Brain Magnetic Resonance Research Articles

Neurološki simptom kao prvi znak kolorektalnog karcinoma

Aim: To present a rare case of a patient with an isolated brain metastasis from colorectal carcinoma in whom an extensive diagnostic workup failed to identify the primary tumour. Case report: A 73-year-old female patient was admitted to the Neurology Department due to weakness in the right side of her body. Imaging with multislice computed tomography (MSCT) and brain magnetic resonance (MR) confirmed an expansive lesion in the left frontal region with surrounding oedema. Radiological findings suggested either a metastasis or a high-grade primary glioma. As the diagnostic workup did not reveal the primary tumour, the patient was transferred to Clinical Hospital Dubrava for surgical treatment. An intracranial tumour resection was performed with a histopathological confirmation of colon carcinoma metastasis. Postoperative brain irradiation was administered. A positron emission tomography scan (PET) showed focal accumulation of radioisotopes without clear pathological substrate in the appendix and cecoascedent junction area. Repeated colonoscopy did not find any pathological mucosal changes. The patient declined systemic treatment. Overall survival was 11 months, in line with literature data. Conclusion: The diagnosis of brain metastases from colorectal carcinoma is very rare, and even rarer is the diagnosis of isolated brain metastases without extracranial progression, counting for <0.5%. Therapeutic options for brain metastases from colorectal carcinoma include surgery, radio-neurosurgery, stereotactic radiotherapy, and brain irradiation. In our opinion, diagnostic evaluation of the central nervous system in patients with colorectal carcinoma should be considered if the patient has neurological symptoms or if there is an increase in carcinoembryonic antigen (CEA) levels without signs of extracranial disease progression.

La imagen de estasis predice el riesgo de eventos cardioembólicos tras el infarto agudo de miocardio: el estudio ISBITAMI

Introduction and objectivesIn the setting of ST-segment elevation myocardial infarction (STEMI), imaging-based biomarkers could be useful for guiding oral anticoagulation to prevent cardioembolism. Our objective was to test the efficacy of intraventricular blood stasis imaging for predicting a composite primary endpoint of cardioembolic risk during the first 6 months after STEMI. MethodsWe designed a prospective clinical study, Imaging Silent Brain Infarct in Acute Myocardial Infarction (ISBITAMI), including patients with a first STEMI, an ejection fraction ≤ 45% and without atrial fibrillation to assess the performance of stasis metrics to predict cardioembolism. Patients underwent ultrasound-based stasis imaging at enrollment followed by heart and brain magnetic resonance at 1-week and 6-month visits. From the stasis maps, we calculated the average residence time, RT, of blood inside the left ventricle and assessed its performance to predict the primary endpoint. The longitudinal strain of the 4 apical segments was quantified by speckle tracking. ResultsA total of 66 patients were assigned to the primary endpoint. Of them, 17 patients had 1 or more events: 3 strokes, 5 silent brain infarctions, and 13 mural thromboses. No systemic embolisms were observed. RT (OR, 3.73; 95%CI, 1.75-7.9; P<.001) and apical strain (OR, 1.47; 95%CI, 1.13-1.92; P=.004) showed complementary prognostic value. The bivariate model showed a c-index=0.86 (95%CI, 0.73-0.95), a negative predictive value of 1.00 (95%CI, 0.94-1.00), and positive predictive value of 0.45 (95%CI, 0.37-0.77). The results were confirmed in a multiple imputation sensitivity analysis. Conventional ultrasound-based metrics were of limited predictive value. ConclusionsIn patients with STEMI and left ventricular systolic dysfunction in sinus rhythm, the risk of cardioembolism may be assessed by echocardiography by combining stasis and strain imaging. Registered at ClinicalTrials.gov (NCT02917213).

