Automatic assessment of disproportionately enlarged subarachnoid-space hydrocephalus from 3D MRI using two deep learning models.

Background: Disproportionately enlarged subarachnoid-space hydrocephalus (DESH) is a key feature for Hakim disease (idiopathic normal pressure hydrocephalus: iNPH), but subjectively evaluated. To develop automatic quantitative assessment of DESH with automatic segmentation using combined deep learning models. Methods: This study included 180 participants (42 Hakim patients, 138 healthy volunteers; 78 males, 102 females). Overall, 159 three-dimensional (3D) T1-weighted and 180 T2-weighted MRIs were included. As a semantic segmentation, 3D MRIs were automatically segmented in the total ventricles, total subarachnoid space (SAS), high-convexity SAS, and Sylvian fissure and basal cistern on the 3D U-Net model. As an image classification, DESH, ventricular dilatation (VD), tightened sulci in the high convexities (THC), and Sylvian fissure dilatation (SFD) were automatically assessed on the multimodal convolutional neural network (CNN) model. For both deep learning models, 110 T1- and 130 T2-weighted MRIs were used for training, 30 T1- and 30 T2-weighted MRIs for internal validation, and the remaining 19 T1- and 20 T2-weighted MRIs for external validation. Dice score was calculated as (overlapping area) × 2/total area. Results: Automatic region extraction from 3D T1- and T2-weighted MRI was accurate for the total ventricles (mean Dice scores: 0.85 and 0.83), Sylvian fissure and basal cistern (0.70 and 0.69), and high-convexity SAS (0.68 and 0.60), respectively. Automatic determination of DESH, VD, THC, and SFD from the segmented regions on the multimodal CNN model was sufficiently reliable; all of the mean softmax probability scores were exceeded by 0.95. All of the areas under the receiver-operating characteristic curves of the DESH, Venthi, and Sylhi indexes calculated by the segmented regions for detecting DESH were exceeded by 0.97. Conclusion: Using 3D U-Net and a multimodal CNN, DESH was automatically detected with automatically segmented regions from 3D MRIs. Our developed diagnostic support tool can improve the precision of Hakim disease (iNPH) diagnosis.

Modeling cerebrospinal fluid dynamics across the entire intracranial space through integration of four-dimensional flow and intravoxel incoherent motion magnetic resonance imaging.

BACKGROUND: Bidirectional reciprocal motion of cerebrospinal fluid (CSF) was quantified using four-dimensional (4D) flow magnetic resonance imaging (MRI) and intravoxel incoherent motion (IVIM) MRI. To estimate various CSF motions in the entire intracranial region, we attempted to integrate the flow parameters calculated using the two MRI sequences. To elucidate how CSF dynamics deteriorate in Hakim's disease, an age-dependent chronic hydrocephalus, flow parameters were estimated from the two MRI sequences to assess CSF motion in the entire intracranial region. METHODS: This study included 127 healthy volunteers aged ≥ 20 years and 44 patients with Hakim's disease. On 4D flow MRI for measuring CSF motion, velocity encoding was set at 5 cm/s. For the IVIM MRI analysis, the diffusion-weighted sequence was set at six b-values (i.e., 0, 50, 100, 250, 500, and 1000 s/mm2), and the biexponential IVIM fitting method was adapted. The relationships between the fraction of incoherent perfusion (f) on IVIM MRI and 4D flow MRI parameters including velocity amplitude (VA), absolute maximum velocity, stroke volume, net flow volume, and reverse flow rate were comprehensively evaluated in seven locations in the ventricles and subarachnoid spaces. Furthermore, we developed a new parameter for fluid oscillation, the Fluid Oscillation Index (FOI), by integrating these two measurements. In addition, we investigated the relationship between the measurements and indices specific to Hakim's disease and the FOIs in the entire intracranial space. RESULTS: The VA on 4D flow MRI was significantly associated with the mean f-values on IVIM MRI. Therefore, we estimated VA that could not be directly measured on 4D flow MRI from the mean f-values on IVIM MRI in the intracranial CSF space, using the following formula; e0.2(f-85) + 0.25. To quantify fluid oscillation using one integrated parameter with weighting, FOI was calculated as VA × 10 + f × 0.02. In addition, the FOIs at the left foramen of Luschka had the strongest correlations with the Evans index (Pearson's correlation coefficient: 0.78). The other indices related with Hakim's disease were significantly associated with the FOIs at the cerebral aqueduct and bilateral foramina of Luschka. FOI at the cerebral aqueduct was also elevated in healthy controls aged ≥ 60 years. CONCLUSIONS: We estimated pulsatile CSF movements in the entire intracranial CSF space in healthy individuals and patients with Hakim's disease using FOI integrating VA from 4D flow MRI and f-values from IVIM MRI. FOI is useful for quantifying the CSF oscillation.

Higher cerebral blood flow on four-dimensional flow magnetic resonance imaging in young women.

We investigated the reduction in regional brain volume and cerebral blood flow (CBF) with aging and explored potential sex differences in healthy brains. Three-dimensional (3D) T1-weighted magnetic resonance imaging (MRI), time-of-flight magnetic resonance angiography, and four-dimensional (4D) flow MRI were performed on 129 healthy volunteers aged 22-92 years. The brains of healthy volunteers were segmented into 21 subregions using 3D T1-weighted MRI and CBFs in 16 major intracranial arteries were measured using 4D flow MRI. The cortical gray matter volume decreased linearly with aging, whereas the cerebral white matter volume increased until the 40s and then decreased, and the subcortical gray matter volume changed little with aging. The cortical gray matter volume was significantly associated with the total CBF of the major intracranial arteries distal to the circle of Willis; however, the cerebral white matter and subcortical gray matter volumes were not. Generally, women have higher total CBF than men, particularly in their 40s and younger, despite the smaller intracranial volume and smaller diameters of intracranial arteries than men. This may contribute to the higher incidence of subarachnoid hemorrhage due to cerebral aneurysms and migraine in women.

Simultaneous boost radiotherapy versus conventional dose radiotherapy for patients with newly diagnosed glioblastoma: a multi-institutional analysis.

We compared survival outcomes of high-dose concomitant boost radiotherapy (HDCBRT) and conventional dose radiotherapy (CRT) for newly diagnosed glioblastoma (GB). Patients treated with intensity-modulated radiation therapy for newly diagnosed GB were included. In HDCBRT, specific targets received 69, 60, and 51 Gy in 30 fractions, while 60 Gy in 30 fractions was administered with a standard radiotherapy method in CRT. Overall survival (OS) and progression-free survival (PFS) were compared using the Log-rank test, followed by multivariate Cox analysis. The inverse probability of treatment weighting (IPTW) method was also applied to each analysis. Among 102 eligible patients, 45 received HDCBRT and 57 received CRT. With a median follow-up of 16 months, the median survival times of OS and PFS were 21 and 9 months, respectively. No significant differences were observed in OS or PFS in the Kaplan-Meier analyses. In the multivariate analysis, HDCBRT correlated with improved OS (hazard ratio, 0.49; 95% confidence interval, 0.27-0.90; P = 0.021), and this result remained consistent after IPTW adjustments (P = 0.028). Conversely, dose suppression due to the proximity of normal tissues and IMRT field correlated with worse OS and PFS (P = 0.008 and 0.049, respectively). A prospective study with a stricter protocol is warranted to validate the efficacy of HDCBRT for GB.