gms | German Medical Science

Background: In recent years, the field of medical image analysis has grown rapidly. The number of pattern recognition tools and the data set size to be processed has increased enormously. These advances facilitate high-throughput extraction of quantitative features. These features enable the conversion of images into mineable data with subsequent analysis of this data for decision support. This process of extracting a large amount of quantitative information from medical imaging studies to characterize certain aspects of patient health is called “radiomics”. This is in contrast to the common practice of using medical images as images for visual interpretation only.

Methods: Radiomic data include first-, second- and higher-order statistics. These data are combined with other patient data and are mined using sophisticated bioinformatics tools. For example, radiomics models can be created to predict the effect of radiotherapy, the response to treatment, the prognosis of an imaged cancer, the genetic characteristics of cancer (radiogenomics), etc. Technically, Radiomics combines image processing, computer vision, quantitative imaging and machine learning.

Results: In this presentation, different approaches to radiomic investigations will be discussed, including: Radiomics for different imaging modalities (CT, MRI and PET) and the use of registered multimodal imaging data sets as a basis for radiomics as well as longitudinal radiomics and radiomics in combination with biomarkers (“Pan-Omics”). Since radiomics analyses can be performed with standard care images, among other things, it is conceivable that the conversion of digital images into mineable data will eventually become routine. However, there are many challenges for the potential use of radiomics-derived methods in clinical practice, including: standardization and robustness of selected metrics, collection of the required data, creation and validation of the resulting models, registration of image data, reliable segmentation tools, etc.

Conclusion: This talk highlights the process of radiomics, its challenges and its potential power to enable better clinical decision making. The results achieved so far show the enormous potential of this general approach to quantify and use data from medical images.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.