Determinants of cognitive dysfunction in adults with sickle cell–related stroke or suspected neurological morbidity

AbstractThe prognosis of sickle cell disease (SCD) in adults is determined primarily by damage to targeted organs such as the brain. Cognitive dysfunction in SCD is a common chronic neurological manifestation, but studies remain mostly descriptive in adults. The objective of this study was to better characterize the cognitive profile and the association between cognitive dysfunction and brain lesions. We included adult patients with SCD referred for a neurological assessment. An adapted battery of neuropsychological tests was used to assess cognitive deficits. Brain or arterial abnormalities were assessed using brain magnetic resonance imaging/magnetic resonance angiography and a cervical and transcranial Doppler ultrasound. The cognitive profile of 96 patients was characterized by deficits in processing speed (58%), short-term memory (34%), and working memory (24%). Brain infarcts were found in 56% of patients and intracranial vasculopathy in 49%. Twenty percent of patients had no brain abnormalities. Processing speed dysfunction was associated with territorial infarcts (odds ratio [OR], 3.1; P = .03) and education outside of France (OR, 4.7; P = .02). Short-term memory dysfunction was associated with territorial infarcts (OR, 3.4; P = .01) and a low educational level (OR, 8.2; P = .01). Working memory dysfunction was associated with a low educational level (OR, 4.3; P = .05) and vasculopathy (OR, 3.7; P = .03). Cognitive dysfunction appears to be a hallmark sign of SCD, particularly for adults with sickle cell-related stroke or suspected neurological morbidity. Assessment of such dysfunction could be used in longitudinal follow-up and clinical trials.

Anomalies of Midbrain/Hindbrain Development: Malformations of Cerebellum: Diagnosis, Classification, and Rehabilitative Hypothesis

AbstractExtensive research has been conducted on the cerebellum, making it one of the most thoroughly investigated regions of the brain. It plays a fundamental role not only in motor control but also in motor learning and cognition. The development of the cerebellum is a lengthy process, beginning during the embryonic period up to the first years of life. This slow and protracted process makes it a vulnerable organ liable to different insults, responsible for many developmental disorders such as Dandy–Walker syndrome, medulloblastoma, dystroglicanopathy, pontocerebellar hypoplasia, thubulinopathies, and Jubert syndrome. Due to several factors, the true prevalence of cerebellar malformations is not known in most cases. The cerebellum undergoes development through following four fundamental stages:(1) Identification of the cerebellar region at the boundary between the midbrain and hindbrain.(2) Establishment of two cell proliferation compartments: firstly, Purkinje cells and deep cerebellar nuclei emerge from the ventricular zone of the metencephalic alar plate; secondly, granule cell precursors are generated from a separate proliferation compartment known as the upper rhombic lip.(3) Migration of granule cells toward the interior: granule precursor cells constitute the external granular layer (EGL), and during the initial postnatal year, granule cells migrate inward to their final position in the internal granular layer.(4) Formation of cerebellar circuitry and subsequent differentiation.Based on different types of involvement of the structures detected in the brain magnetic resonance, the classification of brainstem and cerebellar anomalies is divided into three categories: (1) mainly the cerebellum, (2) mainly the brain stem, and (3) both involved. This review will outline the developmental processes of the cerebellum and delve into common developmental disorders associated with it, including the Dandy–Walker syndrome, cerebellar hypoplasia, rhomboencephalosynapsis, lissencephaly, and gray matter heterotopias.

Generalizing the Enhanced-Deep-Super-Resolution Neural Network to Brain MR Images: A Retrospective Study on the Cam-CAN Dataset.

The Enhanced-Deep-Super-Resolution (EDSR) model is a state-of-the-art convolutional neural network suitable for improving image spatial resolution. It was previously trained with general-purpose pictures and then, in this work, tested on biomedical magnetic resonance (MR) images, comparing the network outcomes with traditional up-sampling techniques. We explored possible changes in the model response when different MR sequences were analyzed. T1w and T2w MR brain images of 70 human healthy subjects (F:M, 40:30) from the Cambridge Centre for Ageing and Neuroscience (Cam-CAN) repository were down-sampled and then up-sampled using EDSR model and BiCubic (BC) interpolation. Several reference metrics were used to quantitatively assess the performance of up-sampling operations (RMSE, pSNR, SSIM, and HFEN). Two-dimensional and three-dimensional reconstructions were evaluated. Different brain tissues were analyzed individually. The EDSR model was superior to BC interpolation on the selected metrics, both for two- and three- dimensional reconstructions. The reference metrics showed higher quality of EDSR over BC reconstructions for all the analyzed images, with a significant difference of all the criteria in T1w images and of the perception-based SSIM and HFEN in T2w images. The analysis per tissue highlights differences in EDSR performance related to the gray-level values, showing a relative lack of outperformance in reconstructing hyperintense areas. The EDSR model, trained on general-purpose images, better reconstructs MR T1w and T2w images than BC, without any retraining or fine-tuning. These results highlight the excellent generalization ability of the network and lead to possible applications on other MR measurements.

