ABSTRACT
Radiomics involves the extraction of information from medical images that are not visible to the human eye. There is evidence that these features can be used for treatment stratification and outcome prediction. However, there is much discussion about the reproducibility of results between different studies. This paper studies the reproducibility of CT texture features used in radiomics, comparing two feature extraction implementations, namely the MATLAB toolkit and Pyradiomics, when applied to independent datasets of CT scans of patients: (i) the open access RIDER dataset containing a set of repeat CT scans taken 15 min apart for 31 patients (RIDER Scan 1 and Scan 2, respectively) treated for lung cancer; and (ii) the open access HN1 dataset containing 137 patients treated for head and neck cancer. Gross tumor volume (GTV), manually outlined by an experienced observer available on both datasets, was used. The 43 common radiomics features available in MATLAB and Pyradiomics were calculated using two intensity-level quantization methods with and without an intensity threshold. Cases were ranked for each feature for all combinations of quantization parameters, and the Spearman's rank coefficient, rs, calculated. Reproducibility was defined when a highly correlated feature in the RIDER dataset also correlated highly in the HN1 dataset, and vice versa. A total of 29 out of the 43 reported stable features were found to be highly reproducible between MATLAB and Pyradiomics implementations, having a consistently high correlation in rank ordering for RIDER Scan 1 and RIDER Scan 2 (rs > 0.8). 18/43 reported features were common in the RIDER and HN1 datasets, suggesting they may be agnostic to disease site. Useful radiomics features should be selected based on reproducibility. This study identified a set of features that meet this requirement and validated the methodology for evaluating reproducibility between datasets.
ABSTRACT
OBJECTIVE(S): The treatment response to neoadjuvant chemoradiation (nCRT) differs largely in individuals treated for rectal cancer. In this study, we investigated the role of radiomics to predict the pathological response in locally advanced rectal cancers at different treatment time points: (1) before the start of any treatment using baseline T2-weighted MRI (T2W-MR) and (2) at the start of radiation treatment using planning CT. METHODS: Patients on nCRT followed by surgery between June 2017 to December 2019 were included in the study. Histopathological tumour response grading (TRG) was used for classification, and gross tumour volume was defined by the radiation oncologists. Following resampling, 100 and 103 pyradiomic features were extracted from T2W-MR and planning CT images, respectively. Synthetic minority oversampling technique (SMOTE) was used to address class imbalance. Four machine learning classifiers built clinical, radiomic, and merged models. Model performances were evaluated on a held-out test dataset following 3-fold cross-validation using area under the receiver operator characteristic curves (AUC) with bootstrap 95% confidence intervals. RESULTS: One hundred and fifty patients were included; 58/150 with TRG 1 were classified as complete responders, and rest were incomplete responders (IR). Clinical models performed better (AUC = 0.68) compared to radiomics models (AUC = 0.62). Overall, the clinical + T2W-MR model showed best performance (AUC = 0.72) in predicting the pathological response prior to therapy. Clinical + Planning CT-merged models could only achieve the highest AUC of 0.66. CONCLUSION: Merging clinical and baseline T2W-MR radiomics enhances predicting pathological response in rectal cancer. Validation in larger cohorts is warranted, especially for watch and wait strategies.
ABSTRACT
Background and purpose: Radiomics models trained with limited single institution data are often not reproducible and generalisable. We developed radiomics models that predict loco-regional recurrence within two years of radiotherapy with private and public datasets and their combinations, to simulate small and multi-institutional studies and study the responsiveness of the models to feature selection, machine learning algorithms, centre-effect harmonization and increased dataset sizes. Materials and methods: 562 patients histologically confirmed and treated for locally advanced head-and-neck cancer (LA-HNC) from two public and two private datasets; one private dataset exclusively reserved for validation. Clinical contours of primary tumours were not recontoured and were used for Pyradiomics based feature extraction. ComBat harmonization was applied, and LASSO-Logistic Regression (LR) and Support Vector Machine (SVM) models were built. 95% confidence interval (CI) of 1000 bootstrapped area-under-the-Receiver-operating-curves (AUC) provided predictive performance. Responsiveness of the models' performance to the choice of feature selection methods, ComBat harmonization, machine learning classifier, single and pooled data was evaluated. Results: LASSO and SelectKBest selected 14 and 16 features, respectively; three were overlapping. Without ComBat, the LR and SVM models for three institutional data showed AUCs (CI) of 0.513 (0.481-0.559) and 0.632 (0.586-0.665), respectively. Performances following ComBat revealed AUCs of 0.559 (0.536-0.590) and 0.662 (0.606-0.690), respectively. Compared to single cohort AUCs (0.562-0.629), SVM models from pooled data performed significantly better at AUC = 0.680. Conclusions: Multi-institutional retrospective data accentuates the existing variabilities that affect radiomics. Carefully designed prospective, multi-institutional studies and data sharing are necessary for clinically relevant head-and-neck cancer prognostication models.