Quantitative susceptibility mapping through model-based deep image prior (MoDIP).

The data-driven approach of supervised learning methods has limited applicability in solving dipole inversion in Quantitative Susceptibility Mapping (QSM) with varying scan parameters across different objects. To address this generalization issue in supervised QSM methods, we propose a novel training-free model-based unsupervised method called MoDIP (Model-based Deep Image Prior). MoDIP comprises a small, untrained network and a Data Fidelity Optimization (DFO) module. The network converges to an interim state, acting as an implicit prior for image regularization, while the optimization process enforces the physical model of QSM dipole inversion. Experimental results demonstrate MoDIP's excellent generalizability in solving QSM dipole inversion across different scan parameters. It exhibits robustness against pathological brain QSM, achieving over 32 % accuracy improvement than supervised deep learning methods. It is also 33 % more computationally efficient and runs 4 times faster than conventional DIP-based approaches, enabling 3D high-resolution image reconstruction in under 4.5 min.

The therapeutic effect of radiotherapy combined with systemic therapy compared to radiotherapy alone in patients with simple brain metastasis after first-line treatment of limited-stage small cell lung cancer: a retrospective study.

PURPOSE: We aimed to compare the therapeutic effect of radiotherapy (RT) plus systemic therapy (ST) with RT alone in patients with simple brain metastasis (BM) after first-line treatment of limited-stage small cell lung cancer (LS-SCLC). METHODS: The patients were treated at a single center from January 2011 to January 2022. BM only without metastases to other organs was defined as simple BM. The eligible patients were divided into RT alone (monotherapy arm) and RT plus ST (combined therapy arm). Univariate and multivariate Cox proportional hazards analyses were used to examine factors associated with increased risk of extracranial progression. After 1:1 propensity score matching analysis, two groups were compared for extracranial progression-free survival (ePFS), PFS, overall survival (OS), and intracranial PFS (iPFS). RESULTS: 133 patients were identified and 100 were analyzed (monotherapy arm: n = 50, combined therapy arm: n = 50). The ePFS of the combined therapy was significantly longer than that of the monotherapy, with a median ePFS of 13.2 months (95% CI, 6.6-19.8) in combined therapy and 8.2 months (95% CI, 5.7-10.7) in monotherapy (P = 0.04). There were no statistically significant differences in PFS (P = 0.057), OS (P = 0.309), or iPFS (P = 0.448). Multifactorial analysis showed that combined therapy was independently associated with better ePFS compared with monotherapy (HR = 0.617, P = 0.034); more than 5 BMs were associated with worse ePFS compared with 1-5 BMs (HR = 1.808, P = 0.012). CONCLUSIONS: Compared with RT alone, combined therapy improves ePFS in patients with simple BM after first-line treatment of LS-SCLC. Combined therapy and 1-5 BMs reduce the risk of extracranial recurrence.

Affine transformation edited and refined deep neural network for quantitative susceptibility mapping.

Deep neural networks have demonstrated great potential in solving dipole inversion for Quantitative Susceptibility Mapping (QSM). However, the performances of most existing deep learning methods drastically degrade with mismatched sequence parameters such as acquisition orientation and spatial resolution. We propose an end-to-end AFfine Transformation Edited and Refined (AFTER) deep neural network for QSM, which is robust against arbitrary acquisition orientation and spatial resolution up to 0.6 mm isotropic at the finest. The AFTER-QSM neural network starts with a forward affine transformation layer, followed by a Unet for dipole inversion, then an inverse affine transformation layer, followed by a Residual Dense Network (RDN) for QSM refinement. Simulation and in-vivo experiments demonstrated that the proposed AFTER-QSM network architecture had excellent generalizability. It can successfully reconstruct susceptibility maps from highly oblique and anisotropic scans, leading to the best image quality assessments in simulation tests and suppressed streaking artifacts and noise levels for in-vivo experiments compared with other methods. Furthermore, ablation studies showed that the RDN refinement network significantly reduced image blurring and susceptibility underestimation due to affine transformations. In addition, the AFTER-QSM network substantially shortened the reconstruction time from minutes using conventional methods to only a few seconds.

Detecting monocyte trafficking in an animal model of glioblastoma using R2* and quantitative susceptibility mapping.