Stasis imaging predicts the risk of cardioembolic events related to acute myocardial infarction: the ISBITAMI study

Introduction and objectivesIn the setting of ST-segment elevation myocardial infarction (STEMI), imaging-based biomarkers could be useful for guiding oral anticoagulation to prevent cardioembolism. Our objective was to test the efficacy of intraventricular blood stasis imaging for predicting a composite primary endpoint of cardioembolic risk during the first 6 months after STEMI. MethodsWe designed a prospective clinical study, Imaging Silent Brain Infarct in Acute Myocardial Infarction (ISBITAMI), including patients with a first STEMI, an ejection fraction ≤ 45% and without atrial fibrillation to assess the performance of stasis metrics to predict cardioembolism. Patients underwent ultrasound-based stasis imaging at enrollment followed by heart and brain magnetic resonance at 1-week and 6-month visits. From the stasis maps, we calculated the average residence time, RT, of blood inside the left ventricle and assessed its performance to predict the primary endpoint. The longitudinal strain of the 4 apical segments was quantified by speckle tracking. ResultsA total of 66 patients were assigned to the primary endpoint. Of them, 17 patients had 1 or more events: 3 strokes, 5 silent brain infarctions, and 13 mural thromboses. No systemic embolisms were observed. RT (OR, 3.73; 95%CI, 1.75-7.9; P<.001) and apical strain (OR, 1.47; 95%CI, 1.13-1.92; P=.004) showed complementary prognostic value. The bivariate model showed a c-index=0.86 (95%CI, 0.73-0.95), a negative predictive value of 1.00 (95%CI, 0.94-1.00), and positive predictive value of 0.45 (95%CI, 0.37-0.77). The results were confirmed in a multiple imputation sensitivity analysis. Conventional ultrasound-based metrics were of limited predictive value. ConclusionsIn patients with STEMI and left ventricular systolic dysfunction in sinus rhythm, the risk of cardioembolism may be assessed by echocardiography by combining stasis and strain imaging. Registered at ClinicalTrials.gov (NCT02917213).

Brain Age Prediction Using Multi-Hop Graph Attention Combined with Convolutional Neural Network.

Convolutional neural networks (CNNs) have been used widely to predict biological brain age based on brain magnetic resonance (MR) images. However, CNNs focus mainly on spatially local features and their aggregates and barely on the connective information between distant regions. To overcome this issue, we propose a novel multi-hop graph attention (MGA) module that exploits both the local and global connections of image features when combined with CNNs. After insertion between convolutional layers, MGA first converts the convolution-derived feature map into graph-structured data by using patch embedding and embedding-distance-based scoring. Multi-hop connections between the graph nodes are modeled by using the Markov chain process. After performing multi-hop graph attention, MGA re-converts the graph into an updated feature map and transfers it to the next convolutional layer. We combined the MGA module with sSE (spatial squeeze and excitation)-ResNet18 for our final prediction model (MGA-sSE-ResNet18) and performed various hyperparameter evaluations to identify the optimal parameter combinations. With 2788 three-dimensional T1-weighted MR images of healthy subjects, we verified the effectiveness of MGA-sSE-ResNet18 with comparisons to four established, general-purpose CNNs and two representative brain age prediction models. The proposed model yielded an optimal performance with a mean absolute error of 2.822 years and Pearson's correlation coefficient (PCC) of 0.968, demonstrating the potential of the MGA module to improve the accuracy of brain age prediction.

CT and MR Appearance of Teeth: Analysis of Anatomy and Embryology and Implications for Disease.

Abnormalities of dental development and anatomy may suggest the presence of congenital or acquired anomalies. The detection of abnormalities, therefore, is an important skill for radiologists to achieve. Knowledge of dental embryology and an understanding of the radiologic appearances of teeth at various stages of maturation are required for the appreciation of abnormal dental development. While many tooth abnormalities are well-depicted on dedicated dental radiographs, the first encounter with a dental anomaly may be by a radiologist on a computed tomographic (CT) or magnetic resonance (MR) exam performed for other reasons. This article depicts normal dental anatomy and development, describing the appearance of the neonatal dentition on CT and MRI, the modalities most often encountered by clinical radiologists. The radiology and dental literature are reviewed, and key concepts are illustrated with supplemental cases from our institution. The value of knowledge of dental development is investigated using the analysis of consecutive MR brain examinations. Finally, the anatomical principles are applied to the diagnosis of odontogenic infection on CT. Through analysis of the literature and case data, the contrast of dental pathology with normal anatomy and development facilitates the detection and characterization of both congenital and acquired dental disease.

