

3.2.3 Second-order features or textures

3.2.3.1 Gray level cooccurrence matrix features

3.2.3.2 Other second-order structures

3.2.3.2.1 Gray level run length matrix

3.2.3.2.2 Gray level size zone matrix

3.2.3.2.3 Gray level dependence matrix

3.2.4 Other complex features (higher-order features)

3.2.5 Package for features evaluation and standardization initiative

3.2.6 Final remarks

3.3 Radiomic workflow

3.3.1 Image acquisition and reconstruction

3.3.2 Image segmentation

3.3.3 Image preprocessing

3.3.3.1 Voxel size resampling

3.3.3.2 Re-segmentation

3.3.3.3 Intensity discretization

3.3.3.4 Image filtering

3.3.3.5 Feature calculation

3.4 Features selection

3.5 Conclusions: radiomics issues and perspectives

References.

4 Standardization and quality assessment

4.1 Introduction

4.2 The need for standardization in radiomics

4.2.1 Variability in image acquisition

4.2.2 Image preprocessing and segmentation

4.2.3 Feature extraction and calculation

4.3 Quality assessment in radiomics

4.4 Standardization initiatives and guidelines

4.4.1 Radiomics quality score

4.4.2 METhodological RadiomICs Score

4.4.3 The image biomarker standardization initiative

4.4.4 CheckList for Evaluation of Radiomics Research

4.4.5 Checklist for artificial intelligence in medical imaging

4.5 Conclusions-future directions and challenges

5 From statistics to machine learning: principles of data analysis

5.1 The role of data analysis in medicine

5.1.1 The origins

5.1.2 Toward precision medicine

5.2 Fundamental concepts in statistics

5.2.1 Descriptive statistics

5.2.1.1 Central values and measures of variability

5.2.2 Inferential statistics

5.2.2.1 Confidence intervals

5.2.2.2 Hypothesis testing and P value

5.2.2.2.1 Testing between-group differences

5.2.2.2.2 Testing correlations between variables

5.2.2.2.3 Corrections for multiple comparisons

5.3 From traditional statistics to data science

5.3.1 Exploratory data analysis

5.3.1.1 Visualizing data distributions

5.3.1.2 Identifying patterns, trends, and correlations

5.3.2 Empirical approach to inferential statistics

5.3.2.1 Estimating confidence intervals via bootstrapping

5.3.2.2 Estimating P values with permutation tests

5.3.2.2.1 Testing between-group differences

5.3.2.2.2 Testing correlations between variables

5.4 Introduction to machine learning

5.4.1 Supervised learning

5.4.1.1 Regression.

5.4.1.2 Classification

5.4.2 Unsupervised learning

5.4.2.1 Clustering

5.5 Model evaluation and validation

5.5.1 Performance metrics

5.5.1.1 Classification metrics

5.5.1.2 Regression metrics

5.5.2 Data partitioning and cross-validation techniques

5.5.3 Generalization capability of machine learning models

5.6 From traditional machine learning to deep learning

5.6.1 Origins and evolution of deep learning

5.6.2 Traditional machine learning versus deep learning

5.6.3 Bias-variance trade-off and model complexity

5.7 Major concerns in data analysis with machine learning

5.7.1 Bias and confounding variables

5.7.2 Importance of model interpretability

5.8 Conclusions

6 Model validation

6.1 Introduction

6.2 Data partitioning

6.3 Hyperparameter tuning

6.4 Bias-variance tradeoff

6.5 Model testing

6.6 Evaluation metrics

6.6.1 Segmentation

6.6.2 Detection

6.6.3 Discrimination

6.6.4 Regression

6.7 Robustness evaluation

6.8 Validation for medical devices

6.9 Challenges

6.10 Conclusions

7 Deriving knowledge from different sources: radiomics and clinical data

7.1 Beyond images: the importance of clinical data

7.2 The role of biobanks in radiomics research

7.2.1 Biobank definition

7.2.2 Data requirements

7.2.3 Governance of a biobank

7.3 Integrating radiomics and clinical data

