Scan With Me (SWiM): A train-the-trainer program to upskill MRI personnel in low- and middle- income countries.

PURPOSE: Access to magnetic resonance imaging (MRI) in low-and-middle-income countries (LMICs) remains among the poorest in the world. The lack of skilled MRI personnel exacerbates access gaps, reinforcing longstanding health disparities. The SWiM Program aims to sustainably create a network of highly skilled MRI technologists in LMICs who will facilitate the transfer of MRI knowledge and skills to their peers and contribute to implementation of highly valuable imaging protocols for impactful clinical and research use. METHODS: The program introduces a case-based curriculum designed using a novel train-the-trainer approach, integrated with peer collaborative learning to upskill practicing MRI technologists in LMICs. The six-week curriculum uses the Teach-Try-Use approach, which combines self-paced didactic lectures covering the basics of MR image acquisition (teach) with hands-on expert-guided scanning experience (try), and implementation of protocols tailored to provide the best possible images on their infrastructures (use). Each program includes research translation skills training using an established advanced MRI technique relevant to LMICs. A pilot program focused on cardiac MRI (CMR) was conducted to assess the program's curriculum, delivery, and evaluation methods. RESULTS: Forty-three MRI technologists from 16 LMICs participated in the pilot CMR program and over the course of the training, implemented optimized CMR protocols that reduced acquisition times while improving image quality. The training resources and scanner-specific standardized protocols are published openly for public use on an online repository. In general, at the end of the program, learners reported considerable improvements in CMR knowledge and skills. All respondents to the program evaluation survey agreed to recommend the program to their colleagues, while 87% indicated interest in returning to help train others. DISCUSSION: The SWiM Program is the first masterclass in MRI acquisition for practicing imaging technologists in LMICS. The program holds the potential to help reduce disparities in MRI expertise and access. The support of the MRI community, imaging societies, and funding agencies will increase its reach and further its impact in democratizing MRI.

The Changing Epidemiology of Central Nervous System Infection: Can Radiologists Keep Up?

Diagnostic radiologists can increase their clinical value by supplementing image pattern recognition with knowledge of epidemiology and geographic distribution of central nervous system (CNS) infections and their causative organisms. This article reviews the changing global disease patterns, as well as zoonotic outbreaks of henipaviruses, coronaviruses, and other emerging, reemerging, and vector-borne organisms; case examples highlight typical imaging features of CNS infections and their mimics. Technical advances in neuroimaging help to enhance the value of radiologists to the multidisciplinary team and the responses to future pandemic preparation.

Beyond Pattern Recognition: Radiology-Pathology-Clinical Correlation.

Radiology-pathology correlation is essential for multidisciplinary collaboration in diagnosis and understanding the mechanism of CNS damage in infectious processes. The microscopic acute inflammatory processes are well established and are supplemented by a variety of less-invasive microbial and immunohistochemical investigations. Understanding the pathogenesis of pathogen spread and neuroinvasion, vascular and immune-mediated brain, and spinal cord damage are essential for interpreting radiological images.

Acute radiological features facilitate diagnosis and prognosis of anti-N-methyl-d-aspartate receptor (NMDAR) and anti-voltage-gated potassium channel (VGKC) encephalitis in adults.

BACKGROUND: Anti-N-methyl-d-aspartate receptor (NMDAR) and anti-voltage-gated potassium channel (VGKC) encephalitis are the commonest antibody-associated autoimmune encephalitides (AIE). Acute clinical features have been well-described, but data on the role of radiological findings in diagnosis and prognosis of AIE are limited. METHODS: Anti-NMDAR and anti-VGKC encephalitis patients from the National Neuroscience Institute were identified. We compared clinical and paraclinical features, at acute presentation and on follow-up between and within groups. RESULTS: Twenty-six anti-NMDAR and 11 anti-VGKC encephalitis patients were reviewed. At acute presentation, dysautonomia (57.7%) and impairment of consciousness (84.6%) occurred exclusively in anti-NMDAR encephalitis. Cerebrospinal fluid pleocytosis was more common in anti-NMDAR encephalitis (88.5% vs 20.0%, p = 0.003), while ictal electroencephalography abnormalities were more frequent in anti-VGKC encephalitis (11.5% vs 45.5%, p = 0.022). On acute imaging, leptomeningeal enhancement was seen only in anti-NMDAR encephalitis (37.5%), while hippocampal T2 hyperintensities supported the diagnosis of anti-VGKC encephalitis (63.6% vs 12.5%, p = 0.002). At follow-up (median 53.0 months, range 13.0-119.0), anti-NMDAR encephalitis patients had better modified Rankin scale scores (median 0.0 vs 3.0, p = 0.023). Relapses occurred equally in both groups. Anti-VGKC encephalitis patients with abnormal acute MRI were more likely to have poor outcomes compared to those with normal imaging (100% vs 25%, p = 0.008), whereas acute imaging features in anti-NMDAR encephalitis did not predict long-term outcomes. CONCLUSIONS: Acute MRI findings can aid in early diagnosis and prognostication in suspected AIE. Leptomeningeal enhancement in anti-NMDAR encephalitis and hippocampal lesions in anti-VGKC encephalitis, together with typical clinical features, may allow distinction between these antibody subtypes, and specific abnormal imaging features in anti-VGKC encephalitis may be used as a prognostic marker.