BACKGROUND: The role of tumor-associated macrophages (TAMs) in glioblastoma (GBM) disease progression has received increasing attention. Recent advances have shown that TAMs can be re-programmed to exert a pro-inflammatory, anti-tumor effect to control GBMs. However, imaging methods capable of differentiating tumor progression from immunotherapy treatment effects have been lacking, making timely assessment of treatment response difficult. We showed that tracking monocytes using iron oxide nanoparticle (USPIO) with MRI can be a sensitive imaging method to detect therapy response directed at the innate immune system. METHODS: We implanted syngeneic mouse glioma stem cells into C57/BL6 mice and treated the animals with either niacin (a stimulator of innate immunity) or vehicle. Animals were imaged using an anatomical MRI sequence, R2* mapping, and quantitative susceptibility mapping (QSM) before and after USPIO injection. RESULTS: Compared to vehicles, niacin-treated animals showed significantly higher susceptibility and R2*, representing USPIO and monocyte infiltration into the tumor. We observed a significant reduction in tumor size in the niacin-treated group 7 days later. We validated our MRI results with flow cytometry and immunofluoresence, which showed that niacin decreased pro-inflammatory Ly6C high monocytes in the blood but increased CD16/32 pro-inflammatory macrophages within the tumor, consistent with migration of these pro-inflammatory innate immune cells from the blood to the tumor. CONCLUSION: MRI with USPIO injection can detect therapeutic responses of innate immune stimulating agents before changes in tumor size have occurred, providing a potential complementary imaging technique to monitor cancer immunotherapies. MANUSCRIPT HIGHLIGHT: We show that iron oxide nanoparticles (USPIOs) can be used to label innate immune cells and detect the trafficking of pro-inflammatory monocytes into the glioblastoma. This preceded changes in tumor size, making it a more sensitive imaging technique.

Distortion-corrected image reconstruction with deep learning on an MRI-Linac.

PURPOSE: MRI is increasingly utilized for image-guided radiotherapy due to its outstanding soft-tissue contrast and lack of ionizing radiation. However, geometric distortions caused by gradient nonlinearities (GNLs) limit anatomical accuracy, potentially compromising the quality of tumor treatments. In addition, slow MR acquisition and reconstruction limit the potential for effective image guidance. Here, we demonstrate a deep learning-based method that rapidly reconstructs distortion-corrected images from raw k-space data for MR-guided radiotherapy applications. METHODS: We leverage recent advances in interpretable unrolling networks to develop a Distortion-Corrected Reconstruction Network (DCReconNet) that applies convolutional neural networks (CNNs) to learn effective regularizations and nonuniform fast Fourier transforms for GNL-encoding. DCReconNet was trained on a public MR brain dataset from 11 healthy volunteers for fully sampled and accelerated techniques, including parallel imaging (PI) and compressed sensing (CS). The performance of DCReconNet was tested on phantom, brain, pelvis, and lung images acquired on a 1.0T MRI-Linac. The DCReconNet, CS-, PI-and UNet-based reconstructed image quality was measured by structural similarity (SSIM) and RMS error (RMSE) for numerical comparisons. The computation time and residual distortion for each method were also reported. RESULTS: Imaging results demonstrated that DCReconNet better preserves image structures compared to CS- and PI-based reconstruction methods. DCReconNet resulted in the highest SSIM (0.95 median value) and lowest RMSE (<0.04) on simulated brain images with four times acceleration. DCReconNet is over 10-times faster than iterative, regularized reconstruction methods. CONCLUSIONS: DCReconNet provides fast and geometrically accurate image reconstruction and has the potential for MRI-guided radiotherapy applications.

Quantitative Susceptibility Mapping Changes Relate to Gait Issues in Parkinson's Disease.

