MR Intensity Normalization Methods Impact Sequence Specific Radiomics Prognostic Model Performance in Primary and Recurrent High-Grade Glioma.

PURPOSE: This study investigates the impact of different intensity normalization (IN) methods on the overall survival (OS) radiomics models' performance of MR sequences in primary (pHGG) and recurrent high-grade glioma (rHGG). METHODS: MR scans acquired before radiotherapy were retrieved from two independent cohorts (rHGG C1: 197, pHGG C2: 141) from multiple scanners (15, 14). The sequences are T1 weighted (w), contrast-enhanced T1w (T1wce), T2w, and T2w-FLAIR. Sequence-specific significant features (SF) associated with OS, extracted from the tumour volume, were derived after applying 15 different IN methods. Survival analyses were conducted using Cox proportional hazard (CPH) and Poisson regression (POI) models. A ranking score was assigned based on the 10-fold cross-validated (CV) concordance index (C-I), mean square error (MSE), and the Akaike information criterion (AICs), to evaluate the methods' performance. RESULTS: Scatter plots of the 10-CV C-I and MSE against the AIC showed an impact on the survival predictions between the IN methods and MR sequences (C1/C2 C-I range: 0.62-0.71/0.61-0.72, MSE range: 0.20-0.42/0.13-0.22). White stripe showed stable results for T1wce (C1/C2 C-I: 0.71/0.65, MSE: 0.21/0.14). Combat (0.68/0.62, 0.22/0.15) and histogram matching (HM, 0.67/0.64, 0.22/0.15) showed consistent prediction results for T2w models. They were also the top-performing methods for T1w in C2 (Combat: 0.67, 0.13; HM: 0.67, 0.13); however, only HM achieved high predictions in C1 (0.66, 0.22). After eliminating IN impacted SF using Spearman's rank-order correlation coefficient, a mean decrease in the C-I and MSE of 0.05 and 0.03 was observed in all four sequences. CONCLUSION: The IN method impacted the predictive power of survival models; thus, performance is sequence-dependent.

MR-Class: A Python Tool for Brain MR Image Classification Utilizing One-vs-All DCNNs to Deal with the Open-Set Recognition Problem.

BACKGROUND: MR image classification in datasets collected from multiple sources is complicated by inconsistent and missing DICOM metadata. Therefore, we aimed to establish a method for the efficient automatic classification of MR brain sequences. METHODS: Deep convolutional neural networks (DCNN) were trained as one-vs-all classifiers to differentiate between six classes: T1 weighted (w), contrast-enhanced T1w, T2w, T2w-FLAIR, ADC, and SWI. Each classifier yields a probability, allowing threshold-based and relative probability assignment while excluding images with low probability (label: unknown, open-set recognition problem). Data from three high-grade glioma (HGG) cohorts was assessed; C1 (320 patients, 20,101 MRI images) was used for training, while C2 (197, 11,333) and C3 (256, 3522) were for testing. Two raters manually checked images through an interactive labeling tool. Finally, MR-Class' added value was evaluated via radiomics model performance for progression-free survival (PFS) prediction in C2, utilizing the concordance index (C-I). RESULTS: Approximately 10% of annotation errors were observed in each cohort between the DICOM series descriptions and the derived labels. MR-Class accuracy was 96.7% [95% Cl: 95.8, 97.3] for C2 and 94.4% [93.6, 96.1] for C3. A total of 620 images were misclassified; manual assessment of those frequently showed motion artifacts or alterations of anatomy by large tumors. Implementation of MR-Class increased the PFS model C-I by 14.6% on average, compared to a model trained without MR-Class. CONCLUSIONS: We provide a DCNN-based method for the sequence classification of brain MR images and demonstrate its usability in two independent HGG datasets.

Deep Learning-based Automatic Lung Segmentation on Multiresolution CT Scans from Healthy and Fibrotic Lungs in Mice.

