Brain and cord imaging features in neuromyelitis optica spectrum disorders

The new era of artificial intelligence in neuroradiology: current research and promising tools

Radiology has a number of characteristics that make it an especially suitable medical discipline for early artificial intelligence (AI) adoption. These include having a well-established digital workflow, standardized protocols for image storage, and numerous well-defined interpretive activities. The more than 200 commercial radiologic AI-based products recently approved by the Food and Drug Administration (FDA) to assist radiologists in a number of narrow image-analysis tasks such as image enhancement, workflow triage, and quantification, corroborate this observation. However, in order to leverage AI to boost efficacy and efficiency, and to overcome substantial obstacles to widespread successful clinical use of these products, radiologists should become familiarized with the emerging applications in their particular areas of expertise. In light of this, in this article we survey the existing literature on the application of AI-based techniques in neuroradiology, focusing on conditions such as vascular diseases, epilepsy, and demyelinating and neurodegenerative conditions. We also introduce some of the algorithms behind the applications, briefly discuss a few of the challenges of generalization in the use of AI models in neuroradiology, and skate over the most relevant commercially available solutions adopted in clinical practice. If well designed, AI algorithms have the potential to radically improve radiology, strengthening image analysis, enhancing the value of quantitative imaging techniques, and mitigating diagnostic errors.

Moving towards a new era for the treatment of neuromyelitis optica spectrum disorders

Neuromyelitis optica spectrum disorders (NMOSD) include a rare group of autoimmune conditions that primarily affect the central nervous system. They are characterized by inflammation and damage to the optic nerves, brain and spinal cord, leading to severe vision impairment, locomotor disability and sphynteric disturbances. In the majority of cases, NMOSD arises due to specific serum immunoglobulin G (IgG) autoantibodies targeting aquaporin 4 (AQP4-IgG), which is the most prevalent water-channel protein of the central nervous system. Early diagnosis and treatment are crucial to manage symptoms and prevent long-term disability in NMOSD patients. NMOSD were previously associated with a poor prognosis. However, recently, a number of randomized controlled trials have demonstrated that biological therapies acting on key elements of NMOSD pathogenesis, such as B cells, interleukin-6 (IL-6) pathway, and complement, have impressive efficacy in preventing the occurrence of clinical relapses. The approval of the initial drugs marks a revolutionary advancement in the treatment of NMOSD patients, significantly transforming therapeutic options and positively impacting their prognosis. In this review, we will provide an updated overview of the key immunopathological, clinical, laboratory, and neuroimaging aspects of NMOSD. Additionally, we will critically examine the latest advancements in NMOSD treatment approaches. Lastly, we will discuss key aspects regarding optimization of treatment strategies and their monitoring.

T1 hypointense brain lesions in NMOSD and its relevance with disability: a single institution cross-sectional study

BACKGROUND

T1 hypointense lesions are considered a surrogate marker of tissue destruction. Although there is a shortage of evidence about T1 hypointense brain lesions, black holes, in patients with Neuromyelitis Optica Spectrum Disorder (NMOSD), the clinical significance of these lesions is not well determined.

OBJECTIVES

The impact of T1 hypointense brain lesions on the clinical status and the disability level of patients with NMOSD was sought in this study.

METHODS

A total of 83 patients with the final diagnosis of NMOSD were recruited. Aquaporin-4 measures were collected. The expanded disability status scale (EDSS) and MRI studies were also extracted. T1 hypointense and T2/FLAIR hyperintense lesions were investigated. The correlation of MRI findings, AQP-4, and EDSS was assessed.

RESULTS

T1 hypointense brain lesions were detected in 22 patients. Mean ± SD EDSS was 3.7 ± 1.5 and significantly higher in patients with brain T1 hypointense lesions than those without them (p-value = 0.01). Noticeably, patients with more than four T1 hypointense lesions had EDSS scores ≥ 4. The presence of T2/FLAIR hyperintense brain lesions correlated with EDSS (3.6 ± 1.6 vs 2.3 ± 1.7; p-value = 0.01). EDSS was similar between those with and without positive AQP-4 (2.7 ± 1.6 vs. 3.2 ± 1.7; p-value = 0.17). Also, positive AQP-4 was not more prevalent in patients with T1 hypointense brain lesions than those without them (50.9 vs 45.4%; p-value = 0.8).