BACKGROUND: Quantitative susceptibility mapping (QSM) demonstrates elevated iron content in Parkinson's disease (PD) patients within the basal ganglia, though it has infrequently been studied in relation to gait difficulties including freezing of gait (FOG). Our purpose was to relate QSM of basal ganglia and extra-basal ganglia structures with qualitative and quantitative gait measures in PD. METHODS: This case-control study included PD and cognitively unimpaired (CU) participants from the Comprehensive Assessment of Neurodegeneration and Dementia study. Whole brain QSM was acquired at 3T. Region of interests (ROIs) were drawn blinded manually in the caudate nucleus, putamen, globus pallidus, pulvinar nucleus of the thalamus, red nucleus, substantia nigra, and dentate nucleus. Susceptibilities of ROIs were compared between PD and CU. Items from the FOG questionnaire and quantitative gait measures from PD participants were compared to susceptibilities. RESULTS: Twenty-nine participants with PD and 27 CU participants were included. There was no difference in susceptibility values in any ROI when comparing CU versus PD (p > 0.05 for all). PD participants with gait impairment (n = 23) had significantly higher susceptibility in the putamen (p = 0.008), red nucleus (p = 0.01), and caudate nucleus (p = 0.03) compared to those without gait impairment (n = 6). PD participants with FOG (n = 12) had significantly higher susceptibility in the globus pallidus (p = 0.03) compared to those without FOG (n = 17). Among quantitative gait measures, only stride time variability was significantly different between those with and without FOG (p = 0.04). CONCLUSION: Susceptibilities in basal ganglia and extra-basal ganglia structures are related to qualitative measures of gait impairment and FOG in PD.

Instant tissue field and magnetic susceptibility mapping from MRI raw phase using Laplacian enhanced deep neural networks.

Quantitative susceptibility mapping (QSM) is an MRI post-processing technique that produces spatially resolved magnetic susceptibility maps from phase data. However, the traditional QSM reconstruction pipeline involves multiple non-trivial steps, including phase unwrapping, background field removal, and dipole inversion. These intermediate steps not only increase the reconstruction time but accumulates errors. This study aims to overcome existing limitations by developing a Laplacian-of-Trigonometric-functions (LoT) enhanced deep neural network for near-instant quantitative field and susceptibility mapping (i.e., iQFM and iQSM) from raw MRI phase data. The proposed iQFM and iQSM methods were compared with established reconstruction pipelines on simulated and in vivo datasets. In addition, experiments on patients with intracranial hemorrhage and multiple sclerosis were also performed to test the generalization of the proposed neural networks. The proposed iQFM and iQSM methods in healthy subjects yielded comparable results to those involving the intermediate steps while dramatically improving reconstruction accuracies on intracranial hemorrhages with large susceptibilities. High susceptibility contrast between multiple sclerosis lesions and healthy tissue was also achieved using the proposed methods. Comparative studies indicated that the most significant contributor to iQFM and iQSM over conventional multi-step methods was the elimination of traditional Laplacian unwrapping. The reconstruction time on the order of minutes for traditional approaches was shortened to around 0.1 s using the trained iQFM and iQSM neural networks.

Risk factors of esophageal fistula induced by re-radiotherapy for recurrent esophageal cancer with local primary site.

PURPOSE: The purpose of the present study was to investigate risk factors for esophageal fistula (EF) in patients with recurrent esophageal cancer receiving re-radiotherapy with or without chemotherapy. METHODS: We reviewed retrospectively the clinical characters and dosimetric parameters of 96 patients with recurrent esophageal cancer treated with re-radiotherapy in Cancer Hospital Affiliated to Shandong First Medical University between August 2014 and January 2021.Univariate and multivariate logistic regression analyses were provided to determine the risk factors of EF induced by re-radiotherapy. RESULTS: The median time interval between two radiotherapy was 23.35 months (range, 4.30 to 238.10 months). EF occurred in 19 patients (19.79%). In univariate analysis, age, T stage, the biologically equivalent dose in the re-radiotherapy, total biologically equivalent dose, hyperfractionated radiotherapy, ulcerative esophageal cancer, the length of tumor and the maximum thickness of tumor had a correlation with the prevalence of EF. In addition, age (HR = 0.170, 95%CI 0.030-0.951, p = 0.044), T stage (HR = 8.369, 95%CI 1.729-40.522, p = 0.008), ulcerative esophageal cancer (HR = 5.810, 95%CI 1.316-25.650, p = 0.020) and the maximum thickness of tumor (HR = 1.314, 95%CI 1.098-1.572, p = 0.003) were risk factors of EF in multivariate logistic regression analysis. CONCLUSIONS: The incidence of EF was significantly increased in patients with recurrent esophageal cancer who underwent re-radiotherapy. This study revealed that age, T stage, ulcerative esophageal cancer and the maximum thickness of the tumor were risk factors associated with EF. In clinical work, patients with risk factors for EF ought to be highly concerned and individualized treatment plans should be taken to reduce the occurrence of EF.