Purpose: To develop a model to accurately segment mouse lungs with varying levels of fibrosis and investigate its applicability to mouse images with different resolutions. Materials and Methods: In this experimental retrospective study, a U-Net was trained to automatically segment lungs on mouse CT images. The model was trained (n = 1200), validated (n = 300), and tested (n = 154) on longitudinally acquired and semiautomatically segmented CT images, which included both healthy and irradiated mice (group A). A second independent group of 237 mice (group B) was used for external testing. The Dice score coefficient (DSC) and Hausdorff distance (HD) were used as metrics to quantify segmentation accuracy. Transfer learning was applied to adapt the model to high-spatial-resolution mouse micro-CT segmentation (n = 20; group C [n = 16 for training and n = 4 for testing]). Results: The trained model yielded a high median DSC in both test datasets: 0.984 (interquartile range [IQR], 0.977-0.988) in group A and 0.966 (IQR, 0.955-0.972) in group B. The median HD in both test datasets was 0.47 mm (IQR, 0-0.51 mm [group A]) and 0.31 mm (IQR, 0.30-0.32 mm [group B]). Spatially resolved quantification of differences toward reference masks revealed two hot spots close to the air-tissue interfaces, which are particularly prone to deviation. Finally, for the higher-resolution mouse CT images, the median DSC was 0.905 (IQR, 0.902-0.929) and the median 95th percentile of the HD was 0.33 mm (IQR, 2.61-2.78 mm). Conclusion: The developed deep learning-based method for mouse lung segmentation performed well independently of disease state (healthy, fibrotic, emphysematous lungs) and CT resolution.Keywords: Deep Learning, Lung Fibrosis, Radiation Therapy, Segmentation, Animal Studies, CT, Thorax, Lung Supplemental material is available for this article. Published under a CC BY 4.0 license.

The impact of tumor metabolic activity assessed by 18F-FET amino acid PET imaging in particle radiotherapy of high-grade glioma patients.

Background: Selective uptake of (18)F-fluoro-ethyl-tyrosine (18F-FET) is used in high-grade glioma (HGG) to assess tumor metabolic activity via positron emission tomography (PET). We aim to investigate its value for target volume definition, as a prognosticator, and associations with whole-blood transcriptome liquid biopsy (WBT lbx) for which we recently reported feasibility to mirror tumor characteristics and response to particle irradiation in recurrent HGG (rHGG). Methods: 18F-FET-PET data from n = 43 patients with primary glioblastoma (pGBM) and n = 33 patients with rHGG were assessed. pGBM patients were irradiated with photons and sequential proton/carbon boost, and rHGG patients were treated with carbon re-irradiation (CIR). WBT (Illumina HumanHT-12 Expression BeadChips) lbx was available for n = 9 patients from the rHGG cohort. PET isocontours (40%-70% SUVmax, 10% steps) and MRI-based treatment volumes (MRIvol) were compared using the conformity index (CI) (pGBM, n = 16; rHGG, n = 27). Associations with WBT lbx data were tested on gene expression level and inferred pathways activity scores (PROGENy) and from transcriptome estimated cell fractions (CIBERSORT, xCell). Results: In pGBM, median SUVmax was higher in PET acquired pre-radiotherapy (4.1, range (R) 1.5-7.8; n = 20) vs. during radiotherapy (3.3, R 1.5-5.7, n = 23; p = 0.03) and in non-resected (4.7, R 2.9-7.9; n = 11) vs. resected tumors (3.3, R 1.5-7.8, n = 32; p = 0.01). In rHGG, a trend toward higher SUVmax values in grade IV tumors was observed (p = 0.13). Median MRIvol was 32.34 (R 8.75-108.77) cm3 in pGBM (n = 16) and 20.77 (R 0.63-128.44) cm3 in rHGG patients (n = 27). The highest median CI was observed for 40% (pGBM, 0.31) and 50% (rHGG, 0.43, all tumors) isodose, with 70% (40%) isodose in grade III (IV) rHGG tumors (median CI, 0.38 and 0.49). High SUVmax was linked to shorter survival in pGBM (>3.3, p = 0.001, OR 6.0 [2.1-17.4]) and rHGG (>2.8, p = 0.02, OR 4.1 [1.2-13.9]). SUVmax showed associations with inferred monocyte fractions, hypoxia, and TGFbeta pathway activity and links to immune checkpoint gene expression from WBT lbx. Conclusion: The benefits of 18F-FET-PET imaging on gross tumor volume (GTV) definition for particle radiotherapy warrant further evaluation. SUVmax might assist in prognostic stratification of HGG patients for particle radiotherapy, highlights heterogeneity in rHGG, and is positively associated with unfavorable signatures in peripheral whole-blood transcriptomes.

