Brain extraction is a critical preprocessing step in the analysis of neuroimaging studies conducted with magnetic resonance imaging (MRI) and influences the accuracy of downstream analyses. The ...majority of brain extraction algorithms are, however, optimized for processing healthy brains and thus frequently fail in the presence of pathologically altered brain or when applied to heterogeneous MRI datasets. Here we introduce a new, rigorously validated algorithm (termed HD‐BET) relying on artificial neural networks that aim to overcome these limitations. We demonstrate that HD‐BET outperforms six popular, publicly available brain extraction algorithms in several large‐scale neuroimaging datasets, including one from a prospective multicentric trial in neuro‐oncology, yielding state‐of‐the‐art performance with median improvements of +1.16 to +2.50 points for the Dice coefficient and −0.66 to −2.51 mm for the Hausdorff distance. Importantly, the HD‐BET algorithm, which shows robust performance in the presence of pathology or treatment‐induced tissue alterations, is applicable to a broad range of MRI sequence types and is not influenced by variations in MRI hardware and acquisition parameters encountered in both research and clinical practice. For broader accessibility, the HD‐BET prediction algorithm is made freely available (www.neuroAI-HD.org) and may become an essential component for robust, automated, high‐throughput processing of MRI neuroimaging data.
Purpose:
To study the performance of different dual energy computed tomography (DECT) techniques, which are available today, and future multi energy CT (MECT) employing novel photon counting ...detectors in an image‐based material decomposition task.
Methods:
The material decomposition performance of different energy‐resolved CT acquisition techniques is assessed and compared in a simulation study of virtual non‐contrast imaging and iodine quantification. The material‐specific images are obtained via a statistically optimal image‐based material decomposition. A projection‐based maximum likelihood approach was used for comparison with the authors’ image‐based method. The different dedicated dual energy CT techniques are simulated employing realistic noise models and x‐ray spectra. The authors compare dual source DECT with fast kV switching DECT and the dual layer sandwich detector DECT approach. Subsequent scanning and a subtraction method are studied as well. Further, the authors benchmark future MECT with novel photon counting detectors in a dedicated DECT application against the performance of today's DECT using a realistic model. Additionally, possible dual source concepts employing photon counting detectors are studied.
Results:
The DECT comparison study shows that dual source DECT has the best performance, followed by the fast kV switching technique and the sandwich detector approach. Comparing DECT with future MECT, the authors found noticeable material image quality improvements for an ideal photon counting detector; however, a realistic detector model with multiple energy bins predicts a performance on the level of dual source DECT at 100 kV/Sn 140 kV. Employing photon counting detectors in dual source concepts can improve the performance again above the level of a single realistic photon counting detector and also above the level of dual source DECT.
Conclusions:
Substantial differences in the performance of today's DECT approaches were found for the application of virtual non‐contrast and iodine imaging. Future MECT with realistic photon counting detectors currently can only perform comparably to dual source DECT at 100 kV/Sn 140 kV. Dual source concepts with photon counting detectors could be a solution to this problem, promising a better performance.
Objectives
To evaluate the detection rates of targeted and systematic biopsies in magnetic resonance imaging (MRI) and ultrasound (US) image‐fusion transperineal prostate biopsy for patients with ...previous benign transrectal biopsies in two high‐volume centres.
Patients and Methods
A two centre prospective outcome study of 487 patients with previous benign biopsies that underwent transperineal MRI/US fusion‐guided targeted and systematic saturation biopsy from 2012 to 2015. Multiparametric MRI (mpMRI) was reported according to Prostate Imaging Reporting and Data System (PI‐RADS) Version 1. Detection of Gleason score 7–10 prostate cancer on biopsy was the primary outcome. Positive (PPV) and negative (NPV) predictive values including 95% confidence intervals (95% CIs) were calculated. Detection rates of targeted and systematic biopsies were compared using McNemar's test.
