Fluid-attenuated Inversion Recovery Research Articles

Automated segmentation of MRI white matter hyperintensities in 8,421 patients with acute ischemic stroke.

To date, only a few small studies have attempted deep learning-based automatic segmentation of white matter hyperintensity (WMH) lesions in patients with cerebral infarction, which is complicated because stroke-related lesions can obscure WMH borders. We developed and validated deep learning algorithms to segment WMH lesions accurately in patients with cerebral infarction, using multisite datasets involving 8,421 patients with acute ischemic stroke. We included 8,421 stroke patients from 9 centers in Korea. 2D UNet and SE-Unet models were trained using 2,408 FLAIR MRI from 3 hospitals and validated using 6,013 FLAIR MRIs from 6 hospitals. WMH segmentation performance was assessed by calculating DSC, correlation coefficient, and concordance correlation coefficient compared to a human-segmented gold standard. In addition, we obtained an uncertainty index that represents overall ambiguity in the voxel classification for WMH segmentation in each patient based on the Kullback-Leibler divergence. In the training dataset, the mean age was 67.4±13.0 years and 60.4% were men. The mean (95% CI) DSCs for Unet in internal testing and external validation were respectively 0.659 (0.649-0.669) and 0.710 (0.707-0.714), which were slightly lower than the reliability between humans (DSC=0.744; 95% CI=0.738-0.751; P=.031). Compared with the Unet, the SE-Unet demonstrated better performance, achieving a mean DSC of 0.675 (0.666-0.685; P<.001) in the internal testing and 0.722 (0.719-0.726; P<.001) in the external validation; moreover, it achieved high DSC values (ranging from 0.672 to 0.744) across multiple validation datasets. We observed a significant correlation between WMH volumes that were segmented automatically and manually for the Unet (r=0.917, P<.0001) and even stronger for the SE-Unet (r=0.933, P<.0001). The SE-Unet also attained a high concordance correlation coefficient (ranging from 0.841 to 0.956) in external test datasets. In addition, the uncertainty indices in the majority of patients (86%) in the external datasets were below 0.35, with an average DSC of 0.744 in these patients. We developed and validated deep learning algorithms to segment WMH in patients with acute cerebral infarction using the largest-ever MRI datasets. In addition, we showed that the uncertainty index can be used to identify cases where automatic WMH segmentation is less accurate and requires human review. WMH = white matter hyperintensity; CNN = convolutional neural networks; SE = squeeze-and-excitation; KL = Kullback-Leibler; ReLU = rectified linear unit; LKW = last known well; mRS = modified Rankin Scale; NIHSS = National Institute of Health Stroke Scale; LAA = large artery atherosclerosis; SVO = small vessel occlusion; CE = cardioembolism.

Central vein sign and trigeminal lesions of multiple sclerosis visualised by 7T MRI

BackgroundAlthough trigeminal nerve involvement is a characteristic of multiple sclerosis (MS), its prevalence across studies varies greatly due to MRI resolution and cohort selection bias. The mechanism behind the site...

Efficient segmentation of active and inactive plaques in FLAIR-images using DeepLabV3Plus SE with efficientnetb0 backbone in multiple sclerosis

This research paper introduces an efficient approach for the segmentation of active and inactive plaques within Fluid-attenuated inversion recovery (FLAIR) images, employing a convolutional neural network (CNN) model known as DeepLabV3Plus SE with the EfficientNetB0 backbone in Multiple sclerosis (MS), and demonstrates its superior performance compared to other CNN architectures. The study encompasses various critical components, including dataset pre-processing techniques, the utilization of the Squeeze and Excitation Network (SE-Block), and the atrous spatial separable pyramid Block to enhance segmentation capabilities. Detailed descriptions of pre-processing procedures, such as removing the cranial bone segment, image resizing, and normalization, are provided. This study analyzed a cross-sectional cohort of 100 MS patients with active brain plaques, examining 5000 MRI slices. After filtering, 1500 slices were utilized for labeling and deep learning. The training process adopts the dice coefficient as the loss function and utilizes Adam optimization. The study evaluated the model's performance using multiple metrics, including intersection over union (IOU), Dice Score, Precision, Recall, and F1-Score, and offers a comparative analysis with other CNN architectures. Results demonstrate the superior segmentation ability of the proposed model, as evidenced by an IOU of 69.87, Dice Score of 76.24, Precision of 88.89, Recall of 73.52, and F1-Score of 80.47 for the DeepLabV3+SE_EfficientNetB0 model. This research contributes to the advancement of plaque segmentation in FLAIR images and offers a compelling approach with substantial potential for medical image analysis and diagnosis.

