Usefulness of deep learning-based noise reduction for 1.5 T MRI brain images.

AIM: To evaluate 1.5 T magnetic resonance imaging (MRI) brain images with denoising procedures using deep learning-based reconstruction (dDLR) relative to the original 1.5 and 3 T images. MATERIALS AND METHODS: Eleven volunteers underwent MRI at 3 and 1.5 T. Two-dimensional fast spin-echo T2-weighted imaging (T2WI), fluid-attenuated inversion recovery (FLAIR) imaging and diffusion-weighted imaging (DWI) sequences were performed. The dDLR method was applied to the 1.5 T data (dDLR-1.5 T), then the image quality of the dDLR-1.5 T data relative to the original 1.5 T and 3 T data was qualitatively and quantitatively assessed based on the structure similarity (SSIM) index; the signal-to-noise ratios (SNRs) of the grey matter (GM) and white matter (WM); and the contrast-to-noise ratios (CNRs) between the GM and WM (CNRgm-wm) and between the striatum (ST) and WM (CNRst-wm). RESULTS: The perceived image quality, and SNRs and CNRs were significantly higher for the dDLR-1.5 T images versus the 1.5 T images for all sequences and almost comparable or even superior to those of the 3 T images. For DWI, the SNRs and CNRst-wm were significantly higher for the dDLR-1.5 T images versus the 3 T images. CONCLUSION: The dDLR technique improved the image quality of 1.5 T brain MRI images. With respect to qualitative and quantitative measurements, the denoised 1.5 T brain images were almost equivalent or even superior to the 3 T brain images.

Predicting prognosis of resected hepatocellular carcinoma by radiomics analysis with random survival forest.

RATIONALE AND OBJECTIVES: To investigate the impact of random survival forest (RSF) classifier trained by radiomics features over the prediction of the overall survival of patients with resectable hepatocellular carcinoma (HCC). MATERIALS AND METHODS: The dynamic computed tomography data of 127 patients (97 men, 30 women; mean age, 68 years) newly diagnosed with resectable HCC were retrospectively analyzed. After manually setting the region of interest to include the tumor within the slice at its maximum diameter, texture analyses were performed with or without a Laplacian of Gaussian filter. Using the extracted 96 histogram based texture features, RSFs were trained using 5-fold cross-validation to predict the individual risk for each patient on disease free survival (DFS) and overall survival (OS). The associations between individual risk and DFS or OS were evaluated using Kaplan-Meier analysis. The effects of the predicted individual risk and clinical variables upon OS were analyzed using a multivariate Cox proportional hazards model. RESULTS: Among the 96 histogram based texture features, RSF extracted 8 of high importance for DFS and 15 for OS. The RSF trained by these features distinguished two patient groups with high and low predicted individual risk (P=1.1×10-4 for DFS, 4.8×10-7 for OS). Based on the multivariate Cox proportional hazards model, high predicted individual risk (hazard ratio=1.06 per 1% increase, P=8.4×10-8) and vascular invasion (hazard ratio=1.74, P=0.039) were the only unfavorable prognostic factors. CONCLUSIONS: The combination of radiomics analysis and RSF might be useful in predicting the prognosis of patients with resectable HCC.

Imaging Differences between Neuromyelitis Optica Spectrum Disorders and Multiple Sclerosis: A Multi-Institutional Study in Japan.

