Radiomics in Brain Tumor: Image Assessment, Quantitative Feature Descriptors, and Machine-Learning Approaches

Abstract

Radiomics describes a broad set of computational methods that extract quantitative features from radiographic images. The resulting features can be used to inform imaging diagnosis, prognosis, and therapy response in oncology. However, major challenges remain for methodological developments to optimize feature extraction and provide rapid information flow in clinical settings. Equally important, to be clinically useful, predictive radiomic properties must be clearly linked to meaningful biological characteristics as well as qualitative imaging properties familiar to radiologist. Here we use a cross-disciplinary approach to highlight studies in radiomics. We review brain tumor radiological studies (e.g., imaging interpretation) through computational models (e.g., computer vision and machine learning) that provide novel clinical insights. In particular, we outline current quantitative image feature extraction and prediction strategies with different level of available clinical classes for supporting clinical decision making. We further discuss machine learning challenges and data opportunities to advance radiomic studies.

A Triple‐Classification Radiomics Model for the Differentiation of Primary Chordoma, Giant Cell Tumor, and Metastatic Tumor of Sacrum Based on T2‐Weighted and Contrast‐Enhanced T1‐Weighted MRI

“…Samji et al found LAVA‐Flex may depict skeletal metastases with a high sensitivity and specificity, improve image quality and reduce artifact . In consideration of these benefits, CE T1w LAVA‐Flex may generate high‐quality imaging of pelvic tumors, such as showing enhancing regions within the tumor and distinguishing necrosis and solid tumor . Our result demonstrated that radiomics features extracted from joint T2w FS and CE T1w images outperformed those from either T2w FS or CE T1w images alone, which is consistent with the results of a previous study …”

“…MRI plays an important role in the diagnosis of sacral tumors, in which T2‐weighted fat saturation (T2w FS) fast recovery fast spin echo and contrast‐enhanced T1‐weighted (CE T1w) liver imaging with volume acceleration‐flexible (LAVA‐Flex) sequences are frequently used. T2w FS images are sensitive to water tissue content, and CE T1w can show enhancing regions within the tumor and distinguish necrosis and solid tumor . Compared with conventional fat suppression T1‐weighed images, LAVA‐Flex can offer superior image quality, less motion artifact, and more homogenous fat suppression, as has been shown in skeletal tumors …”

Brain tumor diagnosis based on artificial neural network and a chaos whale optimization algorithm

“…Images feature extraction is one of the important operations in image processing to the comparison of content‐based images. Feature extraction has been used extensively in medical imaging …”

“…Once feature selection is applied, some features are pruned and some others are kept for use in training for tumor classification. In this research, the optimal features selection is obtained by minimizing the Matthews correlation coefficient (MCC) function as the objective function and with the help of the CWOA which is explained before …”

Novel Quantitative Imaging for Predicting Response to Therapy: Techniques and Clinical Applications

“…Within the oncology space, radiomics has found applications as a means of diagnosis, a prognostic tool predicting response to therapy 35 across organ systems, including but not limited to brain, 36–39 head and neck, 40–42 breast, 43–50 lung, 42,51–53 prostate, 54–58 rectum, 59–64 and liver. 65–69 In the context of lung cancer, radiomics has successfully allowed detection of malignancies in screening CT scans, 53 provided a means to differentiate between benign and malignant lesions, 51 enabled the prediction of risk of recurrence post-therapy, 52 and provided a means to noninvasively assess response to therapy 70 as well as helped to identify patients who would most benefit from therapy.…”