Neurology of COVID-19 in Singapore.

PURPOSE: To describe the spectrum of COVID-19 neurology in Singapore. METHOD: We prospectively studied all microbiologically-confirmed COVID-19 patients in Singapore, who were referred for any neurological complaint within three months of COVID-19 onset. Neurological diagnoses and relationship to COVID-19 was made by consensus guided by contemporaneous literature, refined using recent case definitions. RESULTS: 47,572 patients (median age 34 years, 98% males) were diagnosed with COVID-19 in Singapore between 19 March to 19 July 2020. We identified 90 patients (median age 38, 98.9% males) with neurological disorders; 39 with varying certainty of relationship to COVID-19 categorised as: i) Central nervous system syndromes-4 acute disseminated encephalomyelitis (ADEM) and encephalitis, ii) Cerebrovascular disorders-19 acute ischaemic stroke and transient ischaemic attack (AIS/TIA), 4 cerebral venous thrombosis (CVT), 2 intracerebral haemorrhage, iii) Peripheral nervous system-7 mono/polyneuropathies, and a novel group, iv) Autonomic nervous system-4 limited dysautonomic syndromes. Fifty-one other patients had pre/co-existent neurological conditions unrelated to COVID-19. Encephalitis/ADEM is delayed, occurring in critical COVID-19, while CVT and dysautonomia occurred relatively early, and largely in mild infections. AIS/TIA was variable in onset, occurring in patients with differing COVID-19 severity; remarkably 63.2% were asymptomatic. CVT was more frequent than expected and occurred in mild/asymptomatic patients. There were no neurological complications in all 81 paediatric COVID-19 cases. CONCLUSION: COVID-19 neurology has a wide spectrum of dysimmune-thrombotic disorders. We encountered relatively few neurological complications, probably because our outbreak involved largely young men with mild/asymptomatic COVID-19. It is also widely perceived that the pandemic did not unduly affect the Singapore healthcare system.

Phenotypic bases of NOTCH2NLC GGC expansion positive neuronal intranuclear inclusion disease in a Southeast Asian cohort.

Neuronal intranuclear inclusion disease (NIID) is a neurodegenerative disorder associated with GGC repeats of >60 to 500 copies in the 5'-untranslated region of NOTCH2NLC. The clinical and genetic characterization of NIID outside of East Asia remains unknown. We identified twelve patients who underwent genetic testing using long-read sequencing or repeat primed polymerase chain reaction. All were positive for a GGC repeat expansion; the median repeat length was 107 (range 92-138). Ten were Chinese and two of Malay ethnicity. Age at onset ranged from 50 to 69 years. Eight (66.7%) patients had dementia, while four (33.3%) patients were oligosymptomatic, without typical NIID symptoms of dementia, Parkinsonism, or muscle weakness. GGA interruptions within the GGC expansion were present in four patients; the number of GGA interruptions was highest (6.71%) in the patient with the earliest age at onset (50 years). Median plasma neurofilament light level was 47.3 pg/mL in seven patients (range 26-380 pg/mL). The highest level (380 pg/mL) was found in one patient who experienced an encephalitic episode. Overall, we describe a cohort of genetically confirmed NIID patients from Southeast Asia and provide further information that the presence of GGA interruptions within GGC repeat expansions may serve as a potential genetic modifier in NIID.

