Radiomics in oncology is an emerging translational field of research in biomedical image analysis. It combines extracted quantitative features from clinical standard-of-care imaging and based on such features developing subsequent predictive models directed towards personalized therapies. With the recent advancements in deep learning methods in medical imaging, radiomics has promising benefits for both feature extraction processes and predictive model development. Therefore, radiomics aim to redefine the way information technology intersects and interfaces with medicine, and it is expected to play a key role in non-invasive cancer characterization.
Cancer characterization is an area of scientific and technological development that will continue to have a profound impact on society in the foreseen future. Providing a timely diagnosis and personalized treatment can assist in increasing survival rates, improving the quality of life while reducing healthcare costs. Mechanisms occurring at genetic and tissue characteristics at the molecular level can be highlighted in macroscopic features of medical images. In a parametric image, each voxel can be associated with a value of a certain physiological change. The generated quantitative parametric map reflects reactions and activities deemed important for a better understanding of cancer complexity and response to treatment. Changes in the local physiology of cancer are usually defined by multiple complex mechanisms related to the blood supply, tissue hydrodynamics, and behavior of neighboring tissue regions. Quantification of the intrinsic variations of cancer, such as tumor spatiotemporal heterogeneity – alongside the challenge of big data analytics in healthcare and with the different treatment methods – poses a significant challenge to treatment decisions.
This special issue aims to showcase the latest methodological developments within the realm of computational parametric radiomics with a particular interest in cancer characterization.
Research papers with novel methodological developments that deal with the detection, localization, characterization, staging, biopsy guidance, and active surveillance of cancer are encouraged. Specifically, this includes topics such as, but not limited to:
- Image registration
- Image segmentation
- Image fusion and synthesis
- Image acquisition and reconstruction
- Novel deep learning methods for medical imaging
- Statistics for medical imaging
- Motion analysis
- Texture analysis
- Big data analytics
- Information Visualization
- Imaging and genomics
- Image-guided surgery
- Uncertainty estimation
- Multimodal image processing and analysis
- Functional and molecular imaging
- Interpretability and explainable algorithms
- Generative modeling
- Learning with noisy or limited data
- Statistical and mathematical models
- Shape modeling
- Surgical vision and augmented reality
- Computer-aided detection and diagnosis
- Computational anatomy and physiology
- Transfer learning, domain adaptation, data harmonization
- Geometric learning, geometric deep learning, geometric statistics
Please note: Manuscripts consisting solely of bioinformatics or computational analysis of public genomic or transcriptomic databases which are not accompanied by validation (independent cohort or biological validation in vitro or in vivo) are out of scope for Frontiers in Oncology.
Radiomics in oncology is an emerging translational field of research in biomedical image analysis. It combines extracted quantitative features from clinical standard-of-care imaging and based on such features developing subsequent predictive models directed towards personalized therapies. With the recent advancements in deep learning methods in medical imaging, radiomics has promising benefits for both feature extraction processes and predictive model development. Therefore, radiomics aim to redefine the way information technology intersects and interfaces with medicine, and it is expected to play a key role in non-invasive cancer characterization.
Cancer characterization is an area of scientific and technological development that will continue to have a profound impact on society in the foreseen future. Providing a timely diagnosis and personalized treatment can assist in increasing survival rates, improving the quality of life while reducing healthcare costs. Mechanisms occurring at genetic and tissue characteristics at the molecular level can be highlighted in macroscopic features of medical images. In a parametric image, each voxel can be associated with a value of a certain physiological change. The generated quantitative parametric map reflects reactions and activities deemed important for a better understanding of cancer complexity and response to treatment. Changes in the local physiology of cancer are usually defined by multiple complex mechanisms related to the blood supply, tissue hydrodynamics, and behavior of neighboring tissue regions. Quantification of the intrinsic variations of cancer, such as tumor spatiotemporal heterogeneity – alongside the challenge of big data analytics in healthcare and with the different treatment methods – poses a significant challenge to treatment decisions.
This special issue aims to showcase the latest methodological developments within the realm of computational parametric radiomics with a particular interest in cancer characterization.
Research papers with novel methodological developments that deal with the detection, localization, characterization, staging, biopsy guidance, and active surveillance of cancer are encouraged. Specifically, this includes topics such as, but not limited to:
- Image registration
- Image segmentation
- Image fusion and synthesis
- Image acquisition and reconstruction
- Novel deep learning methods for medical imaging
- Statistics for medical imaging
- Motion analysis
- Texture analysis
- Big data analytics
- Information Visualization
- Imaging and genomics
- Image-guided surgery
- Uncertainty estimation
- Multimodal image processing and analysis
- Functional and molecular imaging
- Interpretability and explainable algorithms
- Generative modeling
- Learning with noisy or limited data
- Statistical and mathematical models
- Shape modeling
- Surgical vision and augmented reality
- Computer-aided detection and diagnosis
- Computational anatomy and physiology
- Transfer learning, domain adaptation, data harmonization
- Geometric learning, geometric deep learning, geometric statistics
Please note: Manuscripts consisting solely of bioinformatics or computational analysis of public genomic or transcriptomic databases which are not accompanied by validation (independent cohort or biological validation in vitro or in vivo) are out of scope for Frontiers in Oncology.