An effective and robust lattice Boltzmann model guided by atlas for hippocampal subregions segmentation.

Given the varying vulnerability of the rostral and caudal regions of the hippocampus to neuropathology in the Alzheimer's disease (AD) continuum, accurately assessing structural changes in these subregions is crucial for early AD detection. The development of reliable and robust automatic segmentation methods for hippocampal subregions (HS) is of utmostimportance. Our aim is to propose and validate a HS segmentation model that is both training-free and highly generalizable. This method should exhibit comparable accuracy and efficiency to state-of-the-art techniques. The segmented HS can serve as a biomarker for studying the progression ofAD. We utilized the functional magnetic resonance imaging of the Brain's Integrated Registration and Segmentation Tool (FIRST) to segment the entire hippocampus. By intersecting the segmentation results with the Brainnetome (BN) atlas, we obtained coarse segmentation of the four HS regions. This coarse segmentation was then employed as a shape prior term in the lattice Boltzmann (LB) model, as well as for initializing contours. Additionally, image gradients and local gray levels were integrated into the external force terms of the LB model to refine the coarse segmentation results. We assessed the segmentation accuracy of the model using the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset and evaluated the potential of the segmentation results as AD biomarkers on both the ADNI and Xuanwudatasets. The median Dice similarity coefficients (DSC) for the left caudal, right caudal, left rostral, and right rostral hippocampus were 0.87, 0.88, 0.88, and 0.89, respectively. The proportion of segmentation results with a DSC exceeding 0.8 was 77%, 78%, 77%, and 94% for the respective regions. In terms of volume, the correlation coefficients between the segmentation results of the four HS regions and the gold standard were 0.95, 0.93, 0.96, and 0.96, respectively. Regarding asymmetry, the correlation coefficient between the segmentation result's right caudal minus left caudal and the corresponding gold standard was 0.91, while for right rostral minus left rostral, it was 0.93. Over time, we observed a decline in the volumes of the four HS regions and the total hippocampal volume of mild cognitive impairment (MCI) converters. Analysis of inter-group differences revealed that, except for the right rostral region in the ADNI dataset, the p-values for the four HS regions in the normal controls (NC), MCI, and AD groups from both datasets were all below 0.05. The right caudal hippocampal volume demonstrated correlation coefficients of 0.47 and 0.43 with the mini-mental state examination (MMSE) and Montreal cognitive assessment (MoCA), respectively. Similarly, the left rostral hippocampal volume showed correlation coefficients of 0.50 and 0.58 with MMSE and MoCA,respectively. Our framework allows for direct application to different brain magnetic resonance (MR) datasets without the need for training. It eliminates the requirement for complex image preprocessing steps while achieving segmentation accuracy comparable to deep learning (DL) methods even with small sample sizes. Compared to traditional active contour models (ACM) and atlas-based methods, our approach exhibits significant speed advantages. The segmented HS regions hold promise as potential biomarkers for studying the progression ofAD.

Erythropoietin in Neonates with Perinatal Asphyxia Undergoing Therapeutic Hypothermia—A Prospective Cohort Study

Aims To assess the safety and feasibility of erythropoietin (EPO) in asphyxiated neonates undergoing therapeutic hypothermia (TH). Subjects and methods This study enrolled 60 neonates with HIE undergoing TH. 30 neonates were divided into EPO with TH group, where neonates received EPO (dose), at a dose rate of 1000 IU/kg/48 hours, for four doses and the remaining 30 neonates undergoing TH were into control group. Magnetic resonance imaging brain was undertaken between 10 and 14 days of life in surviving neonates. The Bayley Scales of Infant (BSI) Development IV was performed at regular intervals up to the age of 12 months. Results Amplitude-integrated electroencephalogram (aEEG) showed burst suppression (19% vs. 11%), low voltage (10% vs. 4%), and flat trace (7% vs. 3%) in control group in comparison with EPO concurrent with TH group which was statistically significant. Brain magnetic-resonance imaging (MRI) done at 12 days (±2 days) showed significant brain injury patterns such as severe brain injury (4% vs. 9% P = .05) and regional specific HIE (7% vs. 13%, P = .03) in control group of only TH neonates. At 12 months, neurodevelopment outcomes in EPO with TH group neonates showed a favorable outcome. Conclusion rhEPO concurrent with TH in HIE neonates resulted in significantly less severe brain injuries in MRI brain in HIE neonates. aEEG changes were less statistically significant in EPO with TH group with favorable neurodevelopmental outcomes at 12 months of age.