BACKGROUND AND PURPOSE: Both clinical and imaging criteria must be met to diagnose neuromyelitis optica spectrum disorders and multiple sclerosis. However, neuromyelitis optica spectrum disorders are often misdiagnosed as MS because of an overlap in MR imaging features. The purpose of this study was to confirm imaging differences between neuromyelitis optica spectrum disorders and MS with visually detailed quantitative analyses of large-sample data. MATERIALS AND METHODS: We retrospectively examined 89 consecutive patients with neuromyelitis optica spectrum disorders (median age, 51 years; range, 16-85 years; females, 77; aquaporin 4 immunoglobulin G-positive, 93%) and 89 with MS (median age, 36 years; range, 18-67 years; females, 68; relapsing-remitting MS, 89%; primary-progressive MS, 7%; secondary-progressive MS, 2%) from 9 institutions across Japan (April 2008 to December 2012). Two neuroradiologists visually evaluated the number, location, and size of all lesions using the Mann-Whitney U test or the Fisher exact test. RESULTS: We enrolled 79 patients with neuromyelitis optica spectrum disorders and 87 with MS for brain analysis, 57 with neuromyelitis optica spectrum disorders and 55 with MS for spinal cord analysis, and 42 with neuromyelitis optica spectrum disorders and 14 with MS for optic nerve analysis. We identified 911 brain lesions in neuromyelitis optica spectrum disorders, 1659 brain lesions in MS, 86 spinal cord lesions in neuromyelitis optica spectrum disorders, and 102 spinal cord lesions in MS. The frequencies of periventricular white matter and deep white matter lesions were 17% and 68% in neuromyelitis optica spectrum disorders versus 41% and 42% in MS, respectively (location of brain lesions, P < .001). We found a significant difference in the distribution of spinal cord lesions between these 2 diseases (P = .024): More thoracic lesions than cervical lesions were present in neuromyelitis optica spectrum disorders (cervical versus thoracic, 29% versus 71%), whereas they were equally distributed in MS (46% versus 54%). Furthermore, thoracic lesions were significantly longer than cervical lesions in neuromyelitis optica spectrum disorders (P = .001), but not in MS (P = .80). CONCLUSIONS: Visually detailed quantitative analyses confirmed imaging differences, especially in brain and spinal cord lesions, between neuromyelitis optica spectrum disorders and MS. These observations may have clinical implications.

The relationship of waist circumference and body mass index to grey matter volume in community dwelling adults with mild obesity.

Objective: Previous work has shown that high body mass index (BMI) is associated with low grey matter volume. However, evidence on the relationship between waist circumference (WC) and brain volume is relatively scarce. Moreover, the influence of mild obesity (as indexed by WC and BMI) on brain volume remains unclear. This study explored the relationships between WC and BMI and grey matter volume in a large sample of Japanese adults. Methods: The participants were 792 community-dwelling adults (523 men and 269 women). Brain magnetic resonance images were collected, and the correlation between WC or BMI and global grey matter volume were analysed. The relationships between WC or BMI and regional grey matter volume were also investigated using voxel-based morphometry. Results: Global grey matter volume was not correlated with WC or BMI. Voxel-based morphometry analysis revealed significant negative correlations between both WC and BMI and regional grey matter volume. The areas correlated with each index were more widespread in men than in women. In women, the total area of the regions significantly correlated with WC was slightly greater than that of the regions significantly correlated with BMI. Conclusions: Results show that both WC and BMI were inversely related to regional grey matter volume, even in Japanese adults with somewhat mild obesity. Especially in populations with less obesity, such as the female participants in current study, WC may be more sensitive than BMI as a marker of grey matter volume differences associated with obesity.

Synthetic MRI in the Detection of Multiple Sclerosis Plaques.

BACKGROUND AND PURPOSE: Synthetic MR imaging enables the creation of various contrast-weighted images including double inversion recovery and phase-sensitive inversion recovery from a single MR imaging quantification scan. Here, we assessed whether synthetic MR imaging is suitable for detecting MS plaques. MATERIALS AND METHODS: Quantitative and conventional MR imaging data on 12 patients with MS were retrospectively analyzed. Synthetic T2-weighted, FLAIR, double inversion recovery, and phase-sensitive inversion recovery images were produced after quantification of T1 and T2 values and proton density. Double inversion recovery images were optimized for each patient by adjusting the TI. The number of visible plaques was determined by a radiologist for a set of these 4 types of synthetic MR images and a set of conventional T1-weighted inversion recovery, T2-weighted, and FLAIR images. Conventional 3D double inversion recovery and other available images were used as the criterion standard. The total acquisition time of synthetic MR imaging was 7 minutes 12 seconds and that of conventional MR imaging was 6 minutes 29 seconds The lesion-to-WM contrast and lesion-to-WM contrast-to-noise ratio were calculated and compared between synthetic and conventional double inversion recovery images. RESULTS: The total plaques detected by synthetic and conventional MR images were 157 and 139, respectively (P = .014). The lesion-to-WM contrast and contrast-to-noise ratio on synthetic double inversion recovery images were superior to those on conventional double inversion recovery images (P = .001 and < 0.001, respectively). CONCLUSIONS: Synthetic MR imaging enabled detection of more MS plaques than conventional MR imaging in a comparable acquisition time. The contrast for MS plaques on synthetic double inversion recovery images was better than on conventional double inversion recovery images.