Results
The median (interquartile range) PSA level was 9.0 (6.7–13.4) ng/mL. PI‐RADS 3–5 mpMRI lesions were reported in 343 (70%) patients and Gleason score 7–10 prostate cancer was detected in 149 (31%). The PPV (95% CI) for detecting Gleason score 7–10 prostate cancer was 0.20 (±0.07) for PI‐RADS 3, 0.32 (±0.09) for PI‐RADS 4, and 0.70 (±0.08) for PI‐RADS 5. The NPV (95% CI) of PI‐RADS 1–2 was 0.92 (±0.04) for Gleason score 7–10 and 0.99 (±0.02) for Gleason score ≥4 + 3 cancer. Systematic biopsies alone found 125/138 (91%) Gleason score 7–10 cancers. In patients with suspicious lesions (PI‐RADS 4–5) on mpMRI, systematic biopsies would not have detected 12/113 significant prostate cancers (11%), while targeted biopsies alone would have failed to diagnose 10/113 (9%). In equivocal lesions (PI‐RADS 3), targeted biopsy alone would not have diagnosed 14/25 (56%) of Gleason score 7–10 cancers, whereas systematic biopsies alone would have missed 1/25 (4%). Combination with PSA density improved the area under the curve of PI‐RADS from 0.822 to 0.846.
Conclusion
In patients with high probability mpMRI lesions, the highest detection rates of Gleason score 7–10 cancer still required combined targeted and systematic MRI/US image‐fusion; however, systematic biopsy alone may be sufficient in patients with equivocal lesions. Repeated prostate biopsies may not be needed at all for patients with a low PSA density and a negative mpMRI read by experienced radiologists.
To compare changes in signal intensity (SI) ratios of the dentate nucleus (DN) and the globus pallidus (GP) to those of other structures on unenhanced T1-weighted magnetic resonance (MR) images ...between linear and macrocyclic gadolinium-based contrast agents (GBCAs).
The study was approved by the ethical committee of the University of Heidelberg (reference no. S-324/2014). Owing to the retrospective character of the study, the ethical committee did not require any written informed consent. Two groups of 50 patients who underwent at least six consecutive MR imaging examinations with the exclusive use of either a linear GBCA (gadopentetate dimeglumine) or a macrocyclic GBCA (gadoterate meglumine) were analyzed retrospectively. The difference in mean SI ratios of DN to pons and GP to thalamus on unenhanced T1-weighted images from the last and first examinations was calculated. One-sample and independent-sample t tests were used to assess the difference in SI ratios for both groups, and regression analysis was performed to account for potential confounders.
The SI ratio difference in the linear group was greater than 0 (mean DN difference ± standard deviation, 0.0407 ± 0.0398 P < .001; GP, 0.0287 ± 0.0275 P < .001) and significantly larger (DN, P < .001 and standardized difference of 1.16; GP, P < .001 and standardized difference of 0.81) than that in the macrocyclic group, which did not differ from 0 (DN, 0.0016 ± 0.0266 P = .680; GP, 0.0031 ± 0.0354 P = .538). The SI ratio difference between the last and first examinations for the DN remained significantly different between the two groups in the regression analysis (P < .001).
This study indicates that an SI increase in the DN and GP on T1-weighted images is caused by serial application of the linear GBCA gadopentetate dimeglumine but not by the macrocyclic GBCA gadoterate meglumine. Clinical implications of this observation remain unclear.
Background
Patients with newly diagnosed inoperable glioma receive chemoradiotherapy (CRT). Standard Response Assessment in Neuro‐Oncology (RANO) takes a minimum of 4 weeks after the end of ...treatment.
Purpose/Hypothesis
To investigate whether chemical exchange saturation transfer (CEST) MRI enables earlier assessment of response to CRT in glioma patients.
Study Type
Longitudinal prospective study.
Population
Twelve brain tumor patients who underwent definitive CRT were included in this study. Three longitudinal CEST MRI measurements were performed for each patient at 7T: first before, second immediately after completion of CRT, and a third measurement as a 6‐week follow‐up.
Field Strength/Sequence
Conventional MRI (contrast‐enhanced, T2w and diffusion‐weighted imaging) at 3T and T2w and CEST MRI at 7T was performed for all patients.