CONCLUSION

We demonstrated that the presence of the brain T1-hypointense lesions corresponds to a higher disability level in NMOSD.

Updates in NMOSD and MOGAD Diagnosis and Treatment: A Tale of Two Central Nervous System Autoimmune Inflammatory Disorders

Aquaporin-4-IgG positive neuromyelitis optica spectrum disorder (AQP4+NMOSD) and myelin-oligodendrocyte glycoprotein antibody-associated disease (MOGAD) are antibody-associated diseases targeting astrocytes and oligodendrocytes, respectively. Their recognition as distinct entities has led to each having its own diagnostic criteria that require a combination of clinical, serologic, and MRI features. The therapeutic approach to acute attacks in AQP4+NMOSD and MOGAD is similar. There is now class 1 evidence to support attack-prevention medications for AQP4+NMOSD. MOGAD lacks proven treatments although clinical trials are now underway. In this review, we will outline similarities and differences between AQP4+NMOSD and MOGAD in terms of diagnosis and treatment.

M-DDC: MRI based demyelinative diseases classification with U-Net segmentation and convolutional network

Childhood demyelinative diseases classification (DDC) with brain magnetic resonance imaging (MRI) is crucial to clinical diagnosis. But few attentions have been paid to DDC in the past. How to accurately differentiate pediatric-onset neuromyelitis optica spectrum disorder (NMOSD) from acute disseminated encephalomyelitis (ADEM) based on MRI is challenging in DDC. In this paper, a novel architecture M-DDC based on joint U-Net segmentation network and deep convolutional network is developed. The U-Net segmentation can provide pixel-level structure information, that helps the lesion areas location and size estimation. The classification branch in DDC can detect the regions of interest inside MRIs, including the white matter regions where lesions appear. The performance of the proposed method is evaluated on MRIs of 201 subjects recorded from the Children's Hospital of Zhejiang University School of Medicine. The comparisons show that the proposed DDC achieves the highest accuracy of 99.19% and dice of 71.1% for ADEM and NMOSD classification and segmentation, respectively.

Joint radiomics and spatial distribution model for MRI-based discrimination of multiple sclerosis, neuromyelitis optica spectrum disorder, and myelin-oligodendrocyte-glycoprotein-IgG-associated disorder

OBJECTIVE

To develop a discrimination pipeline concerning both radiomics and spatial distribution features of brain lesions for discrimination of multiple sclerosis (MS), aquaporin-4-IgG-seropositive neuromyelitis optica spectrum disorder (NMOSD), and myelin-oligodendrocyte-glycoprotein-IgG-associated disorder (MOGAD).

METHODS

Hyperintensity T2 lesions were delineated in 212 brain MRI scans of MS (n = 63), NMOSD (n = 87), and MOGAD (n = 45) patients. To avoid the effect of fixed training/test dataset sampling when developing machine learning models, patients were allocated into 4 sub-groups for cross-validation. For each scan, 351 radiomics and 27 spatial distribution features were extracted. Three models, i.e., multi-lesion radiomics, spatial distribution, and joint models, were constructed using random forest and logistic regression algorithms for differentiating: MS from the others (MS models) and MOGAD from NMOSD (MOG-NMO models), respectively. Then, the joint models were combined with demographic characteristics (i.e., age and sex) to create MS and MOG-NMO discriminators, respectively, based on which a three-disease discrimination pipeline was generated and compared with radiologists.

RESULTS

For classification of both MS-others and MOG-NMO, the joint models performed better than radiomics or spatial distribution model solely. The MS discriminator achieved AUC = 0.909 ± 0.027 and bias-corrected C-index = 0.909 ± 0.027, and the MOG-NMO discriminator achieved AUC = 0.880 ± 0.064 and bias-corrected C-index = 0.883 ± 0.068. The three-disease discrimination pipeline differentiated MS, NMOSD, and MOGAD patients with 75.0% accuracy, prominently outperforming the three radiologists (47.6%, 56.6%, and 66.0%).

CONCLUSIONS

The proposed pipeline integrating multi-lesion radiomics and spatial distribution features could effectively differentiate MS, NMOSD, and MOGAD.