Brain volume and magnetic susceptibility differences in children and adolescents with prenatal alcohol exposure.

BACKGROUND: Prenatal alcohol exposure (PAE) can negatively affect brain development thereby increasing the risk of cognitive deficits, behavioral challenges, and mental health problems. Brain iron is important for a number of physiological processes for healthy brain development. Animal studies show that PAE reduced brain iron; however, this has not been investigated in human children with PAE. METHODS: We studied 20 children and adolescents with PAE and 44 unexposed children and adolescents aged 7.5 to 15 years. All children underwent quantitative susceptibility mapping and T1-weighted magnetic resonance imaging scans. Susceptibility and volume measurements of the caudate, putamen, pallidum, thalamus, amygdala, hippocampus, and nucleus accumbens were extracted using FreeSurfer. ANCOVAs were used to compare volume and susceptibility between groups for each region of interest, controlling for age and gender. For structures where susceptibility differed by group, we also tested for an association between intelligence quotient (IQ) and susceptibility. RESULTS: There were no significant group differences in susceptibility after multiple comparison correction, though the PAE group had higher susceptibility in the thalamus compared to unexposed participants before correction (p = 0.032, q = 0.230). There was no association between IQ and thalamus susceptibility. The PAE group had significantly lower volume in the bilateral caudate, bilateral pallidum, and left putamen. CONCLUSIONS: These findings suggest susceptibility may be altered in children and adolescents with PAE, though more research is needed. Volume reductions are consistent with previous literature and likely underlie cognitive and behavioral deficits associated with PAE.

The effects of resistance exercise on body composition and physical function in prostate cancer patients undergoing androgen deprivation therapy: an update systematic review and meta-analysis.

OBJECTIVE: The aim of the meta-analysis was to explore effects of resistance exercise (RE) on body composition and physical function in patients with prostate cancer (PCa). DATA SOURCES: We searched the electronic databases of Pubmed, Embase, Cochrane, and web of science. Published studies have been collected from these databases. Search terms include resistance training, strength training, RE, androgen suppression therapy, androgen deprivation therapy and PCa, with a deadline of 31 March 2022. MAIN RESULTS: These studies showed significant improvements of body composition(Lean body mass MD: 1.12 95% CI [0.48, 1.76], p < 0.01; Body fat rate MD: -1.12 95% CI [-1.99,-0.24], p < 0.05; Appendicular skeletal mass MD: 0.74 95% CI [0.45, 1.03], p < 0.01) and physical function (leg press MD: 77.95 95% CI [38.90, 117.00], p < 0.01; stair climb MD:-0.30 95% CI [-0.49, -0.12], p < 0.01). In addition, the improvement of Body fat mass (MD: -0.21 95% CI [-0.79, 0.37], p > 0.05), 400 m walk (MD: -21.74 95% CI [-45.53, 2.05], p > 0.05) and times up and go (MD: -0.50 95% CI [-1.03, 0.03], p > 0.05) were not obvious. Subgroup analyses showed that RE for ≥ 6 months (compared with RE intervention for < 6 months) and starting exercise immediately after androgen deprivation therapy (ADT) (compared with delayed exercise after ADT) resulted in more significant improvements in body composition. Furthermore, the results showed that the exercise intensity of 8-12 RM significantly improved body composition. CONCLUSIONS: RE seems to be a promising approach in order to improve body composition and physical function in PCa patients to offset their treatment-related side effects. RE should be used as a means of rehabilitation and care for PCa. Starting exercise immediately after ADT and extending exercise time while choosing the right intensity can better improve the patients' body composition and function. REGISTRATION NUMBER: INPLASY202280019.

Accelerating quantitative susceptibility and R2* mapping using incoherent undersampling and deep neural network reconstruction.