NOTCH2NLC GGC Repeat Expansions Are Associated with Sporadic Essential Tremor: Variable Disease Expressivity on Long-Term Follow-up.

We screened 662 subjects comprising 462 essential tremor (ET) subjects (285 sporadic, 125 with family history, and 52 probands from well-characterized ET pedigrees) and 200 controls and identified pathogenic NOTCH2NLC GGC repeat expansions in 4 sporadic ET patients. Two patients were followed up for >1 decade; one with 90 repeats remained an ET phenotype that did not evolve after 40 years, whereas another patient with 107 repeats developed motor symptoms and cognitive impairment after 8 to 10 years. Neuroimaging in this patient revealed severe leukoencephalopathy; diffusion-weighted imaging hyperintensity in the corticomedullary junction and skin biopsy revealed intranuclear inclusions suggestive of intranuclear inclusion body disease (NIID). No GGC repeats of >60 units were detected in familial ET cases and controls, although 4 ET patients carried 47 to 53 "intermediate" repeats. NOTCH2NLC GGC repeat expansions can be associated with sporadic ET. Carriers presenting with a pure ET phenotype may or may not convert to NIID up to 4 decades after initial tremor onset. ANN NEUROL 2020;88:614-618.

Comparison of machine learning classifiers for differentiation of grade 1 from higher gradings in meningioma: A multicenter radiomics study.

BACKGROUND AND PURPOSE: Advanced imaging analysis for the prediction of tumor biology and modelling of clinically relevant parameters using computed imaging features is part of the emerging field of radiomics research. Here we test the hypothesis that a machine learning approach can distinguish grade 1 from higher gradings in meningioma patients using radiomics features derived from a heterogenous multicenter dataset of multi-paramedic MRI. METHODS: A total of 138 patients from 5 international centers that underwent MRI prior to surgical resection of intracranial meningiomas were included. Segmentation was performed manually on co-registered multi-parametric MR images using apparent diffusion coefficient (ADC) maps, T1-weighted (T1), post-contrast T1-weighted (T1c), subtraction maps (Sub, T1c - T1), T2-weighted fluid-attenuated inversion recovery (FLAIR) and T2-weighted (T2) images. Feature selection was performed and using cross-validation to separate training from testing data, four machine learning classifiers were scored on combinations of MRI modalities: random forest (RF), extreme gradient boosting (XGBoost), support vector machine (SVM) and multilayer perceptron (MLP). RESULTS: The best AUC of 0.97 (1.0 and 0.97 for sensitivity and specificity) was observed for the combination of ADC, ADC of the peritumoral edema, T1, T1c, Sub and FLAIR-derived features using only 16 of the 10,914 possible features and XGBoost. CONCLUSIONS: Machine learning using radiomics features derived from multi-parametric MRI is capable of high AUC scores with high sensitivity and specificity in classifying meningiomas between low and higher gradings despite heterogeneous protocols across different centers. Feature selection can be performed effectively even when extracting a large amount of data for radiomics fingerprinting.

Image Thresholding Improves 3-Dimensional Convolutional Neural Network Diagnosis of Different Acute Brain Hemorrhages on Computed Tomography Scans.

Intracranial hemorrhage is a medical emergency that requires urgent diagnosis and immediate treatment to improve patient outcome. Machine learning algorithms can be used to perform medical image classification and assist clinicians in diagnosing radiological scans. In this paper, we apply 3-dimensional convolutional neural networks (3D CNN) to classify computed tomography (CT) brain scans into normal scans (N) and abnormal scans containing subarachnoid hemorrhage (SAH), intraparenchymal hemorrhage (IPH), acute subdural hemorrhage (ASDH) and brain polytrauma hemorrhage (BPH). The dataset used consists of 399 volumetric CT brain images representing approximately 12,000 images from the National Neuroscience Institute, Singapore. We used a 3D CNN to perform both 2-class (normal versus a specific abnormal class) and 4-class classification (between normal, SAH, IPH, ASDH). We apply image thresholding at the image pre-processing step, that improves 3D CNN classification accuracy and performance by accentuating the pixel intensities that contribute most to feature discrimination. For 2-class classification, the F1 scores for various pairs of medical diagnoses ranged from 0.706 to 0.902 without thresholding. With thresholding implemented, the F1 scores improved and ranged from 0.919 to 0.952. Our results are comparable to, and in some cases, exceed the results published in other work applying 3D CNN to CT or magnetic resonance imaging (MRI) brain scan classification. This work represents a direct application of a 3D CNN to a real hospital scenario involving a medically emergent CT brain diagnosis.