CLINICAL RELEVANCE STATEMENT

The discrimination pipeline merging both radiomics and spatial distribution features of brain lesions may facilitate the differential diagnoses of multiple sclerosis, neuromyelitis optica spectrum disorder, and myelin-oligodendrocyte-glycoprotein-IgG-associated disorder.

KEY POINTS

• Our study introduces an approach by combining radiomics and spatial distribution models. • The joint model exhibited superior performance in distinguishing multiple sclerosis from aquaporin-4-IgG-seropositive neuromyelitis optica spectrum disorder and myelin-oligodendrocyte-glycoprotein-IgG-associated disorder as well as discriminating the latter two diseases. • The three-disease discrimination pipeline showcased remarkable accuracy, surpassing the performance of experienced radiologists, highlighting its potential as a valuable diagnostic tool.

Understanding the Pathophysiology and Magnetic Resonance Imaging of Multiple Sclerosis and Neuromyelitis Optica Spectrum Disorders

Magnetic resonance imaging (MRI) has been extensively applied in the study of multiple sclerosis (MS), substantially contributing to diagnosis, differential diagnosis, and disease monitoring. MRI studies have significantly contributed to the understanding of MS through the characterization of typical radiological features and their clinical or prognostic implications using conventional MRI pulse sequences and further with the application of advanced imaging techniques sensitive to microstructural damage. Interpretation of results has often been validated by MRI-pathology studies. However, the application of MRI techniques in the study of neuromyelitis optica spectrum disorders (NMOSD) remains an emerging field, and MRI studies have focused on radiological correlates of NMOSD and its pathophysiology to aid in diagnosis, improve monitoring, and identify relevant prognostic factors. In this review, we discuss the main contributions of MRI to the understanding of MS and NMOSD, focusing on the most novel discoveries to clarify differences in the pathophysiology of focal inflammation initiation and perpetuation, involvement of normal-appearing tissue, potential entry routes of pathogenic elements into the CNS, and existence of primary or secondary mechanisms of neurodegeneration.

Case report: Hyponatremia is an initial presentation of Neuromyelitis optica spectrum disorder

OBJECTIVE

Neuromyelitis optica spectrum disorders (NMOSD) is often misdiagnosed or delayed because of the complex and diverse clinical manifestations, especially the atypical initial presentation. Hyponatremia can be an infrequently isolated initial presentation of NMOSD and is associated with hypothalamus involvement. Awareness of this mechanism will help clinicians to identify NMOSD early, treat it in time and improve the prognosis.

METHODS

We describe a 36-year-old woman who developed repeated hyponatremia and then experienced diplopia. Serum AQP4, MOG, MBP and GFAP antibody were detected, and NMOSD was finally diagnosed.

RESULTS

She responded well to high-dose glucocorticoids. Sequential treatment with mycophenolate mofetil (MMF) was prescribed. Two-month follow-up revealed full recovery. So far, after 10 months, the patient still has no recurrence.

CONCLUSION

For young patients, repeated hyponatremia, with or without slight fever, and no evidence of obvious infection, brain magnetic resonance imaging (MRI) and serum AQP4/MOG antibody detection may be useful to determine whether there is a possibility of NMOSD.

MRI to differentiate multiple sclerosis, neuromyelitis optica, and myelin oligodendrocyte glycoprotein antibody disease

Differentiating multiple sclerosis (MS) from other relapsing inflammatory autoimmune diseases of the central nervous system such as neuromyelitis optica spectrum disorder (NMOSD) and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD) is crucial in clinical practice. The differential diagnosis may be challenging but making the correct ultimate diagnosis is critical, since prognosis and treatments differ, and inappropriate therapy may promote disability. In the last two decades, significant advances have been made in MS, NMOSD, and MOGAD including new diagnostic criteria with better characterization of typical clinical symptoms and suggestive imaging (magnetic resonance imaging [MRI]) lesions. MRI is invaluable in making the ultimate diagnosis. An increasing amount of new evidence with respect to the specificity of observed lesions as well as the associated dynamic changes in the acute and follow-up phase in each condition has been reported in distinct studies recently published. Additionally, differences in brain (including the optic nerve) and spinal cord lesion patterns between MS, aquaporin4-antibody-positive NMOSD, and MOGAD have been described. We therefore present a narrative review on the most relevant findings in brain, spinal cord, and optic nerve lesions on conventional MRI for distinguishing adult patients with MS from NMOSD and MOGAD in clinical practice. In this context, cortical and central vein sign lesions, brain and spinal cord lesions characteristic of MS, NMOSD, and MOGAD, optic nerve involvement, role of MRI at follow-up, and new proposed diagnostic criteria to differentiate MS from NMOSD and MOGAD were discussed.

