Search

Background: Treatment plans for squamous cell carcinoma of the head and neck (SCCHN) are individually decided in tumor board meetings but some treatment decision-steps lack objective prognostic estimates. Our purpose was to explore the potential of radiomics for SCCHN therapy-specific survival prognostication and to increase the models' interpretability by ranking the features based on their predictive importance. Methods: We included 157 SCCHN patients (male, 119; female, 38; mean age, 64.39 ± 10.71 years) with baseline head and neck CT between 09/2014 and 08/2020 in this retrospective study. Patients were stratified according to their treatment. Using independent training and test datasets with cross-validation and 100 iterations, we identified, ranked and inter-correlated prognostic signatures using elastic net (EN) and random survival forest (RSF). We benchmarked the models against clinical parameters. Inter-reader variation was analyzed using intraclass-correlation coefficients (ICC). Results: EN and RSF achieved top prognostication performances of AUC = 0.795 (95% CI 0.767–0.822) and AUC = 0.811 (95% CI 0.782–0.839). RSF prognostication slightly outperformed the EN for the complete (ΔAUC 0.035, p = 0.002) and radiochemotherapy (ΔAUC 0.092, p < 0.001) cohort. RSF was superior to most clinical benchmarking (p ≤ 0.006). The inter-reader correlation was moderate or high for all features classes (ICC ≥ 0.77 (± 0.19)). Shape features had the highest prognostic importance, followed by texture features. Conclusions: EN and RSF built on radiomics features may be used for survival prognostication. The prognostically leading features may vary between treatment subgroups. This warrants further validation to potentially aid clinical treatment decision making in the future. [ABSTRACT FROM AUTHOR]

Background: Arterial hyperenhancement and washout on computed tomography and magnetic resonance imaging (MRI) are described by all major guidelines as specific criteria for non-invasive diagnosis of hepatocellular carcinoma (HCC). However, publications on the quantitative assessment of washout in MRI are lacking. Therefore, we evaluated a method for quantitatively measuring and defining washout in MRI in order to determine a cutoff value that allows objective HCC diagnosis., Methods: We analyzed all patients who underwent liver transplantation for cirrhosis or liver resection for HCC at our institution between 2003 and 2014. Washout was quantitatively investigated by placing a 25-mm 2 region of interest (ROI) over each nodule and two 25-mm 2 ROIs over adjacent liver parenchyma. The percentage signal ratio (PSR = 100 × ratio of signal intensity of adjacent liver to that of the lesion) was calculated for each series in both groups. Accordingly, this quantitative measurement was compared to a qualitative approach., Results: A total of 16 hypervascularized non-HCC nodules and 69 HCC nodules were identified. Interobserver reliability was reasonably good for the measurement of PSRs and readers showed a substantial agreement for the qualitative assessment. In the HCC group, the median PSR was 116.2 at equilibrium and 112.9 in the delayed phase. In the non-HCC group, the median PSR was 93.8 at equilibrium and 96.0 in the delayed phase. Receiver operating characteristic analysis indicated areas under the curve of 0.902 (p < 0.001) and 0.873 (p < 0.001) at equilibrium and in the delayed phase. PSR values of 102 at equilibrium and 101.5 in the delayed phase led to the highest Youden's index of 0.82 and 0.77, respectively. These PSR cutoffs yielded sensitivities of 82 and 77 %, respectively, with specificities of 100 %. The sensitivity for the qualitative assessment of washout was 88 and 93 % and the specificity was 48 and 56 %. For the classification of HCC, sensitivity yielded 95 and 97 % and specificity was 68 and 56 %, respectively., Conclusion: Quantitatively measuring HCC washout in MRI is easy and reproducible. It can objectify and support diagnosis of HCC. However, the quantitative measurement of washout can only serve as one of several components of HCC assessment.