Visualization of Cerebrospinal Fluid Outflow and Egress along the Nerve Roots of the Lumbar Spine.

Intrinsic cerebrospinal fluid (CSF) dynamics in the brain have been extensively studied, particularly the egress sites of tagged intrinsic CSF in the meninges. Although spinal CSF recirculates within the central nervous system (CNS), we hypothesized that CSF outflows from the lumbar spinal canal. We aimed to visualize and semi-quantify the outflow using non-contrast MRI techniques. We utilized a 3 Tesla clinical MRI with a 16-channel spine coil, employing time-spatial labeling inversion (Time-SLIP) with tag-on and tag-off acquisitions, T2-weighted coronal 2D fluid-attenuated inversion recovery (FLAIR) and T2-weighted coronal 3D centric ky-kz single-shot FSE (cSSFSE). Images were acquired using time-spatial labeling inversion pulse (Time-SLIP) with tag-on and tag-off acquisitions with varying TI periods. Ten healthy volunteers with no known spinal diseases participated. Variations in tagged CSF outflow were observed across different thoracolumbar nerve root segments in all participants. We quantified CSF outflow at all lumbar levels and the psoas region. There was no significant difference among the ROIs for signal intensity. The tagged CSF outflow from the spinal canal is small but demonstrates egress to surrounding tissues. This finding may pave the way for exploring intrathecal drug delivery, understanding of CSF-related pathologies and its potential as a biomarker for peripheral neuropathy and radiculopathy.

An unenhanced 3D-FLAIR sequence using long repetition time and constant flip angle to image endolymphatic hydrops.

To evaluate a three-dimensional fluid-attenuated inversion recovery (3D-FLAIR) sequence using a long repetition time (TR) and constant flip angle (CFA) in differentiating between perilymph and endolymph in a phantom study, and unenhanced endolymphatic hydrops (EH) imaging in a patient study. Three solutions in similar ion and protein concentrations with endolymph, perilymph, and cerebrospinal fluid were prepared for variable flip angle (VFA) 3D-FLAIR (TR 10,000 ms) and CFA (120°) 3D-FLAIR using different TR (10,000, 16,000, and 20,000 ms). Fifty-two patients with probable or definite Meniere's disease received unenhanced CFA (120°) 3D-FLAIR using a long TR (20,000 ms) and 4-h-delay enhanced CFA (120°) 3D-FLAIR (TR 16,000 ms). Image quality, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) of them were compared. Agreement in the evaluation of the EH degree between them was analyzed. In the phantom study, CNRs between perilymphatic and endolymphatic samples of VFA 3D-FLAIR (TR 10,000 ms) and CFA 3D-FLAIR (TR 10,000, 16,000, and 20,000 ms) were 6.66 ± 1.30, 17.90 ± 2.76, 23.87 ± 3.09, and 28.22 ± 3.15 (p < 0.001). In patient study, average score (3.65 ± 0.48 vs. 4.19 ± 0.40), SNR (34.56 ± 9.80 vs. 51.40 ± 11.27), and CNR (30.66 ± 10.55 vs. 45.08 ± 12.27) of unenhanced 3D-FLAIR were lower than enhanced 3D-FLAIR (p < 0.001). Evaluations of the two sequences showed excellent agreement in the cochlear and vestibule (Kappa value: 0.898 and 0.909). The CFA 3D-FLAIR sequence using a long TR could be used in unenhanced EH imaging with high accuracy. Unenhanced imaging of endolymphatic hydrops is valuable in the diagnosis and follow-up of patients, especially those who cannot receive contrast-enhanced MRI. Ion and protein concentration differences can be utilized in differentiating endolymph and perilymph on MRI. Endolymphatic and perilymphatic samples could be differentiated in vitro on this 3D-FLAIR sequence. This unenhanced 3D-FLAIR sequence is in excellent agreement with the enhanced constant flip angle 3D-FLAIR sequence.