Frequency of new asymptomatic MRI lesions during attacks and follow-up of patients with NMOSD in a real-world setting

BACKGROUND

We aimed to assess the frequency of new asymptomatic lesions on brain and spinal imaging (magnetic resonance imaging (MRI)) and their association with subsequent relapses in a large cohort of neuromyelitis optica spectrum disorder (NMOSD) patients in Argentina.

METHODS

We retrospectively reviewed 675 MRI (225 performed during an attack and 450 during the relapse-free period (performed at least 3 months from the last attack)) of NMOSD patients who had at least 2 years of clinical and MRI follow-up since disease onset. Kaplan-Meier (KM) curves were used for depicting time from remission MRI to subsequent relapse.

RESULTS

We included 135 NMOSD patients (64.4% were aquaporin-4-immunoglobulin G (AQP4-IgG)-positive). We found that 26 (19.26%) and 66 (48.88%) of patients experienced at least one new asymptomatic MRI lesion during both the relapse-free period and attacks, respectively. The most frequent asymptomatic MRI lesions were optic nerves followed by short-segment myelitis during the relapse-free period and attacks. KM curves did not show differences in the time taken to develop a new relapse.

CONCLUSION

Our findings showed that new asymptomatic lesions are relatively frequent. However, the presence of new asymptomatic MRI lesions during the relapse-free period and at relapses was not associated with a shorter time to developing subsequent relapses.

Differentiation between multiple sclerosis and neuromyelitis optica spectrum disorder using a deep learning model

Multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD) are autoimmune inflammatory disorders of the central nervous system (CNS) with similar characteristics. The differential diagnosis between MS and NMOSD is critical for initiating early effective therapy. In this study, we developed a deep learning model to differentiate between multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD) using brain magnetic resonance imaging (MRI) data. The model was based on a modified ResNet18 convolution neural network trained with 5-channel images created by selecting five 2D slices of 3D FLAIR images. The accuracy of the model was 76.1%, with a sensitivity of 77.3% and a specificity of 74.8%. Positive and negative predictive values were 76.9% and 78.6%, respectively, with an area under the curve of 0.85. Application of Grad-CAM to the model revealed that white matter lesions were the major classifier. This compact model may aid in the differential diagnosis of MS and NMOSD in clinical practice.

The confound of hemodynamic response function variability in human resting-state functional MRI studies

Functional magnetic resonance imaging (fMRI) is an indirect measure of neural activity with the hemodynamic response function (HRF) coupling it with unmeasured neural activity. The HRF, modulated by several non-neural factors, is variable across brain regions, individuals and populations. Yet, a majority of human resting-state fMRI connectivity studies continue to assume a non-variable HRF. In this article, with supportive prior evidence, we argue that HRF variability cannot be ignored as it substantially confounds within-subject connectivity estimates and between-subjects connectivity group differences. We also discuss its clinical relevance with connectivity impairments confounded by HRF aberrations in several disorders. We present limited data on HRF differences between women and men, which resulted in a 15.4% median error in functional connectivity estimates in a group-level comparison. We also discuss the implications of HRF variability for fMRI studies in the spinal cord. There is a need for more dialogue within the community on the HRF confound, and we hope that our article is a catalyst in the process.

Comparing clinical and imaging features of patients with MOG antibody-positivity and with and without oligoclonal bands

Introduction

Myelin-oligodendrocyte glycoprotein antibody (MOG)-associated disorder (MOGAD) is a recently identified immune-mediated inflammatory disorder of the central nervous system (CNS). The significance of oligoclonal bands (OCBs) is not fully elucidated. This study investigated the clinical differences between patients with MOGAD who tested positive or negative for OCBs.

