Artificial intelligence-driven radiomics study in cancer: the role of feature engineering and modeling

Abstract

Modern medicine is reliant on various medical imaging technologies for non-invasively observing patients’ anatomy. However, the interpretation of medical images can be highly subjective and dependent on the expertise of clinicians. Moreover, some potentially useful quantitative information in medical images, especially that which is not visible to the naked eye, is often ignored during clinical practice. In contrast, radiomics performs high-throughput feature extraction from medical images, which enables quantitative analysis of medical images and prediction of various clinical endpoints. Studies have reported that radiomics exhibits promising performance in diagnosis and predicting treatment responses and prognosis, demonstrating its potential to be a non-invasive auxiliary tool for personalized medicine. However, radiomics remains in a developmental phase as numerous technical challenges have yet to be solved, especially in feature engineering and statistical modeling. In this review, we introduce the current utility of radiomics by summarizing research on its application in the diagnosis, prognosis, and prediction of treatment responses in patients with cancer. We focus on machine learning approaches, for feature extraction and selection during feature engineering and for imbalanced datasets and multi-modality fusion during statistical modeling. Furthermore, we introduce the stability, reproducibility, and interpretability of features, and the generalizability and interpretability of models. Finally, we offer possible solutions to current challenges in radiomics research.

CT whole lung radiomic nomogram: a potential biomarker for lung function evaluation and identification of COPD

“…Second, the maximal redundancy-minimal relevance algorithm was used to eliminate the redundant and irrelevant features. Minimal redundancy maximal relevance has been proven to be an effective and reliable feature selection method for radiomic, which can consider both the importance of features and the correlation between features to find the optimal feature subset [ 14 , 15 ]. Finally, the least absolute shrinkage and selection operator (LASSO) regression algorithm and penalty parameter adjustment were used for tenfold cross-validation.…”

Radiomics: The New Promise for Differentiating Progression, Recurrence, Pseudoprogression, and Radionecrosis in Glioma and Glioblastoma Multiforme

“…In some studies, radiologists manually contoured lesions [106]. This may lead to observer bias, high inter-reader variability, the derivation of unstable radiomics features, and increased variability in image acquisition and reconstruction, which can affect the reproducibility of radiomics features [93,118,119]. ITK-SNAP (http://www.itksnap.org/) was the main segmentation software used in studies, although other software was employed, too.…”

A non-negative spike-and-slab lasso generalized linear stacking prediction modeling method for high-dimensional omics data

“…An ideal model should own interpretability, such that the end-users can comprehend and utilize it effectively [ 5 ]. Sparse models such as the Lasso and Elastic net, are considered to be more interpretable since they emphasize the limited number of important features that contribute more to prediction [ 6 ].…”