Radiomics in radiation oncology—basics, methods, and limitations

Lohmann, Philipp; Bousabarah, Khaled; Hoevels, Mauritius; Treuer, Harald

doi:10.1007/s00066-020-01663-3

Cited by 73 publications

(57 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For decades, in medical oncology, patients suffering from cancer underwent diagnosis imaging acquisitions including PET/CT/MRI, where anatomical and functional information were combined to provide prognosis of the disease and an effective treatment plan. The extensive use of advanced hybrid imaging scanners increased the amount of diagnostic data in daily routine, enhancing the need of computational support for fast and accurate diagnosis [2]. Daily clinical applications seem to take more and more advantage of the rapid developments of AI alongside the evolution of computer science.…”

Section: Ai In Oncologymentioning

confidence: 99%

“…Therefore, in contrast to feature-based radiomics, large datasets are necessary to identify a relevant and robust feature subset. One other limitation of deep learning-based radiomics is the high correlation between the features and the input data, as the DLR features are generated from that very data without the application of prior knowledge [2].…”

Section: Deep Learning Radiomics (Dlr) Featuresmentioning

confidence: 99%

See 1 more Smart Citation

Artificial intelligence: Deep learning in oncological radiomics and challenges of interpretability and data harmonization

et al. 2021

View full text Add to dashboard Cite

Over the last decade there has been an extensive evolution in the Artificial Intelligence (AI) field. Modern radiation oncology is based on the exploitation of advanced computational methods aiming to personalization and high diagnostic and therapeutic precision. The quantity of the available imaging data and the increased developments of Machine Learning (ML), particularly Deep Learning (DL), triggered the research on uncovering "hidden" biomarkers and quantitative features from anatomical and functional medical images. Deep Neural Networks (DNN) have achieved outstanding performance and broad implementation in image processing tasks. Lately, DNNs have been considered for radiomics and their potentials for explainable AI (XAI) may help classification and prediction in clinical practice. However, most of them are using limited datasets and lack generalized applicability. In this study we review the basics of radiomics feature extraction, DNNs in image analysis, and major interpretability methods that help enable explainable AI. Furthermore, we discuss the crucial requirement of multicenter recruitment of large datasets, increasing the biomarkers variability, so as to establish the potential clinical value of radiomics and the development of robust explainable AI models.

show abstract

Section: Ai In Oncologymentioning

confidence: 99%

Section: Deep Learning Radiomics (Dlr) Featuresmentioning

confidence: 99%

Artificial intelligence: Deep learning in oncological radiomics and challenges of interpretability and data harmonization

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Images were resampled to a 1 × 1 × 5 mm 3 voxel size using the AFNI package (https://afni.nimh.nih.gov/ (accessed on: 5 May 2021)) [14]. Due to large difference between slice thickness (5 mm) and in-plane spacing (0.5-0.75 mm) in our subjects, there was a risk of introducing artificial information and bias with upsampling and information loss with downsampling [15][16][17]. "As per image biomarker standardization initiative (IBSI) guidelines, in patients with large slice thickness compared to in plane voxel size dimensions, it may be beneficial to perform 2D interpolation.…”

Section: Image Pre-processingmentioning

confidence: 99%

Radiomic Based Machine Learning Performance for a Three Class Problem in Neuro-Oncology: Time to Test the Waters?

Priya

Liu

Ward

et al. 2021

Cancers

View full text Add to dashboard Cite

Prior radiomics studies have focused on two-class brain tumor classification, which limits generalizability. The performance of radiomics in differentiating the three most common malignant brain tumors (glioblastoma (GBM), primary central nervous system lymphoma (PCNSL), and metastatic disease) is assessed; factors affecting the model performance and usefulness of a single sequence versus multiparametric MRI (MP-MRI) remain largely unaddressed. This retrospective study included 253 patients (120 metastatic (lung and brain), 40 PCNSL, and 93 GBM). Radiomic features were extracted for whole a tumor mask (enhancing plus necrotic) and an edema mask (first pipeline), as well as for separate enhancing and necrotic and edema masks (second pipeline). Model performance was evaluated using MP-MRI, individual sequences, and the T1 contrast enhanced (T1-CE) sequence without the edema mask across 45 model/feature selection combinations. The second pipeline showed significantly high performance across all combinations (Brier score: 0.311–0.325). GBRM fit using the full feature set from the T1-CE sequence was the best model. The majority of the top models were built using a full feature set and inbuilt feature selection. No significant difference was seen between the top-performing models for MP-MRI (AUC 0.910) and T1-CE sequence with (AUC 0.908) and without edema masks (AUC 0.894). T1-CE is the single best sequence with comparable performance to that of multiparametric MRI (MP-MRI). Model performance varies based on tumor subregion and the combination of model/feature selection methods.

show abstract

“…Semantic features are used to describe morphologic characteristics of lesions such as shape, size, location, etc., while agnostic features (e.g. textural features) use innovative mathematical procedures in a high-throughput way that may fail to be perceived by the naked eye [16][17][18][19].…”

Section: Introductionmentioning

confidence: 99%

Evaluation of textural-based radiomics features for differentiation of COVID-19 pneumonia from non-COVID pneumonia

Soleymani

Jahanshahi

Hefzi

et al. 2021

Egypt J Radiol Nucl Med

View full text Add to dashboard Cite

Background The false-positive rate of computed tomography (CT) images in the diagnosis of coronavirus disease 2019 (COVID-19) is a challenge for the management in the pandemic. The main purpose of this study is to investigate the textural radiomics features on chest CT images of COVID-19 pneumonia patients and compare them with those of non-COVID pneumonia. This is a retrospective study. Some textural radiomics features were extracted from the CT images of 66 patients with COVID-19 pneumonia and 40 with non-COVID pneumonia. For radiomics analysis, the regions of interest (ROIs) were manually identified inside the pulmonary ground-glass opacities. For each ROI, 12 textural features were obtained and, then, statistical analysis was performed to assess the differences in these features between the two study groups. Results 8 of the 12 texture features demonstrated a significant difference (P < 0.05) in two groups, with COVID-19 pneumonia lesions tending to be more heterogeneous in comparison with the non-COVID cases. Among the 8 significant features, only two (homogeneity and energy) were found to be higher in non-COVID cases. Conclusions Textural radiomics features can be used for differentiating COVID-19 pneumonia from non-COVID pneumonia, as a non-invasive method, and help with better prognosis and diagnosis of COVID-19 patients.

show abstract

Radiomics in radiation oncology—basics, methods, and limitations

Cited by 73 publications

References 43 publications

Artificial intelligence: Deep learning in oncological radiomics and challenges of interpretability and data harmonization

Artificial intelligence: Deep learning in oncological radiomics and challenges of interpretability and data harmonization

Radiomic Based Machine Learning Performance for a Three Class Problem in Neuro-Oncology: Time to Test the Waters?

Evaluation of textural-based radiomics features for differentiation of COVID-19 pneumonia from non-COVID pneumonia

Contact Info

Product

Resources

About