Methods

The study was conducted on 23 patients with MOG-IgG-seropositivity who presented with central nervous system (CNS) symptoms. The patients were screened and divided into OCB-positive (n=10) and OCB-negative (n=13) groups, and their demographic, clinical, and magnetic resonance imaging (MRI) features were compared.

Results

The results revealed that patients with OCB-positivity had a significantly higher frequency of relapse, and their IgG index was significantly higher.

Discussion

OCBs were common in MOGAD met the consensus criteria. The study concluded that careful treatment decision-making is necessary in MOG antibody-positive cases with OCB-positivity.

Can early-onset acquired demyelinating syndrome (ADS) hide pediatric Behcet's disease? A case report

Behcet's disease (BD) is a rare vasculitis characterized by multisystemic inflammation. Central nervous system (CNS) involvement is rare and heterogeneous, particularly in the pediatric population. A diagnosis of neuro-Behcet could be highly challenging, especially if neurological manifestations precede other systemic features; however, its timely definition is crucial to prevent long-term sequelae. In this study, we describe the case of a girl who, at 13 months of age, presented with a first episode of encephalopathy compatible with acute disseminated encephalomyelitis, followed, after 6 months, by a neurological relapse characterized by ophthalmoparesis and gait ataxia, in association with new inflammatory lesions in the brain and spinal cord, suggesting a neuromyelitis optica spectrum disorder. The neurological manifestations were successfully treated with high-dose steroids and intravenous immunoglobulins. In the following months, the patient developed a multisystemic involvement suggestive of Behcet's disease, characterized by polyarthritis and uveitis, associated with HLA-B51 positivity. The challenge presented by this unique case required a multidisciplinary approach involving pediatric neurologists, neuro-radiologists, and pediatric rheumatologists, with all of these specialists creating awareness about early-onset acquired demyelinating syndromes (ADSs). Given the rarity of this presentation, we performed a review of the literature focusing on neurological manifestations in BD and differential diagnosis of patients with early-onset ADS.

Tumefactive Demyelination in MOG Ab-Associated Disease, Multiple Sclerosis, and AQP-4-IgG-Positive Neuromyelitis Optica Spectrum Disorder

BACKGROUND AND OBJECTIVES

Studies on tumefactive brain lesions in myelin oligodendrocyte glycoprotein-immunoglobulin G (IgG)-associated disease (MOGAD) are lacking. We sought to characterize the frequency clinical, laboratory, and MRI features of these lesions in MOGAD and compare them with those in multiple sclerosis (MS) and aquaporin-4-IgG-positive neuromyelitis optica spectrum disorder (AQP4+NMOSD).

METHODS

We retrospectively searched 194 patients with MOGAD and 359 patients with AQP4+NMOSD with clinical/MRI details available from the Mayo Clinic databases and included those with ≥1 tumefactive brain lesion (maximum transverse diameter ≥2 cm) on MRI. Patients with tumefactive MS were identified using the Mayo Clinic medical record linkage system. Binary multivariable stepwise logistic regression identified independent predictors of MOGAD diagnosis; Cox proportional regression models were used to assess the risk of relapsing disease and gait aid in patients with tumefactive MOGAD vs those with nontumefactive MOGAD.

RESULTS

We included 108 patients with tumefactive demyelination (MOGAD = 43; AQP4+NMOSD = 16; and MS = 49). Tumefactive lesions were more frequent among those with MOGAD (43/194 [22%]) than among those with AQP4+NMOSD (16/359 [5%], p < 0.001). Risk of relapse and need for gait aid were similar in tumefactive and nontumefactive MOGAD. Clinical features more frequent in MOGAD than in MS included headache (18/43 [42%] vs 10/49 [20%]; p = 0.03) and somnolence (12/43 [28%] vs 2/49 [4%]; p = 0.003), the latter also more frequent than in AQP4+NMOSD (0/16 [0%]; p = 0.02). The presence of peripheral T2-hypointense rim, T1-hypointensity, diffusion restriction (particularly an arc pattern), ring enhancement, and Baló-like or cystic appearance favored MS over MOGAD (p ≤ 0.001). MRI features were broadly similar in MOGAD and AQP4+NMOSD, except for more frequent diffusion restriction in AQP4+NMOSD (10/15 [67%]) than in MOGAD (11/42 [26%], p = 0.005). CSF analysis revealed less frequent positive oligoclonal bands in MOGAD (2/37 [5%]) than in MS (30/43 [70%], p < 0.001) and higher median white cell count in MOGAD than in MS (33 vs 6 cells/μL, p < 0.001). At baseline, independent predictors of MOGAD diagnosis were the presence of somnolence/headache, absence of T2-hypointense rim, lack of T1-hypointensity, and no diffusion restriction (Nagelkerke R 2 = 0.67). Tumefactive lesion resolution was more common in MOGAD than in MS or AQP4+NMOSD and improved model performance.