Retrospective Evaluation of Patients Admitted to the Pediatric Neurology Outpatient Clinic with Headache: Experience of a Tertiary Hospital.

It was aimed to determine the etiological and clinical features of pediatric patients with headache complaints. The files of patients who were admitted to the pediatric neurology outpatient clinic with headache were reviewed retrospectively. Patients' age, gender, features of headache, symptoms accompanying headache, available blood tests, brain magnetic resonance (MR) and electroencephalography (EEG) results were recorded. Of the total 470 patients, aged between 3 and 17 years, 291 (61.9%) were female and 179 (39.1%) were male. The mean age of the patients was 12.38±3.45 years. According to age groups, there were 16 (3.4%) patients under the age of 5, 159 (33.8%) between the ages of 6-11, and 295 (62.8%) patients aged 12-17 years. While 289 (61.5%) patients were diagnosed with primary headache, 122 (26.0%) patients were diagnosed with secondary headache, and headaches of 59 (12.5%) patients could not be classified. The most common primary headaches were tension-type headache (TTN) (n=177, 37.7%) and migraine (n=111, 23.6%). The 86 (70.5%) of the patients with secondary headache were diagnosed with sinusitis. Abnormal neurological examination finding was determined in 8 (1.7%) patients. Brain MR was performed in 439 (93.4%) of all patients and abnormal brain MR findings were detected in 52 (11.8%) patients. EEG was performed in 205 (43.6%) of all patients and abnormal EEG findings were detected in 24 (11.7%) patients. According to age groups, headache was most common in the 12-17 age group. The most common causes of headache were TTN and migraine, respectively. The most common secondary headache cause was sinusitis. We think that physical and neurological examination still maintains its priority in determining the causes of headache.

Conditional spatial biased intuitionistic clustering technique for brain MRI image segmentation.

In clinical research, it is crucial to segment the magnetic resonance (MR) brain image for studying the internal tissues of the brain. To address this challenge in a sustainable manner, a novel approach has been proposed leveraging the power of unsupervised clustering while integrating conditional spatial properties of the image into intuitionistic clustering technique for segmenting MRI images of brain scans. In the proposed technique, an Intuitionistic-based clustering approach incorporates a nuanced understanding of uncertainty inherent in the image data. The measure of uncertainty is achieved through calculation of hesitation degree. The approach introduces a conditional spatial function alongside the intuitionistic membership matrix, enabling the consideration of spatial relationships within the image. Furthermore, by calculating weighted intuitionistic membership matrix, the algorithm gains the ability to adapt its smoothing behavior based on the local context. The main advantages are enhanced robustness with homogenous segments, lower sensitivity to noise, intensity inhomogeneity and accommodation of degree of hesitation or uncertainty that may exist in the real-world datasets. A comparative analysis of synthetic and real datasets of MR brain images proves the efficiency of the suggested approach over different algorithms. The paper investigates how the suggested research methodology performs in medical industry under different circumstances including both qualitative and quantitative parameters such as segmentation accuracy, similarity index, true positive ratio, false positive ratio. The experimental outcomes demonstrate that the suggested algorithm outperforms in retaining image details and achieving segmentation accuracy.