Periventricular Leukomalacia: Comparison of Parenchymal Signal and Volume Changes on Brain MRI in Paediatric Cases with Healthy Peers

Background/aims: Previous MRI studies have shown white matter (WM) and gray matter (GM) of cerebrum and cerebellum, corpus callosum (CC) abnormalities in periventricular leukomalacia (PVL). But, the WM FLAIR signal ratio in MRI may provide quantitative data in the diagnosis and follow-up, as a new radiologic method. Thalamic involvement may be a biomarker for neuronal damage and disease severity. We aimed to re-investigate both WM and GM volume changes of cerebrum and cerebellum, CC surface area in PVL, and to evaluate the diagnostic accuracy of the thalamus L sign and FLAIR signal ratio. Methods: MRI scans of 30 pediatric patients with PVL and 42 healthy controls were analyzed to examine WM and GM volume changes, FLAIR signal ratio, CC surface area, and thalamus L sign. Volumetric analyses were done with the Volbrain program. Results: Decreased subcortical GM volumes were found in PVL (p0.001). Conclusions: Our results emphasise the role of MRI in the detection of PVL, the evaluation of GM changes and brain damage, and the importance of thalamus L sign and FLAIR signal ratio in the evaluation of the severity of the disease. Comprehensive studies in this direction may contribute to the development of targeted treatment strategies aimed at reducing cognitive and motor impairments in PVL.

MR imaging of endolymphatic hydrops in Ménière’s disease: feasibility at 1.5 T

BackgroundMénière’s disease is a chronic condition of the inner ear that causes vertigo, tinnitus and hearing loss. Its diagnosis relies on clinical criteria that are subjective and pure-tone audiometry results that are not specific. Its pathological substrate is endolymphatic hydrops. Its imaging was recently made possible by the late-enhanced 3D FLAIR sequence. This technique was primarily tested on 3 T. Our objective was to prove its feasibility using a 1.5 T magnet.MethodsWe conducted a prospective study including 30 patients who fulfilled the Bárány society criteria for Ménière’s disease. We performed the late-enhanced 3D FLAIR sequence on all patients. We used it to look for and grade endolymphatic hydrops in the utricle, the saccule and the cochlear canal using the Kahn method.ResultsWe found endolymphatic hydrops in all of the 30 patients who fulfilled the diagnostic criteria for Ménière’s disease. We had no false positives and only one false negative with a patient presenting with bilateral disease clinically but having endolymphatic hydrops only on one side. Thus, our correspondence rate between clinical and imaging findings was 97%.ConclusionsIt is possible to diagnose endolymphatic hydrops with the late-enhanced 3D FLAIR sequence using a 1.5 T MRI machine. Since Ménière’s disease diagnosis is sometimes tricky, imaging endolymphatic hydrops can aid in the diagnosis when the clinical picture is incomplete. It also helps guide invasive treatment plans. Feasibility at 1.5 T ensures broader access to the late-enhanced 3D FLAIR sequence. Beyond the scope of Ménière’s disease, this sequence offers the possibility to better understand pressure-related inner ear diseases.

Unveiling Opportunities for Intervention: A Prospective Cohort Study Investigating the Clinical Significance of Diffusion-Weighted Imaging (DWI)-Fluid-Attenuated Inversion Recovery (FLAIR) Mismatch Beyond the Window Period in Acute Ischemic Stroke.

Acute ischemic stroke causes irreversible damage to the brain parenchymasurrounded by salvageable tissue known as the ischemic penumbra.Magnetic resonance imaging (MRI), particularly the mismatch between abnormal diffusion-weighted imaging (DWI) signals and normal fluid-attenuated inversion recovery (FLAIR) signals, plays a critical role in detecting ischemic penumbra. It also allows for the identification of patients who may benefit from reperfusion therapy. Hence, this prospective cohort study aimed to explore the correlation between DWI-FLAIR mismatch and clinical outcomes in acute ischemic stroke patients, specifically those with delayed or uncertain symptom onset, offering potential insights into reperfusion therapy. A total of 38 thrombotic stroke patients aged above 18 were included in this prospective cohort study. Baseline data, including demographics, lifestyle factors, and medical history, were recorded. DWI-FLAIR mismatch was evaluated through brain MRI within 4.5 hours to 12 hours of symptom onset. Of the cohort, 63.2% were males, predominantly in the 61-70 age group. Smoking and alcohol consumption were reported by 15.79% each. DWI-FLAIR mismatch was present in 20 out of 38 subjects. No statistically significant differences were noted in the mean National Institutes of Health Stroke Scale (NIHSS) and Modified Rankin Scale (MRS) scores between subjects with and without DWI-FLAIR mismatch. Thrombolysis in wake-up stroke subjects demonstrated a substantial reduction in mean MRS at discharge (1.29±0.95) and at six to eight weeks (1.71±1.11), suggesting potential benefits on functional outcomes. The prevalence of DWI-FLAIR mismatch was seen in the majority of patients beyond their window period and also showed beneficiary outcomes with a mean reduction in NHISS and MRS scores following thrombolysis.

MRI-Based Inner Ear Assessment and Cochlin Tomoprotein-Based Evaluation of Perilymphatic Fistula in Patients with Sudden Hearing Loss.