Joint Optimization of Photon-Carbon Ion Treatments for Glioblastoma.

PURPOSE: Carbon ions are radiobiologically more effective than photons and are beneficial for treating radioresistant gross tumor volumes (GTV). However, owing to a reduced fractionation effect, they may be disadvantageous for treating infiltrative tumors, in which healthy tissue inside the clinical target volume (CTV) must be protected through fractionation. This work addresses the question: What is the ideal combined photon-carbon ion fluence distribution for treating infiltrative tumors given a specific fraction allocation between photons and carbon ions? METHODS AND MATERIALS: We present a method to simultaneously optimize sequentially delivered intensity modulated photon (IMRT) and carbon ion (CIRT) treatments based on cumulative biological effect, incorporating both the variable relative biological effect of carbon ions and the fractionation effect within the linear quadratic model. The method is demonstrated for 6 glioblastoma patients in comparison with the current clinical standard of independently optimized CIRT-IMRT plans. RESULTS: Compared with the reference plan, joint optimization strategies yield inhomogeneous photon and carbon ion dose distributions that cumulatively deliver a homogeneous biological effect distribution. In the optimal distributions, the dose to CTV is mostly delivered by photons and carbon ions are restricted to the GTV with variations depending on tumor size and location. Improvements in conformity of high-dose regions are reflected by a mean EQD2 reduction of 3.29 ± 1.22 Gy in a dose fall-off margin around the CTV. Carbon ions may deliver higher doses to the center of the GTV, and photon contributions are increased at interfaces with CTV and critical structures. This results in a mean EQD2 reduction of 8.3 ± 2.28 Gy, in which the brain stem abuts the target volumes. CONCLUSIONS: We have developed a biophysical model to optimize combined photon-carbon ion treatments. For 6 glioblastoma patient cases, we show that our approach results in a more targeted application of carbon ions that (1) reduces dose in normal tissues within the target volume, which can only be protected through fractionation; and (2) boosts central target volume regions to reduce integral dose. Joint optimization of IMRT-CIRT treatments enable the exploration of a new spectrum of plans that can better address physical and radiobiological treatment planning challenges.

Deep Convolution Neural Network (DCNN) Multiplane Approach to Synthetic CT Generation From MR images-Application in Brain Proton Therapy.

PURPOSE: The first aim of this work is to present a novel deep convolution neural network (DCNN) multiplane approach and compare it to single-plane prediction of synthetic computed tomography (sCT) by using the real computed tomography (CT) as ground truth. The second aim is to demonstrate the feasibility of magnetic resonance imaging (MRI)-based proton therapy planning for the brain by assessing the range shift error within the clinical acceptance threshold. METHODS AND MATERIALS: The image database included 15 pairs of MRI/CT scans of the head. Three DCNNs were trained to estimate, for each voxel, the Hounsfield unit (HU) value from MRI intensities. Each DCNN gave an estimation in the axial, sagittal, and coronal plane, respectively. The median HU among the 3 values was selected to build the sCT. The sCT/CT agreement was evaluated by a mean absolute error (MAE) and mean error, computed within the head contour and on 6 different tissues. Dice similarity coefficients were calculated to assess the geometric overlap of bone and air cavities segmentations. A 3-beam proton therapy plan was simulated for each patient. Beam-by-beam range shift (RS) analysis was conducted to assess the proton-stopping power estimation. RS analysis was performed using clinically accepted thresholds of (1) 3.5% + 1 mm and (2) 2.5% + 1.5 mm of the total range. RESULTS: DCNN multiplane statistically outperformed single-plane prediction of sCT (P < .025). MAE and mean error within the head were 54 ± 7 HU and -4 ± 17 HU (mean ± standard deviation), respectively. Soft tissues were very close to perfect agreement (11 ± 3 HU in terms of MAE). Segmentation of air and bone regions led to a Dice similarity coefficient of 0.92 ± 0.03 and 0.93 ± 0.02, respectively. Proton RS was always below clinical acceptance thresholds, with a relative RS error of 0.14% ± 1.11%. CONCLUSIONS: The multiplane DCNN approach significantly improved the sCT prediction compared with other DCNN methods presented in the literature. The method was demonstrated to be highly accurate for MRI-only proton planning purposes.