Quantitative susceptibility mapping (QSM) and R2* mapping are MRI post-processing methods that quantify tissue magnetic susceptibility and transverse relaxation rate distributions. However, QSM and R2* acquisitions are relatively slow, even with parallel imaging. Incoherent undersampling and compressed sensing reconstruction techniques have been used to accelerate traditional magnitude-based MRI acquisitions; however, most do not recover the full phase signal, as required by QSM, due to its non-convex nature. In this study, a learning-based Deep Complex Residual Network (DCRNet) is proposed to recover both the magnitude and phase images from incoherently undersampled data, enabling high acceleration of QSM and R2* acquisition. Magnitude, phase, R2*, and QSM results from DCRNet were compared with two iterative and one deep learning methods on retrospectively undersampled acquisitions from six healthy volunteers, one intracranial hemorrhage and one multiple sclerosis patients, as well as one prospectively undersampled healthy subject using a 7T scanner. Peak signal to noise ratio (PSNR), structural similarity (SSIM), root-mean-squared error (RMSE), and region-of-interest susceptibility and R2* measurements are reported for numerical comparisons. The proposed DCRNet method substantially reduced artifacts and blurring compared to the other methods and resulted in the highest PSNR, SSIM, and RMSE on the magnitude, R2*, local field, and susceptibility maps. Compared to two iterative and one deep learning methods, the DCRNet method demonstrated a 3.2% to 9.1% accuracy improvement in deep grey matter susceptibility when accelerated by a factor of four. The DCRNet also dramatically shortened the reconstruction time of single 2D brain images from 36-140 seconds using conventional approaches to only 15-70 milliseconds.

xQSM: quantitative susceptibility mapping with octave convolutional and noise-regularized neural networks.

Quantitative susceptibility mapping (QSM) provides a valuable MRI contrast mechanism that has demonstrated broad clinical applications. However, the image reconstruction of QSM is challenging due to its ill-posed dipole inversion process. In this study, a new deep learning method for QSM reconstruction, namely xQSM, was designed by introducing noise regularization and modified octave convolutional layers into a U-net backbone and trained with synthetic and in vivo datasets, respectively. The xQSM method was compared with two recent deep learning (QSMnet+ and DeepQSM) and two conventional dipole inversion (MEDI and iLSQR) methods, using both digital simulations and in vivo experiments. Reconstruction error metrics, including peak signal-to-noise ratio, structural similarity, normalized root mean squared error and deep gray matter susceptibility measurements, were evaluated for comparison of the different methods. The results showed that the proposed xQSM network trained with in vivo datasets achieved the best reconstructions of all the deep learning methods. In particular, it led to, on average, 32.3%, 25.4% and 11.7% improvement in the accuracy of globus pallidus susceptibility estimation for digital simulations and 39.3%, 21.8% and 6.3% improvements for in vivo acquisitions compared with DeepQSM, QSMnet+ and iLSQR, respectively. It also exhibited the highest linearity against different susceptibility intensity scales and demonstrated the most robust generalization capability to various spatial resolutions of all the deep learning methods. In addition, the xQSM method also substantially shortened the reconstruction time from minutes using MEDI to only a few seconds.

Inter-observer variations of the tumor bed delineation for patients after breast conserving surgery in preoperative magnetic resonance and computed tomography scan fusion.