To study the correlation between positive cochlin tomoprotein testing (CTP), magnetic resonance (MR) imaging, and the auditory and vestibular function amongst patients with sudden hearing loss. Prospective case series. We prospectively examined eight patients who presented with sudden hearing loss (>60 dB) with or without vertigo or tinnitus. We performed an ELISA-based CTP detection test using middle ear lavage samples. In addition to the CTP examination, a magnetic resonance imaging (MRI) examination was performed using different sequences (T1 and a T1 sequence with a contrast medium (CM), a T2 sequence, 4 h delayed intravenous gadolinium-enhanced three-dimensional fluid-attenuated inversion recovery (3D FLAIR)). All patients with sudden hearing loss (>60 dB) presented a non-specific contrast enhancement in the cochlea and vestibulum on the affected side on delayed 3D-FLAIR MRI. Four patients had a positive CTP test, suggesting a perilymphatic fistula (PLF). However, no specific MRI signal for a PLF was observed. Using multimodal diagnostic measures, such as CTP testing and different MRI sequences, no correlation could be found in patients with a PLF.

MRI-based radiomics signatures for preoperative prediction of Ki-67 index in primary central nervous system lymphoma

PurposeThe aim of this study was to develop and validate radiomics signatures based on MRI for preoperative prediction of Ki-67 proliferative index (PI) expression in primary central nervous system lymphoma (PCNSL). MethodsA total of 341 patients with PCNSL were retrospectively analyzed, including 286 patients in one center as the training set and 55 patients in another two centers as the external validation set. Radiomics features were extracted and selected from preoperative contrast-enhanced T1-weighted images, fluid attenuation inversion recovery to build radiomics signatures according to the Ki-67 PI. The predictive performances of the radiomics model were evaluated using four classifiers including random forest, K-Nearest Neighbors, Neural Network and Decision Tree. A combined model was built by incorporating radiomics signature, clinical variables and MRI radiological characteristics using multivariate logistic regression analysis, and a nomogram was established to predict the expression of Ki-67 individually. The predictive performances of the models were evaluated using area under receiver operating characteristic curve (AUC) and decision curve analysis (DCA). ResultsRadiomics signatures were independent predictors of the expression level of Ki-67 (OR: 2.523, P < 0.001). RF radiomics models had the highest accuracy (0.934 in the training set and 0.811 in the external validation set) and F1 Score (0.920 in the training set and 0.836 in the external validation set). The clinic-radiologic-radiomics nomogram showed better predictive performance with AUCs of 0.877(95 % CI: 0.837–0.918) in the training set and 0.866(95 % CI: 0.774–0.957) in the external validation set. The calibration curve and DCA demonstrated goodness-of-fit and improved benefits in clinical practice of the nomogram. ConclusionsNomograms integrating MRI-based radiomics and clinical-radiological characteristics could effectively predict Ki-67 PI in primary PCNSL.

Investigation of the relationship between thrombophilic disorders and brain white matter lesions in migraine with aura.

Migraine is associated with several genetic or acquired comorbidities. Studies conducted in recent years emphasize that the frequency of thrombophilia is high in migraine, especially migraine with aura (MA). Similarly, the presence of white matter lesions (WMLs) on brain magnetic resonance imaging (MRI) scans has been associated with migraine for many years. Based on the knowledge that both WMLs and thrombophilia variants are frequently observed in MA, we aimed to investigate whether there is a relationship between genetic thrombophilia and the presence of WMLs in these patients. The levels of proteins S and C, antithrombin III activities, activated protein C (APC) resistance, antiphospholipid immunoglobulin G/immunoglobulin M (IgG/IgM) and anticardiolipin IgG/IgM antibodies were investigated in 66 MA patients between the ages of 18 and 49 years who presented no cardiovascular risk factors. The presence of WMLs and the Fazekas grade was determined from the brain magnetic resonance imaging (MRI) scans' T2-weighted and fluid-attenuated inversion recovery (FLAIR) sequence taken from the patients. The rates of WMLs were compared in patients with and without thrombophilia. Thrombophilia was detected in 34.8% of the patients, and 27.3% were determined to have WMLs in brain MRI scans. The WMLs were detected in 23.3% of the patients without thrombophilia, in 34.8% of those with thrombophilia, and in 50% of the subjects with multiple thrombophilia disorders. Among the thrombophilia disorders, only APC resistance was significantly more common in patients with WMLs. The results of the present study showed that thrombophilia may be a mechanism that should be investigated in the etiology of increased WMLs in MA.