Utility of a Multiparametric Quantitative MRI Model That Assesses Myelin and Edema for Evaluating Plaques, Periplaque White Matter, and Normal-Appearing White Matter in Patients with Multiple Sclerosis: A Feasibility Study.

BACKGROUND AND PURPOSE: T1 and T2 values and proton density can now be quantified on the basis of a single MR acquisition. The myelin and edema in a voxel can also be estimated from these values. The purpose of this study was to evaluate a multiparametric quantitative MR imaging model that assesses myelin and edema for characterizing plaques, periplaque white matter, and normal-appearing white matter in patients with MS. MATERIALS AND METHODS: We examined 3T quantitative MR imaging data from 21 patients with MS. The myelin partial volume, excess parenchymal water partial volume, the inverse of T1 and transverse T2 relaxation times (R1, R2), and proton density were compared among plaques, periplaque white matter, and normal-appearing white matter. RESULTS: All metrics differed significantly across the 3 groups (P < .001). Those in plaques differed most from those in normal-appearing white matter. The percentage changes of the metrics in plaques and periplaque white matter relative to normal-appearing white matter were significantly more different from zero for myelin partial volume (mean, -61.59 ± 20.28% [plaque relative to normal-appearing white matter], and mean, -10.51 ± 11.41% [periplaque white matter relative to normal-appearing white matter]), and excess parenchymal water partial volume (13.82 × 103 ± 49.47 × 103% and 51.33 × 102 ± 155.31 × 102%) than for R1 (-35.23 ± 13.93% and -6.08 ± 8.66%), R2 (-21.06 ± 11.39% and -4.79 ± 6.79%), and proton density (23.37 ± 10.30% and 3.37 ± 4.24%). CONCLUSIONS: Multiparametric quantitative MR imaging captures white matter damage in MS. Myelin partial volume and excess parenchymal water partial volume are more sensitive to the MS disease process than R1, R2, and proton density.

Influence of parameter settings in voxel-based morphometry 8. Using DARTEL and region-of-interest on reproducibility in gray matter volumetry.

OBJECTIVES: To investigate whether reproducibility of gray matter volumetry is influenced by parameter settings for VBM 8 using Diffeomorphic Anatomical Registration Through Exponentiated Lie Algebra (DARTEL) with region-of-interest (ROI) analyses. METHODS: We prepared three-dimensional T1-weighted magnetic resonance images (3D-T1WIs) of 21 healthy subjects. All subjects were imaged with each of five MRI systems. Voxel-based morphometry 8 (VBM 8) and WFU PickAtlas software were used for gray matter volumetry. The bilateral ROI labels used were those provided as default settings with the software: Frontal Lobe, Hippocampus, Occipital Lobe, Orbital Gyrus, Parietal Lobe, Putamen, and Temporal Lobe. All 3D-T1WIs were segmented to gray matter with six parameters of VBM 8, with each parameter having between three and eight selectable levels. Reproducibility was evaluated as the standard deviation (mm³) of measured values for the five MRI systems. RESULTS: Reproducibility was influenced by 'Bias regularization (BiasR)', 'Bias FWHM', and 'De-noising filter' settings, but not by 'MRF weighting', 'Sampling distance', or 'Warping regularization' settings. Reproducibility in BiasR was influenced by ROI. Superior reproducibility was observed in Frontal Lobe with the BiasR1 setting, and in Hippocampus, Parietal Lobe, and Putamen with the BiasR3*, BiasR1, and BiasR5 settings, respectively. CONCLUSION: Reproducibility of gray matter volumetry was influenced by parameter settings in VBM 8 using DARTEL and ROI. In multi-center studies, the use of appropriate settings in VBM 8 with DARTEL results in reduced scanner effect.

Late enhancement of brain-behavior correlations during response inhibition.

Previous neuroimaging studies of response inhibition have examined correlations between behavioral efficiency and brain activity, but the temporal stability of the correlations has largely been ignored. The present functional magnetic resonance imaging (fMRI) study demonstrates the temporal changes of the brain activity associated with performance efficiency that led to more robust brain-behavior correlations in a later part of the experimental sessions. Participants performed a stop-signal task requiring inhibition of inappropriate responses, where more efficient behavioral performance is reflected in a shorter stop-signal reaction time (SSRT). Among across-subject negative correlations between the brain activity and the SSRT, the majority of the negative correlations were observed in the second half of experimental sessions. In the cerebellar region that showed the greatest difference in correlations between the second and the first halves, the brain activity increased in efficient performers, whereas the brain activity decreased in poor performers. These results suggest the existence of multiple brain mechanisms that increase and decrease the brain activity depending on the behavioral efficiency of the performers. More practically, these results indicate that robust brain-behavior correlations can more effectively be detected in a later part of the experimental sessions.