PURPOSE: Tumor bed (TB) delineation based on preoperative magnetic resonance imaging (pre-MRI) fused with postoperative computed tomography (post-CT) were compared to post-CT only to define pre-MRI may aid in improving the accuracy of delineation. METHODS AND MATERIALS: The pre-MRI imaging of 10 patients underwent radiotherapy (RT) after breast conserving surgery (BCS) were reviewed. Post-CT scans were acquired in the same prone position as pre-MRI. Pre-MRI and post-CT automatically match and then manual alignment was given to enhance fusion consistency. Three radiation oncologists and 2 radiologists delineated the clinical target volume (CTV) for CT-based. The gross target volume (GTV) of pre-MRI-based was determined by the volume of tumor acquired with 6 sequences: T1, T2, T2W-SPAIR, DWI, dyn-eTHRIVE and sdyn-eTHRIVE, expended 10 mm to form the CTV-pre-MRI. Planning target volume (PTV) for each sequence was determined by CTV extended 15 mm, trimmed to 3 mm from skin and the breast-chest wall interface. The variability of the TB delineation were developed as follows: the mean volume, conformity index (CI) and dice coefficient (DC). RESULTS: The mean volumes of CTV and PTV delineated with CT were all larger than those with pre-MRI. The lower inter-observer variability was observed from PTV, especially in sdyn-eTHRIVE in all sequences. For each sequence of pre-MRI, all DCs were larger than post-CT, and the largest DC was observed by sdyn-eTHRIVE sequence fusion to post-CT. The overlap for PTV was significantly improved in the pre-MRI-based compared with the CT-based. CONCLUSIONS: TB volumes based on pre-MRI were smaller than post-CT with CVS increased. Pre-MRI provided a more precise definition of the TB with observers performed a smaller inter-observer variability than CT. Pre-MRI, especially in sdyn-eTHRIVE sequence, should help in reducing treatment volumes with the improved accuracy of TB delineation of adjuvant RT of breast cancer.

Amide signal intensities may be reduced in the motor cortex and the corticospinal tract of ALS patients.

OBJECTIVES: The aim of the study is to assess amide concentration changes in ALS patients compared with healthy controls by using quantitative amide proton transfer (APT) and multiparameter magnetic resonance imaging, and testing its correlation with clinical scores. METHODS: Sixteen ALS patients and sixteen healthy controls were recruited as part of the Canadian ALS Neuroimaging Consortium, and multimodal magnetic resonance imaging was performed at 3 T, including APT and diffusion imaging. Lorentz fitting was used to quantify the amide effect. Clinical disability was evaluated using the revised ALS functional rating scale (ALSFRS-R), and its correlation with image characteristics was assessed. The diagnostic performance of different imaging parameters was evaluated with receiver operating characteristic analysis. RESULTS: Our results showed that the amide peak was significantly different between the motor cortex and other gray matter territories within the brain of ALS patients (p < 0.001). Compared with controls, amide signal intensities in ALS were significantly reduced in the motor cortex (p < 0.001) and corticospinal tract (p = 0.046), while abnormalities were not detected using routine imaging methods. There was no significant correlation between amide and ALSFRS-R score. The diagnostic accuracy of the amide peak was superior to that of diffusion imaging. CONCLUSIONS: This study demonstrated changes of amide signal intensities in the motor cortex and corticospinal tract of ALS patients. KEY POINTS: • The neurodegenerative disease amyotrophic lateral sclerosis (ALS) has a lack of objective imaging indicators for diagnosis and assessment. • Analysis of amide proton transfer imaging revealed changes in the motor cortex and corticospinal tract of ALS patients that were not visible on standard magnetic resonance imaging. • The diagnostic accuracy of the amide peak was superior to that of diffusion imaging.

Atypical presentations of coronavirus disease 2019 (COVID-19) from onset to readmission.

BACKGROUND: To investigate the CT imaging and clinical features of three atypical presentations of coronavirus disease 2019 (COVID-19), namely (1) asymptomatic, (2) CT imaging-negative, and (3) re-detectable positive (RP), during all disease stages. METHODS: A consecutive cohort of 79 COVID-19 patients was retrospectively recruited from five independent institutions. For each presentation type, all patients were classified into atypical vs. typical groups (i.e., asymptomatic vs.symptomatic, CT imaging-negative vs. CT imaging-positive, and RP and non-RP,respectively). The chi-square test, Student's t test, and Kruskal-Wallis H test were performed to compare CT imaging and clinical features of atypical vs. typical patients for all three presentation categories. RESULTS: In our COVID-19 cohort, we found 12.7% asymptomatic patients, 13.9% CT imaging-negative patients, and 8.9% RP patients. The asymptomatic patients had fewer hospitalization days (P=0.043), lower total scores for bilateral lung involvement (P< 0.001), and fewer ground-glass opacities (GGOs) in the peripheral area (P< 0.001) than symptomatic patients. The CT imaging-negative patients were younger (P=0.002), had a higher lymphocyte count (P=0.038), had a higher lymphocyte rate (P=0.008), and had more asymptomatic infections (P=0.002) than the CT imaging-positive patients. The RP patients with moderate COVID-19 had lower total scores of for bilateral lung involvement (P=0.030) and a smaller portion of the left lung affected (P=0.024) than non-RP patients. Compared to their first hospitalization, RP patients had a shorter hospitalization period (P< 0.001) and fewer days from the onset of illness to last RNA negative conversion (P< 0.001) at readmission. CONCLUSIONS: Significant CT imaging and clinical feature differences were found between atypical and typical COVID-19 patients for all three atypical presentation categories investigated in this study, which may help provide complementary information for the effective management of COVID-19.