Assessment
The mean relaxation‐compensated relayed nuclear‐Overhauser‐effect CEST signal (rNOE) and the mean downfield‐rNOE‐suppressed amide proton transfer (dns‐APT) CEST signal were investigated. Additionally, choline‐to‐N‐acetyl‐aspartate ratios (Cho/NAA) were evaluated using single‐voxel 1H‐MRS in six of these patients. Performance of obtained contrasts was analyzed in assessing treatment response as classified according to the updated RANO criteria.
Statistical Test
Unpaired Student's t‐test.
Results
The rNOE signal significantly separated stable and progressive disease directly after the end of therapy (post‐treatment normalized to pre‐treatment mean ± SD: rNOEresponder = 1.090 ± 0.110, rNOEnon‐responder = 0.808 ± 0.155, P = 0.015). In contrast, no significant difference was observed between either group when assessing the normalized dns‐APT (dns‐APTresponder = 0.953 ± 0.384, dns‐APTnon‐responder = 0.972 ± 0.477, P = 0.95). In the smaller MRS subcohort, normalized Cho/NAA decreased in therapy responders (Cho/NAAresponder = 0.632 ± 0.007, Cho/NAAnon‐responder = 0.946 ± 0.124, P = 0.070).
Data Conclusion
rNOE mediated CEST imaging at 7T allowed for discrimination of responders and non‐responders immediately after the end of CRT, additionally supported by 1H‐MRS data. This is at least 4 weeks earlier than the standard clinical evaluation according to RANO. Therefore, CEST MRI may enable early response assessment in glioma patients.
Level of Evidence: 1
Technical Efficacy Stage: 5
J. Magn. Reson. Imaging 2019;50:1268–1277.
Purpose
Relaxation‐compensated CEST‐MRI (i.e., the inverse metrics magnetization transfer ratio and apparent exchange‐dependent relaxation) has already been shown to provide valuable information for ...brain tumor diagnosis at ultrahigh magnetic field strengths. This study aims at translating the established acquisition protocol at 7 T to a clinically relevant magnetic field strength of 3 T.
Methods
Protein model solutions were analyzed at multiple magnetic field strengths to assess the spectral widths of the amide proton transfer and relayed nuclear Overhauser effect (rNOE) signals at 3 T. This prior knowledge of the spectral range of CEST signals enabled a reliable and stable Lorentzian‐fitting also at 3 T where distinct peaks are no longer resolved in the Z‐spectrum. In comparison to the established acquisition protocol at 7 T, also the image readout was extended to three dimensions.
Results
The observed spectral range of CEST signals at 3 T was approximately ±15 ppm. Final relaxation‐compensated amide proton transfer and relayed nuclear Overhauser effect contrasts were in line with previous results at 7 T. Examination of a patient with glioblastoma demonstrated the applicability of this acquisition protocol in a clinical setting.
Conclusion
The presented acquisition protocol allows relaxation‐compensated CEST‐MRI at 3 T with a 3D coverage of the human brain. Translation to a clinically relevant magnetic field strength of 3 T opens the door to trials with a large number of participants, thus enabling a comprehensive assessment of the clinical relevance of relaxation compensation in CEST‐MRI.
The Response Assessment in Neuro-Oncology (RANO) criteria and requirements for a uniform protocol have been introduced to standardise assessment of MRI scans in both clinical trials and clinical ...practice. However, these criteria mainly rely on manual two-dimensional measurements of contrast-enhancing (CE) target lesions and thus restrict both reliability and accurate assessment of tumour burden and treatment response. We aimed to develop a framework relying on artificial neural networks (ANNs) for fully automated quantitative analysis of MRI in neuro-oncology to overcome the inherent limitations of manual assessment of tumour burden.
In this retrospective study, we compiled a single-institution dataset of MRI data from patients with brain tumours being treated at Heidelberg University Hospital (Heidelberg, Germany; Heidelberg training dataset) to develop and train an ANN for automated identification and volumetric segmentation of CE tumours and non-enhancing T2-signal abnormalities (NEs) on MRI. Independent testing and large-scale application of the ANN for tumour segmentation was done in a single-institution longitudinal testing dataset from the Heidelberg University Hospital and in a multi-institutional longitudinal testing dataset from the prospective randomised phase 2 and 3 European Organisation for Research and Treatment of Cancer (EORTC)-26101 trial (NCT01290939), acquired at 38 institutions across Europe. In both longitudinal datasets, spatial and temporal tumour volume dynamics were automatically quantified to calculate time to progression, which was compared with time to progression determined by RANO, both in terms of reliability and as a surrogate endpoint for predicting overall survival. We integrated this approach for fully automated quantitative analysis of MRI in neuro-oncology within an application-ready software infrastructure and applied it in a simulated clinical environment of patients with brain tumours from the Heidelberg University Hospital (Heidelberg simulation dataset).