Amyloid/Tau/Neurodegeneration (ATN) image‐based biomarker trajectories of Alzheimer’s disease through 5‐year longitudinal follow‐up

AbstractBackgroundTo investigate the progression of Alzheimer’s disease (AD) in image‐based Amyloid/Tau/Neurodegeneration (ATN) biomarkers through 5‐year longitudinal follow‐up.MethodAmong the participants who completed 18F‐florbetaben (FBB) for amyloid and 18F‐flortaucipir (FTP) for tau positron emission tomography (PET), and brain magnetic resonance (MR) imaging for neurodegeneration at a first visit, we enrolled the 145 participants who the studies at least twice [67 amyloid negative cognitive unimpaired (A‐CU), 10 amyloid positive cognitive unimpaired (A+CU), 68 amyloid positive cognitive impaired (A+CI)]. 70 participants completed the scans twice (25.2±3.1, months), and 75 participants completed the scans three times (57.9±7.4, months). We estimated the temporal trajectories of ATN biomarkers using a restricted cubic spline model and modified Euler method between baseline and annual change rates. The standardized uptake value ratio of FBB/FTP PET and hippocampal volume were used for representing each ATN biomarker, respectively.ResultThe FBB curve of the global cortex and AD‐specific regions reaches a plateau with a sigmoid appearance. FTP curves in the regions for Braak stages I‐II and III‐IV show similar sigmoid appearance while the global cortical and stage V‐VI tau curves show an exponential increase continuously. The hippocampal volume curve shows a linear decrease until the time point when the tau accumulation begins in the Braak stage I‐II, then decreases to a reversed sigmoid appearance.ConclusionLike the already known saturation of amyloid burden, it was also shown that the tau burden of the region for Braak stages I‐II and III‐IV saturated, but not in Braak stages V‐VI. Hippocampal atrophy seems to accelerate when tau begins to accumulate in early Braak stage regions. In future work, it is necessary to enroll patients at an advanced stage and collaborate with an international cohort.

Exploring shared neural substrates underlying cognition and gait variability in adults without dementia

BackgroundHigh gait variability is associated with neurodegeneration and cognitive impairments and is predictive of cognitive impairment and dementia. The objective of this study was to identify cortical or subcortical structures of the brain shared by gait variability measured using a body-worn tri-axial accelerometer (TAA) and cognitive function.MethodsThis study is a part of a larger population-based cohort study on cognitive aging and dementia. The study included 207 participants without dementia, with a mean age of 72.6, and 45.4% of them are females. We conducted standardized diagnostic interview including a detailed medical history, physical and neurological examinations, and laboratory tests for cognitive impairment. We obtained gait variability during walking using a body-worn TAA along and measured cortical thickness and subcortical volume from brain magnetic resonance (MR) images. We cross-sectionally investigated the cortical and subcortical neural structures associated with gait variability and the shared neural substrates of gait variability and cognitive function.ResultsHigher gait variability was associated with the lower cognitive function and thinner cortical gray matter but not smaller subcortical structures. Among the clusters exhibiting correlations with gait variability, one that included the inferior temporal, entorhinal, parahippocampal, fusiform, and lingual regions in the left hemisphere was also associated with global cognitive and verbal memory function. Mediation analysis results revealed that the cluster’s cortical thickness played a mediating role in the association between gait variability and cognitive function.ConclusionGait variability and cognitive function may share neural substrates, specifically in regions related to memory and visuospatial navigation.

Accurate personalized survival prediction for amyotrophic lateral sclerosis patients

Amyotrophic Lateral Sclerosis (ALS) is a rapidly progressive neurodegenerative disease. Accurately predicting the survival time for ALS patients can help patients and clinicians to plan for future treatment and care. We describe the application of a machine-learned tool that incorporates clinical features and cortical thickness from brain magnetic resonance (MR) images to estimate the time until a composite respiratory failure event for ALS patients, and presents the prediction as individual survival distributions (ISDs). These ISDs provide the probability of survival (none of the respiratory failures) at multiple future time points, for each individual patient. Our learner considers several survival prediction models, and selects the best model to provide predictions. We evaluate our learned model using the mean absolute error margin (MAE-margin), a modified version of mean absolute error that handles data with censored outcomes. We show that our tool can provide helpful information for patients and clinicians in planning future treatment.

Improved segmentation of basal ganglia from MR images using convolutional neural network with crossover-typed skip connection.