On the value of QSM from MPRAGE for segmenting and quantifying iron-rich deep gray matter.

PURPOSE: Quantitative susceptibility mapping (QSM) has been employed for both iron evaluation and segmentation of deep gray matter (DGM), but QSM sequences are not typically used in standard brain volumetric studies, which use T1-weighted magnetization-prepared rapid acquisition with gradient echo (MPRAGE) with short TE. Here, QSM produced directly from standard MPRAGE phase ( QSMMPRAGE ) is evaluated for segmentation and quantification of highly iron-rich DGM regions. METHODS: Simulations were used to explore quality and possible limitations. In addition, QSM from a standard multi-echo gradient-echo ( QSMGRE ) was compared to QSMMPRAGE in 40 healthy adults at 3T. DGM structures with weak contrast on MPRAGE magnitude were evaluated for improving segmentation with QSMMPRAGE , with focus on the iron-rich globus pallidus (GP). Furthermore, susceptibility quantification was assessed on six DGM nuclei and compared to standard QSMGRE . RESULTS: Limited by TE and signal-to-noise ratio, only iron-rich regions like GP and dentate nucleus produced adequate contrast on QSMMPRAGE , confining applications to such regions. QSMMPRAGE improved GP segmentation with mean Dice scores raised by 9.0%, and mean volumetric differences reduced by 9.7%. Simulations suggested that phase contrast-to-noise ratio (CNR) should be above 3.0 to attain segmentation improvement. For quantification purposes, higher CNR is required, and typical QSMMPRAGE provided comparable estimates to QSMGRE in large iron-rich DGM nuclei. CONCLUSION: Despite the short TE of standard MPRAGE, QSMMPRAGE can improve GP segmentation over the use of MPRAGE magnitude alone and roughly quantify high-iron regions in DGM. Thus, reconstructing QSMMPRAGE can be a useful addition to volumetric studies that rarely include standard QSMGRE .

Quantification of brain oxygen extraction fraction using QSM and a hyperoxic challenge.

PURPOSE: To use hyperoxia in combination with QSM to quantify microvascular oxygen extraction fraction (OEF) and cerebral metabolic rate of oxygen (CMRO2 ) in healthy subjects and to cross-validate results with those from hypercapnia QSM-OEF. METHODS: Ten healthy subjects were scanned on a 3T MRI scanner. At baseline normoxia and during hyperoxia (PetO2 = +300 mmHg), QSM data were acquired using a multi-echo gradient-echo (GRE) sequence, and cerebral blood flow data were acquired using a pseudocontinuous arterial spin labeling sequence. The OEF and CMRO2 maps were computed and compared with those from hypercapnia QSM-OEF, acquired in the same subjects, using correlation and Bland-Altman analysis in 16 vascular territories. RESULTS: Hyperoxia QSM-OEF produced physiologically reasonable OEF and CMRO2 values in all subjects (gray-matter region of interest average OEF = 0.42 ± 0.04, average CMRO2 = 181 ± 34 µmol O2 /min/100 g). When compared with hypercapnia QSM-OEF, Bland-Altman plots revealed small deviations (mean OEF difference = 0.015, mean CMRO2 difference = 4.9 µmol O2 /min/100 g, P < .05). Good and excellent correlations of regional OEF and CMRO2 were found for the two methods. In addition, hyperoxia had minimal impact on cerebral blood flow (average gray-matter cerebral blood flow was reduced by 7.5 ± 5.4%). CONCLUSIONS: Hyperoxia in combination with QSM is a robust approach to measure OEF. Compared with hypercapnia, hyperoxia is more comfortable and has minimal impact on cerebral blood flow.