Artificial Intelligence and Radiology in Singapore: Championing a New Age of Augmented Imaging for Unsurpassed Patient Care.

Artificial intelligence (AI) has been positioned as being the most important recent advancement in radiology, if not the most potentially disruptive. Singapore radiologists have been quick to embrace this technology as part of the natural progression of the discipline toward a vision of how clinical medicine, empowered by technology, can achieve our national healthcare objectives of delivering value-based and patient-centric care. In this article, we consider 3 core questions relating to AI in radiology, and review the barriers to the widespread adoption of AI in radiology. We propose solutions and describe a "Centaur" model as a promising avenue for enabling the interfacing between AI and radiologists. Finally, we introduce The Radiological AI, Data Science and Imaging Informatics (RADII) subsection of the Singapore Radiological Society. RADII is an enabling body, which together with key technological and institutional stakeholders, will champion research, development and evaluation of AI for radiology applications.

Histogram Analysis Parameters Apparent Diffusion Coefficient for Distinguishing High and Low-Grade Meningiomas: A Multicenter Study.

Low grade meningiomas have better prognosis than high grade meningiomas. The aim of this study was to measure apparent diffusion coefficient (ADC) histogram analysis parameters in different meningiomas in a large multicenter sample and to analyze the possibility of several parameters for predicting tumor grade and proliferation potential. Overall, 148 meningiomas from 7 institutions were evaluated in this retrospective study. Grade 1 lesions were diagnosed in 101 (68.2%) cases, grade 2 in 41 (27.7%) patients, and grade 3 in 6 (4.1%) patients. All tumors were investigated by MRI (1.5 T scanner) by using diffusion weighted imaging (b values of 0 and 1000 s/mm2). For every lesion, the following parameters were calculated: mean ADC, maximum ADC, minimum ADC, median ADC, mode ADC, ADC percentiles P10, P25, P75, P90, kurtosis, skewness, and entropy. The comparison of ADC values was performed by Mann-Whitney-U test. Correlation between different ADC parameters and KI 67 was calculated by Spearman's rank correlation coefficient. Grade 2/3 meningiomas showed statistically significant lower ADC histogram analysis parameters in comparison to grade 1 tumors, especially ADC median. A threshold value of 0.82 for ADC median to predict tumor grade was estimated (sensitivity = 82.2%, specificity = 63.8%, accuracy = 76.4%, positive and negative predictive values were 83% and 62.5%, respectively). All ADC parameters except maximum ADC showed weak significant correlations with KI 67, especially ADC P25 (P = -.340, P = .0001).

Nipah Virus Infection.

Nipah virus, a paramyxovirus related to Hendra virus, first emerged in Malaysia in 1998. Clinical presentation ranges from asymptomatic infection to fatal encephalitis. Malaysia has had no more cases since 1999, but outbreaks continue to occur in Bangladesh and India. In the Malaysia-Singapore outbreak, transmission occurred primarily through contact with pigs, whereas in Bangladesh and India, it is associated with ingestion of contaminated date palm sap and human-to-human transmission. Bats are the main reservoir for this virus, which can cause disease in humans and animals. There are currently no effective therapeutics, and supportive care and prevention are the mainstays of management.

Progressive Supranuclear Gaze Palsy with Predominant Cerebellar Ataxia: A Case Series with Videos.

Progressive supranuclear palsy (PSP) with predominant cerebellar ataxia (PSP-C) is a rare phenotype of PSP. The clinical and radiological features of this disorder remain poorly characterized. Through a retrospective case series, we aim to characterize the clinical and radiological features of PSP-C. Four patients with PSP-C were identified: patients who presented with prominent cerebellar dysfunction that disappeared with the progression of the disease. Supranuclear gaze palsy occurred at a mean of 2.0 ± 2.3 years after the onset of ataxia. Mild cerebellar volume loss and midbrain atrophy were detected on brain imaging, which are supportive of a diagnosis of PSP. Videos are presented illustrating the co-existence of cerebellar signs and supranuclear gaze palsy and the disappearance of cerebellar signs with disease progression. Better recognition and the development of validated diagnostic criteria would aid in the antemortem recognition of this rare condition.