DISCUSSION

Tumefactive lesions are frequent in MOGAD but not associated with a worse prognosis. The clinical, MRI, and CSF attributes of tumefactive MOGAD differ from those of tumefactive MS and are more similar to those of tumefactive AQP4+NMOSD with the exception of lesion resolution, which favors MOGAD.

Barriers to access and unmet needs to neuromyelitis optica spectrum disorders care in an Argentinean cohort

INTRODUCTION

Neuromyelitis optica spectrum disorder (NMOSD) is a rare but severe neuroimmunological condition associated with a significant financial burden. NMOSD is also associated with increased health care utilization, including neurology outpatient visits, magnetic resonance imaging (MRI) use, long-term medication, among others. We aimed to evaluate real-world patient experiences in access to care and NMOSD burden in an Argentinean cohort.

METHODS

This cross-sectional study used a self-administered survey and was conducted in Argentina (2022). Patients with NMOSD were divided into three groups: private health insurance (PHI), social health insurance (SHI), and public health insurance (PHI, Ministry of Public Health). Differences in access and health care barriers were assessed.

RESULTS

One hundred patients with NMOSD (74 women) with a mean age at diagnosis of 38.7 years were included. Their EDSS was 2.8 and they were followed for 5.2 years. Of them, 51%, 11%, and 13% were employed (full-time: 57.5%), currently unemployed and retired by NMOSD, respectively. 55% of them visited between 2-3 specialists before NMOSD diagnosis. Aquaporin-4-antibody and/or myelin oligodendrocyte glycoprotein-antibody testing was requested in 91% (health insurance covered this partially in 15.3% and 32.9% of the time the test was entirely paid by patient/family). Patients with NMOSD receiving private medical care reported greater access to MRI, outpatient visits, and fewer issues to obtain NMOSD medications compared to those treated at public institutions. A longer mean time to MRI and neurology visit was found in the PHI group when compared with the other two subgroups. Regression analysis showed that private insurance (OR=3.84, p=0.01) was the only independent factor associated with appropriate access to NMOSD medications in Argentina.

CONCLUSION

These findings suggest that barriers to access and utilization of NMOSD care services in Argentina are common. NMOSD patients experienced problems to receive NMOSD medication properly, especially those from the public sector.

Frequency of NMOSD misdiagnosis in a cohort from Latin America: Impact and evaluation of different contributors

BACKGROUND

Neuromyelitis optica spectrum disorder (NMOSD) misdiagnosis (i.e. the incorrect diagnosis of patients who truly have NMOSD) remains an issue in clinical practice. We determined the frequency and factors associated with NMOSD misdiagnosis in patients evaluated in a cohort from Latin America.

METHODS

We retrospectively reviewed the medical records of patients with NMOSD, according to the 2015 diagnostic criteria, from referral clinics in six Latin American countries (Argentina, Chile, Paraguay, Colombia, Ecuador, and Venezuela). Diagnoses prior to NMOSD and ultimate diagnoses, demographic, clinical and paraclinical data, and treatment schemes were evaluated.

RESULTS

A total of 469 patients presented with an established diagnosis of NMOSD (73.2% seropositive) and after evaluation, we determined that 56 (12%) patients had been initially misdiagnosed with a disease other than NMOSD. The most frequent alternative diagnoses were multiple sclerosis (MS; 66.1%), clinically isolated syndrome (17.9%), and cerebrovascular disease (3.6%). NMOSD misdiagnosis was determined by MS/NMOSD specialists in 33.9% of cases. An atypical MS syndrome was found in 86% of misdiagnosed patients, 50% had NMOSD red flags in brain and/or spinal magnetic resonance imaging (MRI), and 71.5% were prescribed disease-modifying drugs.