Accurate and automatic segmentation of basal ganglia from magnetic resonance (MR) images is important for diagnosis and treatment of various brain disorders. However, the basal ganglia segmentation is a challenging task because of the class imbalance and the unclear boundaries among basal ganglia anatomical structures. Thus, we aim to present an encoder-decoder convolutional neural network (CNN)-based method for improved segmentation of basal ganglia by focusing on skip connections that determine the segmentation performance of encoder-decoder CNNs. We also aim to reveal the effect of skip connections on the segmentation of basal ganglia with unclear boundaries. We used the encoder-decoder CNNs with the following five patterns of skip connections: without skip connection, with full-resolution horizontal skip connection, with horizontal skip connections, with vertical skip connections, and with crossover-typed skip connections (the proposed method). We compared and evaluated the performance of the CNNs in the experiment of basal ganglia segmentation using T1-weighted MR brain images of 79 patients. The experimental results showed that the skip connections at each scale level help CNNs to acquire multi-scale image features, the vertical skip connections contribute on acquiring finer image features for segmentation of smaller anatomical structures with more blurred boundaries, and the crossover-typed skip connections, a combination of horizontal and vertical skip connections, provided better segmentation accuracy. This paper investigated the effect of skip connections on the basal ganglia segmentation and revealed the crossover-typed skip connections might be effective for improving the segmentation of basal ganglia with the class imbalance and the unclear boundaries.

Brain Tumor Localization Using N-Cut

A brain tumor is an abnormal collection of tissue in the brain. When tumors form, they are classified as either malignant or benign. It is critical to notice and identify the existence of tumors in brain images since they can be life threatening. This paper illustrates a novel segmentation method in which threshold technique is combined with normalized cut (Ncut) for the segregation of the tumors from brain magnetic resonance (MR) images. Image segmentation is a technique for grouping images. It is a method of splitting an image into sections with comparable attributes such as intensity, texture, colour, and so on. In thresholding, an object is distinguished from the background, and for the proposed segmentation methodology, the threshold value is determined by normalized graph cut. A weighted graph is divided into disjointed sets (groups) in which the similarity within a group is high and the similarity across groups is low. A graph-cut is a grouping approach in which the total weight of edges eliminated between these two pieces is used to calculate the degree of dissimilarity between these two groups. The normalized cut criterion is used to calculate the total likeness within the groups as well as the dissimilarity between the different groups.

Pediatric anti-NMDAR encephalitis with demyelination on brain MRI: A single center study

ObjectiveTo explore the clinical characteristics, immunotherapy response, and prognosis of pediatric anti-N-methyl-D-aspartate receptor (NMDAR) encephalitis associated with demyelination on brain magnetic resonance (MRI). MethodsWe retrospectively reviewed the medical records of children diagnosed with anti-NMDAR encephalitis in our hospital between January 2016 and December 2021. All children with evidence of demyelination on brain MRI were included. ResultsA total of 183 anti-NMDAR encephalitis children were included; 8.7 % (16/183) of them had demyelination on brain MRI. Nine were positive for myelin oligodendrocyte glycoprotein (MOG)-IgG, while two were positive for both MOG-IgG and glial fibrillary acidic protein (GFAP)-IgG. Four patients had a history of acquired demyelinating syndromes and encephalitis, respectively, while nine (56.3 %) had atypical symptoms of anti-NMDAR encephalitis. All children had supratentorial demyelination on brain MRI; four of them had additional infratentorial lesions. All children received first-line immunotherapy; four were administered repeated first-line immunotherapy and/or rituximab because of poor initial response. During the follow-up, 37.5 % (6/16) of the children relapsed, but all responded well to immunotherapy. There were no significant differences in mRS score before immunotherapy, response to first-line immunotherapy, and long-term prognosis between anti-NMDAR encephalitis children with and without demyelination. However, patients with demyelination were more likely to have a history of acquired demyelinating syndromes or unexplained cortical encephalitis and to relapse. ConclusionPediatric anti-NMDAR encephalitis can co-occur with demyelination and has a high rate of MOG-IgG positivity. A history of acquired demyelinating syndromes or unexplained cortical encephalitis and atypical symptoms may indicate demyelination in children with anti-NMDAR encephalitis. Pediatric anti-NMDAR encephalitis with demyelination is more likely to relapse and needs a closer follow-up. However, it remains unknown whether more intensive immunotherapy is required in these patients.