Iterative Motion Correction Technique with Deep Learning Reconstruction for Brain MRI: A Volunteer and Patient Study.

The aim of this study was to investigate the effect of iterative motion correction (IMC) on reducing artifacts in brain magnetic resonance imaging (MRI) with deep learning reconstruction (DLR). The study included 10 volunteers (between September 2023 and December 2023) and 30 patients (between June 2022 and July 2022) for quantitative and qualitative analyses, respectively. Volunteers were instructed to remain still during the first MRI with fluid-attenuated inversion recovery sequence (FLAIR) and to move during the second scan. IMCoff DLR images were reconstructed from the raw data of the former acquisition; IMCon and IMCoff DLR images were reconstructed from the latter acquisition. After registration of the motion images, the structural similarity index measure (SSIM) was calculated using motionless images as reference. For qualitative analyses, IMCon and IMCoff FLAIR DLR images of the patients were reconstructed and evaluated by three blinded readers in terms of motion artifacts, noise, and overall quality. SSIM for IMCon images was 0.952, higher than that for IMCoff images (0.949) (p < 0.001). In qualitative analyses, although noise in IMCon images was rated as increased by two of the three readers (both p < 0.001), all readers agreed that motion artifacts and overall quality were significantly better in IMCon images than in IMCoff images (all p < 0.001). In conclusion, IMC reduced motion artifacts in brain FLAIR DLR images while maintaining similarity to motionless images.

Central Vein Sign and Paramagnetic Rim Lesions: Susceptibility Changes in Brain Tissues and Their Implications for the Study of Multiple Sclerosis Pathology.

Multiple sclerosis (MS) is the most common acquired inflammatory and demyelinating disease in adults. The conventional diagnostic of MS and the follow-up of inflammatory activity is based on the detection of hyperintense foci in T2 and fluid-attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) and lesions with brain-blood barrier (BBB) disruption in the central nervous system (CNS) parenchyma. However, T2/FLAIR hyperintense lesions are not specific to MS and the MS pathology and inflammatory processes go far beyond focal lesions and can be independent of BBB disruption. MRI techniques based on the magnetic susceptibility properties of the tissue, such as T2*, susceptibility-weighted images (SWI), and quantitative susceptibility mapping (QSM) offer tools for advanced MS diagnostic, follow-up, and the assessment of more detailed features of MS dynamic pathology. Susceptibility-weighted techniques are sensitive to the paramagnetic components of biological tissues, such as deoxyhemoglobin. This capability enables the visualization of brain parenchymal veins. Consequently, it presents an opportunity to identify veins within the core of multiple sclerosis (MS) lesions, thereby affirming their venocentric characteristics. This advancement significantly enhances the accuracy of the differential diagnostic process. Another important paramagnetic component in biological tissues is iron. In MS, the dynamic trafficking of iron between different cells, such as oligodendrocytes, astrocytes, and microglia, enables the study of different stages of demyelination and remyelination. Furthermore, the accumulation of iron in activated microglia serves as an indicator of latent inflammatory activity in chronic MS lesions, termed paramagnetic rim lesions (PRLs). PRLs have been correlated with disease progression and degenerative processes, underscoring their significance in MS pathology. This review will elucidate the underlying physical principles of magnetic susceptibility and their implications for the formation and interpretation of T2*, SWI, and QSM sequences. Additionally, it will explore their applications in multiple sclerosis (MS), particularly in detecting the central vein sign (CVS) and PRLs, and assessing iron metabolism. Furthermore, the review will discuss their role in advancing early and precise MS diagnosis and prognostic evaluation, as well as their utility in studying chronic active inflammation and degenerative processes.

Diagnostic Accuracy of Posterior/Anterior Periventricular WMH Ratio to Differentiate CAA From Hypertensive Arteriopathy.