7.3.1 Challenges in collecting clinical data

7.3.2 Privacy regulations: pseudonymization and anonymization

7.3.2.1 Pseudonymization or anonymization?

7.4 Implementation strategies: European imaging and clinical data biobanks

7.4.1 European scenario

7.4.2 EUCAIM: a pan-federated infrastructure

AI disclosure

8 Deep radiomics

8.1 Deep learning models in deep radiomics

8.1.1 Convolutional neural networks

8.1.2 Recurrent neural networks and long short-term memory networks

8.1.3 Transformers

8.1.4 Hybrid architectures

8.2 Clinical applications

8.3 Explainability

8.4 Conclusion: role of radiologists in the era of deep radiomics

9 The multiomics breakthrough in the new age of data

9.1 Introduction

9.2 Multiomics approaches in predictive modeling

9.3 Using other omics as ground truth

9.3.1 Radiomics and pathology

9.3.2 Radiomics and genetic information

9.3.3 Radiomics and proteomics

9.3.4 Radiomics and transcriptomics

9.4 Multimodal models

9.5 From research to clinical practice

9.6 Conclusion

10 The open chapter: applications in healthcare-oncologic applications

10.1 Introduction

10.2 Current applications of innovations in oncology

10.2.1 Artificial intelligence and machine learning in cancer diagnosis and prognosis

10.2.2 Personalized medicine: tailoring treatments to individual patients

10.2.3 Open data and collaborative research

10.3 Challenges in implementing these innovations

10.3.1 Data privacy and security

10.3.2 Bias and fairness in AI models

10.3.3 Integration with clinical workflows

10.3.4 Ethical considerations

10.4 Future potential and directions in oncologic applications

10.4.1 Advanced predictive analytics and real-time decision support

10.4.2 AI-driven drug discovery and precision medicine

10.4.3 Expansion of telemedicine and remote patient monitoring

10.4.4 Global collaboration and data-sharing networks

10.5 Conclusion

11 The open chapter: applications in healthcare-cardiac imaging applications

11.1 Introduction

11.2 Part I: radiomics in coronary imaging

11.2.1 Coronary artery disease

11.2.2 Translational applications

11.2.3 Limitations and possible solutions to promote clinical adaptation

11.3 Part II: radiomics in myocardial tissue characterization

11.3.1 Radiomics in nonischemic cardiomyopathies

11.3.1.1 Hypertrophic cardiomyopathy

11.3.1.2 Dilated cardiomyopathy

11.3.2 Radiomics in ischemic heart disease

11.3.3 Radiomics in myocarditis

11.3.4 The role of radiomic in multienergy and photon counting cardiac CT

11.4 Conclusions

12 The open chapter: applications in healthcare - neuroradiology applications

12.1 Introduction

12.2 Clinical applications of radiomics in neuroradiology

12.2.1 Neuro-oncology

12.2.1.1 Gliomas

12.2.1.1.1 Tumor grading

12.2.1.1.2 Molecular characterization

12.2.1.1.3 Treatment response and pseudoprogression

12.2.2 Brain metastases

12.2.2.1 Differentiation of brain metastases and primary tumors

12.2.3 Meningiomas

12.2.3.1 Meningioma risk stratification

12.2.4 Neurodegenerative diseases

12.2.4.1 Early diagnosis and differential diagnosis

12.2.4.2 Progression tracking and risk prediction

12.2.4.3 Differentiation and diagnosis

12.2.4.4 Disease severity and monitoring

12.2.5 Vascular disorders

12.2.5.1 Infarct core and penumbra assessment

12.2.5.2 Outcome prediction

12.3.6 Pediatric neuroradiology

12.3.6.1 Brain tumors in pediatrics

12.3.6.1.1 Tumor characterization and prognosis

12.3.7 Epilepsy

12.3.7.1 Detection of cortical dysplasia

12.3.8 Developmental and genetic disorders

12.3.8.1 Tracking brain maturation.

Notes:

Includes bibliographical references and index.

Description based on publisher supplied metadata and other sources.

ISBN:

0-443-29243-4

OCLC:

1543217269