Postoperative transient reduced diffusion in the ipsilateral striatum and thalamus.

BACKGROUND AND PURPOSE: Restriction of diffusion has been reported in the early phase of secondary neuronal degeneration, such as wallerian degeneration. The purpose of this study was to investigate postoperative transient reduced diffusion in the ipsilateral striatum and thalamus as a remote effect of surgery. MATERIALS AND METHODS: Six hundred two postoperative MR imaging examinations in 125 patients after cerebral surgery were retrospectively reviewed, focusing on the presence of reduced diffusion in the striatum and/or thalamus. The distribution of reduced diffusion in the striatum was classified into 3 groups: anterior, central, and posterior. Reduced diffusion in the thalamus was also classified on the basis of the anatomic locations of the thalamic nuclei. Further follow-up MRI was available in all patients with postoperative reduced diffusion, and acute infarctions were excluded. The patient medical records were reviewed to evaluate neurologic status. RESULTS: Restriction of diffusion was observed in the striatum and/or thalamus ipsilateral to the surgical site in 17 patients (13.6%). The distribution of signal abnormality correlated with the location of the operation, in concordance with the architecture of the striatocortical and thalamocortical connections. Reduced diffusion was observed from days 7 to 46 after the operation, especially during days 8-21. The signal abnormalities completely resolved on follow-up examinations. The median follow-up period was 202 days (interquartile range, 76-487 days). CONCLUSIONS: Postoperative transient reduced diffusion in the ipsilateral striatum and/or thalamus likely represents an early phase of secondary neuronal degeneration based on its characteristic distribution and time course. Clinically, this reduced diffusion should not be mistaken for postoperative ischemic injury.

Absence of age-related prefrontal NAA change in adults with autism spectrum disorders.

Atypical trajectory of brain growth in autism spectrum disorders (ASDs) has been recognized as a potential etiology of an atypical course of behavioral development. Numerous neuroimaging studies have focused on childhood to investigate atypical age-related change of brain structure and function, because it is a period of neuron and synapse maturation. Recent studies, however, have shown that the atypical age-related structural change of autistic brain expands beyond childhood and constitutes neural underpinnings for lifelong difficulty to behavioral adaptation. Thus, we examined effects of aging on neurochemical aspects of brain maturation using 3-T proton magnetic resonance spectroscopy ((1)H-MRS) with single voxel in the medial prefrontal cortex (PFC) in 24 adult men with non-medicated high-functioning ASDs and 25 age-, IQ- and parental-socioeconomic-background-matched men with typical development (TD). Multivariate analyses of covariance demonstrated significantly high N-acetylaspartate (NAA) level in the ASD subjects compared with the TD subjects (F=4.83, P=0.033). The low NAA level showed a significant positive correlation with advanced age in the TD group (r=-0.618, P=0.001), but was not evident among the ASD individuals (r=0.258, P=0.223). Fisher's r-to-z transformation showed a significant difference in the correlations between the ASD and TD groups (Z=-3.23, P=0.001), which indicated that the age-NAA relationship was significantly specific to people with TD. The current (1)H-MRS study provided new evidence that atypical age-related change of neurochemical aspects of brain maturation in ASD individuals expands beyond childhood and persists during adulthood.

Diagnostic Ability of Fluid-Attenuated Inversion Recovery MR Imaging to Detect Remnant or Recurrent Meningiomas after Resection.

