Image biomarker standardisation initiative (1612.07003v11)

Abstract: The image biomarker standardisation initiative (IBSI) is an independent international collaboration which works towards standardising the extraction of image biomarkers from acquired imaging for the purpose of high-throughput quantitative image analysis (radiomics). Lack of reproducibility and validation of high-throughput quantitative image analysis studies is considered to be a major challenge for the field. Part of this challenge lies in the scantiness of consensus-based guidelines and definitions for the process of translating acquired imaging into high-throughput image biomarkers. The IBSI therefore seeks to provide image biomarker nomenclature and definitions, benchmark data sets, and benchmark values to verify image processing and image biomarker calculations, as well as reporting guidelines, for high-throughput image analysis.

• The paper establishes standardized definitions and a robust workflow to ensure consistent extraction and reporting of radiomic features across studies.
• It proposes a comprehensive image processing protocol—including pre-processing, segmentation, and intensity discretisation—that enhances reproducibility in radiomic analysis.
• It provides benchmark digital phantoms and reference values, facilitating validation processes and accelerating clinical translation in oncology.

Overview of the Image Biomarker Standardisation Initiative

The paper "Image Biomarker Standardisation Initiative" forms the foundational charter for the standardisation of radiomic feature extraction. Radiomics refers to the high-throughput mining of imaging features, transforming medical images into quantitative data that can be analyzed for various medical applications, predominantly in oncology. However, the reproducibility of radiomic studies has been impeded by inconsistent methodologies and the absence of standardised guidelines. This document endeavors to address these issues by providing a comprehensive framework for the standardised extraction and reporting of image biomarkers.

Key Contributions

1. Standardisation Definitions and Nomenclature: The document offers precise definitions and nomenclature for image biomarkers, crucial to mitigating discrepancies arising from varied interpretations in different studies. This includes delineating both qualitative and quantitative biomarkers, with a particular focus on quantitative ones due to their potential for automation and high-throughput computation.
2. Image Processing Workflow: The paper introduces a generalised image processing workflow essential for the computation of radiomic features. This structured workflow encapsulates stages such as image pre-processing, segmentation, and intensity discretisation. By standardising these stages, the IBSI aims to ensure that radiomic features are computed consistently across different studies, facilitating reliable comparisons.
3. Digital Phantoms and Reference Values: The IBSI provides a set of benchmark datasets and digital phantoms, accompanied by detailed reference values for various radiomic features. These resources serve as benchmarks for validating the accuracy and repeatability of radiomic analysis workflows. The consensus on reference values has been meticulously achieved through extensive collaboration and rounds of verification among independent teams.
4. Guidelines and Reporting Tools: Complementing the technical standards, the IBSI also proposes reporting guidelines to enhance the transparency and reproducibility of radiomic studies. These guidelines stress the disclosure of all image processing parameters employed during analyses, enabling others to replicate and verify reported results accurately.

Implications and Future Directions

The standardisation efforts led by IBSI have significant implications for the fields of oncology and personalized medicine. By providing a unified framework, these guidelines can bolster the credibility and clinical utility of radiomic biomarkers, expediting their integration into routine clinical workflows. The standardisation is anticipated to enhance the comparability of studies, reduce data ambiguity, and ultimately improve the formulation of predictive and prognostic models in medical imaging.

Looking forward, the IBSI's methodologies might evolve to incorporate emerging imaging modalities, such as those stemming from multimodal imaging studies. As the integration of AI and machine learning in medical imaging accelerates, the IBSI framework could serve as a foundation for developing more sophisticated algorithms that could automatically handle vast amounts of imaging data with minimized human intervention.

In conclusion, the Image Biomarker Standardisation Initiative marks a pivotal step forward in harmonizing radiomics research by establishing robust standards and workflows. It not only addresses current challenges in reproducibility but also paves the way for more reliable and robust translation of radiomic features from bench to bedside. These initiatives significantly contribute to the potential application of radiomics in clinical decision-making processes and personalized treatment strategies.