Cerebral OEF quantification: A comparison study between quantitative susceptibility mapping and dual-gas calibrated BOLD imaging.

PURPOSE: To compare regional oxygen extraction fraction (OEF) and cerebral metabolic rate of oxygen consumption (CMRO2 ) quantified from the microvascular quantitative susceptibility mapping (QSM) using a hypercapnic gas challenge with those measured by the dual-gas calibrated BOLD imaging (DGC-BOLD) in healthy subjects. METHODS: Ten healthy subjects were scanned using a 3T MR system. The QSM data were acquired with a multi-echo gradient-echo sequence at baseline and hypercapnia. Cerebral blood flow data were acquired using the pseudo-continuous arterial spin labeling technique. Baseline OEF and CMRO2 were calculated using QSM and cerebral blood flow measurements. The DGC-BOLD data were also collected under a hypercapnic and a hyperoxic condition to yield baseline OEF and CMRO2 . The QSM-OEF and CMRO2 maps were compared with DGC-BOLD OEF and CMRO2 maps using region of interest (vascular territories) analysis and Bland-Altman plots. RESULTS: Hypercapnia is a robust stimulus for mapping OEF in combination with QSM. Average OEF in 16 vascular territory regions of interest across 10 subjects was 0.40 ± 0.04 by QSM-OEF and 0.38 ± 0.09 by DGC-BOLD. The average CMRO2 was 176 ± 35 and 167 ± 53 µmol O2 /min/100g by QSM-OEF and DGC-BOLD, respectively. A Bland-Altman plot of regional OEF and CMRO2 in regions of interest revealed a statistically significant but small difference (OEF difference = 0.02, CMRO2 difference = 9 µmol O2 /min/100g, p < .05) between the 2 methods for the 10 healthy subjects. CONCLUSION: Hypercapnic challenge-assisted QSM-OEF is a feasible approach to quantify regional brain OEF and CMRO2 . Compared with DGC-BOLD, hypercapnia QSM-OEF results in smaller intersubject variability and requires only 1 gas challenge.

Extracting more for less: multi-echo MP2RAGE for simultaneous T1 -weighted imaging, T1 mapping, R2∗ mapping, SWI, and QSM from a single acquisition.

PURPOSE: To demonstrate simultaneous T1 -weighted imaging, T1 mapping, R2∗ mapping, SWI, and QSM from a single multi-echo (ME) MP2RAGE acquisition. METHODS: A single-echo (SE) MP2RAGE sequence at 7 tesla was extended to ME with 4 bipolar gradient echo readouts. T1 -weighted images and T1 maps calculated from individual echoes were combined using sum of squares and averaged, respectively. ME-combined SWI and associated minimum intensity projection images were generated with TE-adjusted homodyne filters. A QSM reconstruction pipeline was used, including a phase-offsets correction and coil combination method to properly combine the phase images from the 32 receiver channels. Measurements of susceptibility, R2∗ , and T1 of brain tissue from ME-MP2RAGE were compared with those from standard ME-gradient echo and SE-MP2RAGE. RESULTS: The ME combined T1 -weighted, T1 map, SWI, and minimum intensity projection images showed increased SNRs compared to the SE results. The proposed coil combination method led to QSM results free of phase-singularity artifacts, which were present in the standard adaptive combination method. T1 -weighted, T1 , and susceptibility maps from ME-MP2RAGE were comparable to those obtained from SE-MP2RAGE and ME-gradient echo, whereas R2∗ maps showed increased blurring and reduced SNR. T1 , R2∗ , and susceptibility values of brain tissue from ME-MP2RAGE were consistent with those from SE-MP2RAGE and ME-gradient echo. CONCLUSION: High-resolution structural T1 weighted imaging, T1 mapping, R2∗ mapping, SWI, and QSM can be extracted from a single 8.5-min ME-MP2RAGE acquisition using a customized reconstruction pipeline. This method can be applied to replace separate SE-MP2RAGE and ME-gradient echo acquisitions to significantly shorten total scan time.

Editorial for "Deep-Learning Detection of Cancer Metastasis to the Brain on MRI".

LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY: Stage 1 J. Magn. Reson. Imaging 2020;52:1237-1238.