THU182 Widening The Spectrum Of ABCC8-related Disease: Insulin-mediated Pseudoacromegaly

Abstract Disclosure: I.F. Machado: None. B.B. Mendonca: None. A. Latronico: None. M.Y. Nishi: None. L.G. Gomes: None. Introduction: Insulin-mediated pseudoacromegaly is a rare condition characterized by acromegaloid phenotype without excessive GH secretion in the context of severe hyperinsulinemia. Case report: A 10.5 years old girl with tall stature, obesity, acanthosis nigricans, neurosensorial hearing deficiency and acromegaloid features was referred for clinical investigation. She was born of non-consanguineous parents, full-term, with normal birth weight and stature. She had no history of neonatal hypoglycemia or in infancy. At six years of age, she presented marked accelerated growth velocity and weight gain resulting in bilateral slipped femoral epiphysis. She developed thelarche at 7 years old, and was treated with gonadotropin releasing hormone agonist (GnRHa) from 8-10.5 years. Hearing deficiency was diagnosed at 7 years of age. Her family history was remarkable for paternal sudden cardiac death at age 35. At physical examination, she had an acromegaloid facial appearance with a prominent forehead and jaw, large hands and feet, and severe acanthosis nigricans in neck and axillae. Her weight and height were 99.5 kgs (Z 8.8) and 170 cm (Z 4.6), respectively; target height was 1.66 m (Z 0,72) and BMI was 34.4 (Z 3.5). Her breast and pubic hair development were Tanner stage IV. Laboratory exams showed normal IGF-1 and random GH levels. There was an abnormal response of GH (2.63 ng/mL) following an oral glucose tolerance test (OGTT), considered as false positive in the context of obesity and puberty. Her brain magnetic resonance was normal. She had a hemoglobin glycated of 6.1% and very high basal insulin levels of 98 µUI/mL with a normal glucose value of 81 mg/dL and OGTT-120 minutes insulin levels &amp;gt; 1000 µUI/mL with glucose of 113 mg/dL. Exome sequencing analysis showed a heterozygous missense ABCC8 variant c.622G&amp;gt;A p.(Glu208Lys), located in an exonic hotspot, and allele frequency absent from gnomAD databases. The ACMG classified this variant as pathogenic. Pseudoacromegaly, a heterogeneous condition, is caused by multiple genetic and pathophysiological backgrounds. In insulin-mediated pseudoacromegaly, the defects are mainly associated with postreceptor insulin signaling and action. We described a new type of hyperinsulinemic pseudoacromegaly caused by a mutation in the ABCC8 gene, which encodes the sulfonylurea receptor, a member of the K(ATP) channel superfamily. This gene is associated with congenital hyperinsulinemia causing hypoglycemia, generally when in homozygous state. This clinical report illustrates a new phenotype associated with ABCC8 mutation, and identifies a new genetic cause of hyperinsulinemic pseudoacromegaly. Conclusion: We described a patient with insulin-mediated pseudoacromegaly associated with neurosensorial hearing deficiency caused by a dominant variant in the ABCC8 gene, defining a new genetic cause of this intriguing disorder. Presentation: Thursday, June 15, 2023

The use of artificial intelligence-supported communication technologies in neurological fields: A case study on brain tumor detection

Objective: The global health system is being shaped by multidisciplinary studies on the diagnosis of diseases and the provision of effective treatment services. Information and communication technologies have been developing laboratory and imaging studies through artificial intelligence-supported systems for the last twenty years. Studies with high accuracy levels in the diagnosis and treatment protocols of diseases make important contributions to making healthy decisions. Artificial intelligence applications have been actively used in the treatment processes of neurological cancer cases in the field of health, as in many fields in recent years. Among these applications, the machine learning model has started to be preferred in the detection of brain tumors because it can provide remarkable results. The main purpose of the study is to provide a supportive analysis for the organization of early diagnosis and rapid treatment in areas such as intracranial pressure, tumor treatment and radiotherapy of patients during intensive care processes. Materials and Methods: In this study, the method developed by doctors with machine learning Kaggle and developers of samples in the network through an example of an application that was developed through machine learning on brain tumors, brain tumor detection carried on with the validation of the data sets includes four classifications. Results: The study consists of two different study systems, namely practice and test. Sectional images from 2865 brain magnetic resonance imaging (MRI )and computed tomography (CT) samples were examined as training in the first stage of the application using the convolutional neural network (CNN) model, and the detected tumors were classified. In this context, MRI results were obtained on 2865 samples with 2470 units and 86.23% with tumors, and 395 units and 13.76% no tumors. Conclusion: In the study, samples with tumors were detected in a 3-month period for brain tumor detection with artificial intelligence and classified typologically. Accordingly, the reliability of the application was proven by providing 98.55% verification on 2865 samples, 3 different tumor types and no tumor data.