Periventricular white matter hyperintensities (PVWMHs) in cerebral amyloid angiopathy (CAA) have been reported posterior predominant using semiautomated segmentation method and logarithmic transformation. We aimed to compare PVWMH extent and posterior/anterior distribution between patients with CAA and patients with hypertensive arteriopathy with radiological tools available in daily practice. We retrospectively analyzed confluent PVWMH directly adjacent to lateral ventricles on axial FLAIR (fluid-attenuated inversion recovery) from 108 patients with CAA and 99 patients with hypertensive arteriopathy presenting with hemorrhage-related symptoms consecutively recruited in our stroke database (Nîmes University Hospital, France) between January 2015 and March 2022. For each of the left (L), right (R), anterior (A), and posterior (P) horns of lateral ventricles, the maximal distance between the outer PVWMH border and ventricle border was measured. The sum of anterior left PVWMH and anterior right PVWMH, and posterior left PVWMH and posterior right PVWMH resulted in anterior and posterior extent, respectively. Compared with hypertensive arteriopathy, patients with CAA were older (median, 77 versus 71; P=0.0010) and less frequently male (46% versus 64%; P=0.012) and had less frequent hypertension (45% versus 63%; P=0.013) and more chronic hemorrhages (P<0.0001). CAA showed slightly more extensive anterior right PVWMH (median, 6.50 versus 5.90 mm; P=0.034), far more extensive (all P<0.0001) posterior left PVWMH (median, 13.95 versus 6.95 mm), posterior right PVWMH (median, 14.15 versus 5.45 mm), posterior (median, 27.95 versus 13.00 mm), and total (median, 39.60 versus 24.65 mm) PVWMH, and higher posterior/anterior ratios (median, 1.82 versus 1.01). Age-/sex-adjusted model predicting CAA incorporating total PVWMH extent and posterior/anterior ratios for the given score (-4.3683+0.0268×PVWMH-T+0.3749×posterior/anterior PVWMH ratio+0.0394×age+0.3046 when female) showed highest area under the curve (0.76 [0.70-0.83]), with a 72% [62.50-80.99] sensitivity and 76% [67.18-84.12] specificity. Values above the optimal threshold of 0.22 for the score showed a crude relative risk of 2.75 (2.26-2.37; P<0.0001). Severe posterior PVWMH and high posterior/anterior PVWMH ratio assessed by radiological tools used in daily practice are associated with probable CAA versus hypertensive arteriopathy. URL: https://www.clinicaltrials.gov; Unique identifier: NCT05486897.

The prognostic importance of traumatic axonal injury on early MRI: the Trondheim TAI-MRI grading and quantitative models

ObjectivesWe analysed magnetic resonance imaging (MRI) findings after traumatic brain injury (TBI) aiming to improve the grading of traumatic axonal injury (TAI) to better reflect the outcome.MethodsFour-hundred sixty-three patients (8–70 years) with mild (n = 158), moderate (n = 129), or severe (n = 176) TBI and early MRI were prospectively included. TAI presence, numbers, and volumes at predefined locations were registered on fluid-attenuated inversion recovery (FLAIR) and diffusion-weighted imaging, and presence and numbers on T2*GRE/SWI. Presence and volumes of contusions were registered on FLAIR. We assessed the outcome with the Glasgow Outcome Scale Extended. Multivariable logistic and elastic-net regression analyses were performed.ResultsThe presence of TAI differed between mild (6%), moderate (70%), and severe TBI (95%). In severe TBI, bilateral TAI in mesencephalon or thalami and bilateral TAI in pons predicted worse outcomes and were defined as the worst grades (4 and 5, respectively) in the Trondheim TAI-MRI grading. The Trondheim TAI-MRI grading performed better than the standard TAI grading in severe TBI (pseudo-R2 0.19 vs. 0.16). In moderate-severe TBI, quantitative models including both FLAIR volume of TAI and contusions performed best (pseudo-R2 0.19–0.21). In patients with mild TBI or Glasgow Coma Scale (GCS) score 13, models with the volume of contusions performed best (pseudo-R2 0.25–0.26).ConclusionsWe propose the Trondheim TAI-MRI grading (grades 1–5) with bilateral TAI in mesencephalon or thalami, and bilateral TAI in pons as the worst grades. The predictive value was highest for the quantitative models including FLAIR volume of TAI and contusions (GCS score <13) or FLAIR volume of contusions (GCS score ≥ 13), which emphasise artificial intelligence as a potentially important future tool.Clinical relevance statementThe Trondheim TAI-MRI grading reflects patient outcomes better in severe TBI than today’s standard TAI grading and can be implemented after external validation. The prognostic importance of volumetric models is promising for future use of artificial intelligence technologies.Key PointsTraumatic axonal injury (TAI) is an important injury type in all TBI severities. Studies demonstrating which MRI findings that can serve as future biomarkers are highly warranted.This study proposes the most optimal MRI models for predicting patient outcome at 6 months after TBI; one updated pragmatic model and a volumetric model.The Trondheim TAI-MRI grading, in severe TBI, reflects patient outcome better than today’s standard grading of TAI and the prognostic importance of volumetric models in all severities of TBI is promising for future use of AI.