It has been suggested that a difference in signal intensity (SI) between the resection cavity and normal cerebrospinal fluid (CSF) on fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) in partially resected gliomas indicates subsequent or coincident tumor progression. We considered that this would hold true for resected meningiomas as well. Hence, we aimed to assess whether or not such a difference in the SI during the follow-up evaluation helps predict residual or recurrent tumor in resected meningiomas. We evaluated 63 patients with resected meningiomas. The SI within the resection cavity observed on FLAIR images was qualitatively and quantitatively assessed during follow-up. Qualitative analysis comprised visual comparison of the SI in the resection cavity with that of normal CSF by neuroradiologists. The SI in the resection cavity was quantitatively assessed by region of interest (ROI) analysis and normalized against the background noise and CSF SI. Normalized SI recorded during follow-up was compared with that recorded immediately after resection. Tumor progression was defined as a 20% or greater increase in the diameter of the longest residual or recurrent meningioma (Response Evaluation Criteria in Solid Tumors). The sensitivity and specificity of the elevated SI in the resection cavity for indicating residual or recurrent tumor were calculated. Qualitative analysis by FLAIR MRI showed that patients with remnant tumor following surgery had a prolonged SI increase in the resection cavity. Further, SI increase could not always be observed before recurrence, and both SI increase and regrowth remnant/recurrence could be detected in the same MRI examination. In resected meningiomas, leakage of tumor elements into the resection cavity, presumably tumor cells, manifests as an SI increase on FLAIR images and indicates residual or recurrent tumor. However, unlike the previous reports on partially resected gliomas, we concluded that the SI change does not always precede tumor progression or recurrence.

MR imaging of postischemic neuronal death in the substantia nigra and thalamus following middle cerebral artery occlusion in rats.

The goal of this study was to investigate apparent diffusion coefficient (ADC) and T(2) relaxation time (T(2)) in the substantia nigra and thalamus after middle cerebral artery occlusion in rats. In the substantia nigra ipsilateral to infarct, ADC was significantly lower and T(2) was significantly higher on the third and fourth days, but they did not change significantly on the first, second, eighth and 15th days. In the ipsilateral thalamus, ADC and T(2) did not change significantly between the first and fourth days, but were significantly lower on the eighth and 15th days. This combination of MR findings suggested that secondary degeneration in the thalamus was different from that in the substantia nigra.

Neuro-Behçet's disease: analysis of apparent diffusion coefficients.

Little is known about the diffusion properties of brain lesions in neuro-Behçet's disease (NBD). We looked at 19 NBD lesions (13 active, six chronic) in six patients, using diffusion-weighted MRI (DWI). We calculated the apparent diffusion coefficients (ADC) of the lesions and compared them with normal contralateral brain, expressing this ratio as relative ADC (rADC). The rADC was 1.29+/-0.33 in active and 1.44+/-0.47 in chronic lesions, i.e., significantly higher than in the control regions. Increased diffusivity in both active and chronic phases in NBD is different from the pattern in ischaemic infarcts, so that ADC analysis might be helpful for differentiating between these conditions.

Three-dimensional white matter tractography by diffusion tensor imaging in ischaemic stroke involving the corticospinal tract.

Diffusion tensor MR imaging (DTI) provides information on diffusion anisotropy, which can be expressed with three-dimensional (3D) white matter tractography. We used 3D white matter tractography to show the corticospinal tract in eight patients with acute or early subacute ischaemic stroke involving the posterior limb of the internal capsule or corona radiata and to assess involvement of the tract. Infarcts and the tract were shown simultaneously, providing information on their spatial relationships. In five of the eight patients, 3D fibre tract maps showed the corticospinal tract in close proximity to the infarct but not to pass through it. All these patients recovered well, with maximum improvement from the lowest score on manual muscle testing (MMT) up to the full score through rehabilitation. In the other three patients the corticospinal tract was shown running through the infarct; reduction in MMT did not necessarily improve favourably or last longer, other than in one patient. As 3D white matter tractography can show spatial relationships between the corticospinal tract and an infarct, it might be helpful in prognosis of gross motor function.

Hemorrhagic brain metastases with high signal intensity on diffusion-weighted MR images. A case report.

Diffusion-weighted MR imaging has been applicable to the differential diagnosis of abscesses and necrotic or cystic brain tumors. However, restricted water diffusion is not necessarily specific for brain abscess. We describe ring-enhancing metastases of lung carcinoma characterized by high signal intensity on diffusion-weighted MR images. The signal pattern probably reflected intralesional hemorrhage. The present report adds to the growing literature regarding the differential diagnosis of ring-enhancing brain lesions.

Diffusion property in a hamartomatous lesion of neurofibromatosis type 1.

Though diffusion-weighted MRI has been applied to various intracranial lesions, few reports had been presented about cerebral hamartomatous lesions in patients with neurofibromatosis type 1 (NF1). In this study, we report the interval changes of apparent diffusion coefficient (ADC) in a presumed hamartomatous lesion. In our case, the ADC increased slightly over a 3 year period. This diffusion property may provide specific insight into the etiology of cerebral hamartomatous lesions observed in NF1.