For training of the ANN, MRI data were collected from 455 patients with brain tumours (one MRI per patient) being treated at Heidelberg hospital between July 29, 2009, and March 17, 2017 (Heidelberg training dataset). For independent testing of the ANN, an independent longitudinal dataset of 40 patients, with data from 239 MRI scans, was collected at Heidelberg University Hospital in parallel with the training dataset (Heidelberg test dataset), and 2034 MRI scans from 532 patients at 34 institutions collected between Oct 26, 2011, and Dec 3, 2015, in the EORTC-26101 study were of sufficient quality to be included in the EORTC-26101 test dataset. The ANN yielded excellent performance for accurate detection and segmentation of CE tumours and NE volumes in both longitudinal test datasets (median DICE coefficient for CE tumours 0·89 95% CI 0·86–0·90, and for NEs 0·93 0·92–0·94 in the Heidelberg test dataset; CE tumours 0·91 0·90–0·92, NEs 0·93 0·93–0·94 in the EORTC-26101 test dataset). Time to progression from quantitative ANN-based assessment of tumour response was a significantly better surrogate endpoint than central RANO assessment for predicting overall survival in the EORTC-26101 test dataset (hazard ratios ANN 2·59 95% CI 1·86–3·60 vs central RANO 2·07 1·46–2·92; p<0·0001) and also yielded a 36% margin over RANO (p<0·0001) when comparing reliability values (ie, agreement in the quantitative volumetrically defined time to progression based on radiologist ground truth vs automated assessment with ANN of 87% 266 of 306 with sufficient data compared with 51% 155 of 306 with local vs independent central RANO assessment). In the Heidelberg simulation dataset, which comprised 466 patients with brain tumours, with 595 MRI scans obtained between April 27, and Sept 17, 2018, automated on-demand processing of MRI scans and quantitative tumour response assessment within the simulated clinical environment required 10 min of computation time (average per scan).
Overall, we found that ANN enabled objective and automated assessment of tumour response in neuro-oncology at high throughput and could ultimately serve as a blueprint for the application of ANN in radiology to improve clinical decision making. Future research should focus on prospective validation within clinical trials and application for automated high-throughput imaging biomarker discovery and extension to other diseases.
Medical Faculty Heidelberg Postdoc-Program, Else Kröner-Fresenius Foundation.
Radiation therapy (RT) continues to be one of the mainstays of cancer treatment. Considerable efforts have been recently devoted to integrating MRI into clinical RT planning and monitoring. This ...integration, known as MRI-guided RT, has been motivated by the superior soft-tissue contrast, organ motion visualization, and ability to monitor tumor and tissue physiologic changes provided by MRI compared with CT. Offline MRI is already used for treatment planning at many institutions. Furthermore, MRI-guided linear accelerator systems, allowing use of MRI during treatment, enable improved adaptation to anatomic changes between RT fractions compared with CT guidance. Efforts are underway to develop real-time MRI-guided intrafraction adaptive RT of tumors affected by motion and MRI-derived biomarkers to monitor treatment response and potentially adapt treatment to physiologic changes. These developments in MRI guidance provide the basis for a paradigm change in treatment planning, monitoring, and adaptation. Key challenges to advancing MRI-guided RT include real-time volumetric anatomic imaging, addressing image distortion because of magnetic field inhomogeneities, reproducible quantitative imaging across different MRI systems, and biologic validation of quantitative imaging. This review describes emerging innovations in offline and online MRI-guided RT, exciting opportunities they offer for advancing research and clinical care, hurdles to be overcome, and the need for multidisciplinary collaboration.