Association of audiovestibular findings with fluid-attenuated inversion recovery magnetic resonance imaging in patients with vestibular schwannoma

BackgroundVestibular schwannoma (VS) is a benign tumor of the vestibular nerve. Flair-attenuated inversion recovery (FLAIR) of magnetic resonance imaging (MRI) images are sensitive in detecting high protein contents of fluids. ObjectivesTo investigate the association between signal intensity (SI) on FLAIR images and audiovestibular findings in patients with VS. MethodsMedical records of twenty-five patients with VS were retrospectively analyzed. ResultsLarger tumors were associated with increased FLAIR SI of the cochlea, vestibule, and semicircular canal (SCC) on the affected side compared to those of the unaffected side. Pure-tone audiometry (PTA), and speech audiometry were associated with the SI of the affected cochlea. There was no significant correlation between the SI of the vestibule and vestibular evoked myogenic potential, SI of the SCC, and caloric test or video head impulse test results. ConclusionOur study suggests that tumor size was significantly associated with high SI on FLAIR imaging, and audiological findings were associated with the SI of the affected cochlea. Further studies with larger cohorts are required to confirm the association between vestibular function and FLAIR imaging in VS.

Evaluation of T2W FLAIR MR image quality using artificial intelligence image reconstruction techniques in the pediatric brain

BackgroundArtificial intelligence (AI) reconstruction techniques have the potential to improve image quality and decrease imaging time. However, these techniques must be assessed for safe and effective use in clinical practice.ObjectiveTo assess image quality and diagnostic confidence of AI reconstruction in the pediatric brain on fluid-attenuated inversion recovery (FLAIR) imaging.Materials and methodsThis prospective, institutional review board (IRB)-approved study enrolled 50 pediatric patients (median age=12 years, Q1=10 years, Q3=14 years) undergoing clinical brain MRI. T2-weighted (T2W) FLAIR images were reconstructed by both standard clinical and AI reconstruction algorithms (strong denoising). Images were independently rated by two neuroradiologists on a dedicated research picture archiving and communication system (PACS) to indicate whether AI increased, decreased, or had no effect on image quality compared to standard reconstruction. Quantitative analysis of signal intensities was also performed to calculate apparent signal to noise (aSNR) and apparent contrast to noise (aCNR) ratios.ResultsAI reconstruction was better than standard in 99% (reader 1, 49/50; reader 2, 50/50) for overall image quality, 99% (reader 1, 49/50; reader 2, 50/50) for subjective SNR, and 98% (reader 1, 49/50; reader 2, 49/50) for diagnostic preference. Quantitative analysis revealed significantly higher gray matter aSNR (30.6±6.5), white matter aSNR (21.4±5.6), and gray-white matter aCNR (7.1±1.6) in AI-reconstructed images compared to standard reconstruction (18±2.7, 14.2±2.8, 4.4±0.8, p<0.001) respectively.ConclusionWe conclude that AI reconstruction improved T2W FLAIR image quality in most patients when compared with standard reconstruction in pediatric patients.

DIPG-09. PROGNOSTIC FACTORS IN REIRRADIATED DIFFUSE INTRINSIC PONTINE GLIOMA: CLINICAL AND RADIOMIC ANALYSIS

Abstract BACKGROUND Diffuse Intrinsic Pontine Glioma (DIPG) is a lethal pediatric brain tumor. Radiation therapy (RT) is the standard treatment, with reirradiation considered at progression. However, the prognostic factors for reirradiation remain unclear. This study aims to investigate the outcomes of patients with DIPG undergoing reirradiation and to identify radiomic prognostic factors. METHODS We analyzed patients with histopathologically proven DIPG who underwent reirradiation at our institution. We delineated pre-reirradiation tumors using 3D Slicer software and extracted radiomic features from FLAIR MRI sequences using the Python package PyRadiomics. Finally, we used the least absolute shrinkage and selection operator (Lasso) for Cox’s proportional hazard model with leave-one-out cross-validation to select optimal prognostic factors for survival after reirradiation. We have also compared the reirradiated group with our cohort of 22 non-reirradiated DIPG patients eligible for reirradiation but whose parents chose only supportive care at progression. RESULTS The study included 15 patients who received reirradiation at the first progression, with a total dose of 20 Gy or 24 Gy in 2 Gy fractions. Reirradiation was well tolerated, without grade &amp;gt;2 toxicity. Neurological improvement was observed in 80% of patients. Median survival from the end of the first RT was significantly (p = 0.01) longer in the reirradiated group (68 weeks) compared to the non-reirradiated group (25 weeks). The Cox model demonstrated a concordance of 0.88 (95% CI: 0.83–0.92), revealing that tumor sphericity and structural gray-level heterogeneity in FLAIR MRI images were associated with more prolonged survival in reirradiated patients. CONCLUSIONS Reirradiation is a safe and effective approach for patients with DIPG. MRI-based radiomic models could help predict survival after treatment. Our results revealed that tumors with low sphericity and high homogeneity of gray levels are associated with a worse prognosis after reirradiation.

IMG-27. ASSESSMENT OF VARIABILITY OF PEDIATRIC BRAIN TUMOR IMAGING PROTOCOLS ACROSS INSTITUTIONS: ARE WE FOLLOWING RECOMMENDATIONS?

Abstract BACKGROUND Evaluating pediatric brain tumors requires imaging children with tailored MRI protocols. Considerable variation in pediatric brain tumor imaging protocols exist across institutions, which can cause inaccuracies in staging. Although recently consensus guidelines for pediatric brain tumor imaging protocols have been established, adherence to these guidelines across institutions is unknown. Institutions may perceive guidelines as applicable for research purposes rather than routine clinical practice. We aim to explore variability in pediatric brain tumor imaging protocols across institutions and determine the factors influencing guideline divergence. METHODS AND METHODOLOGY Faculty at institutions routinely imaging pediatric patients with brain tumors were invited to share brain tumor MRI protocols and complete a structured survey. The survey included multiple choice and open-ended queries concerning practice type, the presence of an institutional imaging protocol for pediatric brain tumors and if the protocol was based on specific consensus recommendations. We assessed if protocols included the 7 minimum recommended MRI sequences by the Children’s Oncology Group (COG) guidelines, including slice thickness and gap. RESULTS Twenty emails were sent and responses were received from 12 institutions (the data collection is still ongoing). 2 institutions did not have dedicated protocols for pediatric brain tumors. On average, institutions had 5.7 out of the 7 minimum COG-recommended sequences. All 10 institutions included 3D T1 IR-GRE/TSE, DWI or DTI, and SWI sequences in their protocols. 9 (90%) included post-contrast 3D T1 IR-GRE/TSE, 7 (70%) included T2 TSE/FSE, 7 (70%) included either 3D T2 FLAIR or T2 FLAIR TSE/FSE, and 4 (40%) included T1/TSE/FSE. Across all sequences in all protocols, adherence to recommended imaging parameters was 87% for plane, 71% for slice thickness, and 85% for gap. CONCLUSION Significant variability in pediatric brain tumor imaging protocols exists across institutions Standardizing imaging protocols through collaboration is vital in ensuring uniformity in care.

IMG-08. SYNTHESIZING MISSING MRI SEQUENCES IN PEDIATRIC BRAIN TUMORS USING GENERATIVE ADVERSARIAL NETWORKS; TOWARDS IMPROVED VOLUMETRIC TUMOR ASSESSMENT

Abstract BACKGROUND Standard MRI sequences, such as pre- and post-contrast T1w, T2w, and FLAIR images, are essential for optimizing segmentation of tumor subregions and evaluating treatment responses in pediatric brain tumors (PBTs). However, MRI sets are often incomplete due to imaging artifacts or inconsistent acquisition protocols across various centers. Generative Adversarial Networks (GANs) have been effectively utilized to generate missing MRI sequences for adult brain tumors. This study applies image-to-image translation models using GANs to synthesize missing FLAIR images from T2w images in PBTs. METHODS This retrospective study developed two GAN models, pix2pix in both 2D and 3D, trained and validated on T2w and FLAIR image pairs from 79 and 19 patients, respectively, with pediatric-type diffuse high-grade gliomas (diffuse midline glioma (DMG) including diffuse intrinsic pontine glioma (DIPG)), collected from the Children’s Brain Tumor Network (CBTN). The 2D pix2pix model processes each T2w MRI volume slice-by-slice, generating corresponding FLAIR slices to reconstruct the complete image volume. The 3D pix2pix model allows translation of 3D FLAIR volumes directly from T2w volumes without the need for individual slice processing and reconstruction. The quality of the generated FLAIR volumes was evaluated using Structural Similarity Index (SSI; best closer to 1) and Mean Squared Error (MSE; best near 0) on the validation set. RESULTS The 2D and 3D models achieved robust performance with median SSI, MSE of 0.88, 0.004, and 0.92, 0.002, respectively. CONCLUSIONS The GAN models developed in this study effectively generate missing FLAIR MRI volumes from corresponding T2w images in PBTs. The 3D model, outperforming 2D, speeds the overall processing pipeline by eliminating volume slicing and reconstruction. Future work includes assessing the impact of these synthesized images on the accuracy of our pretrained autosegmentation models in differentiating tumor subregions and measuring tumor volumes in longitudinal MRIs.