CONCLUSIONS

NMOSD misdiagnosis is relatively frequent in Latin America (12%). Misapplication and misinterpretation of clinical and neuroradiological findings are relevant factors associated with misdiagnosis.

Neuroimaging features in inflammatory myelopathies: A review

Spinal cord involvement can be observed in the course of immune-mediated disorders. Although multiple sclerosis (MS) represents the leading cause of inflammatory myelopathy, an increasing number of alternative etiologies must be now considered in the diagnostic work-up of patients presenting with myelitis. These include antibody-mediated disorders and cytotoxic T cell-mediated diseases targeting central nervous system (CNS) antigens, and systemic autoimmune conditions with secondary CNS involvement. Even though clinical features are helpful to orient the diagnostic suspicion (e.g., timing and severity of myelopathy symptoms), the differential diagnosis of inflammatory myelopathies is often challenging due to overlapping features. Moreover, noninflammatory etiologies can sometimes mimic an inflammatory process. In this setting, magnetic resonance imaging (MRI) is becoming a fundamental tool for the characterization of spinal cord damage, revealing a pictorial scenario which is wider than the clinical manifestations. The characterization of spinal cord lesions in terms of longitudinal extension, location on axial plane, involvement of the white matter and/or gray matter, and specific patterns of contrast enhancement, often allows a proper differentiation of these diseases. For instance, besides classical features, such as the presence of longitudinally extensive spinal cord lesions in patients with aquaporin-4-IgG positive neuromyelitis optica spectrum disorder (AQP4+NMOSD), novel radiological signs (e.g., H sign, trident sign) have been recently proposed and successfully applied for the differential diagnosis of inflammatory myelopathies. In this review article, we will discuss the radiological features of spinal cord involvement in autoimmune disorders such as MS, AQP4+NMOSD, myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD), and other recently characterized immune-mediated diseases. The identification of imaging pitfalls and mimics that can lead to misdiagnosis will also be examined. Since spinal cord damage is a major cause of irreversible clinical disability, the recognition of these radiological aspects will help clinicians achieve a correct and prompt diagnosis, treat early with disease-specific treatment and improve patient outcomes.

Neuromyelitis Optica Spectrum Disorders

PURPOSE OF REVIEW

This article reviews the cardinal clinical features, distinct immunopathology, current diagnostic criteria, relapse-related risk factors, emerging biomarkers, and evolving treatment strategies pertaining to neuromyelitis optica spectrum disorders (NMOSD).

RECENT FINDINGS

The discovery of the pathogenic aquaporin-4 (AQP4)-IgG autoantibody and characterization of NMOSD as an autoimmune astrocytopathy have spearheaded the identification of key immunologic therapeutic targets in this disease, including but not limited to the complement system, the interleukin 6 (IL-6) receptor, and B cells. Accordingly, four recent randomized controlled trials have demonstrated the efficacy of three new NMOSD therapies, namely eculizumab, satralizumab, and inebilizumab.

SUMMARY

Currently, NMOSD poses both diagnostic and treatment challenges. It is debated whether individuals who are seropositive for myelin oligodendrocyte glycoprotein (MOG)-IgG belong within the neuromyelitis optica spectrum. This discussion is fueled by disparities in treatment responses between patients who are AQP4-IgG seropositive and seronegative, suggesting different immunopathologic mechanisms may govern these conditions. As our understanding regarding the immune pathophysiology of NMOSD expands, emerging biomarkers, including serum neurofilament light chain and glial fibrillary acidic protein (GFAP), may facilitate earlier relapse detection and inform long-term treatment decisions. Future research focal points should include strategies to optimize relapse management, restorative treatments that augment neurologic recovery, and practical solutions that promote equitable access to approved therapies for all patients with NMOSD.

Myelin Oligodendrocyte Glycoprotein Antibody-Associated Disease (MOGAD): A Review of Clinical and MRI Features, Diagnosis, and Management

Myelin oligodendrocyte glycoprotein (MOG) antibody-associated disease (MOGAD) is the most recently defined inflammatory demyelinating disease of the central nervous system (CNS). Over the last decade, several studies have helped delineate the characteristic clinical-MRI phenotypes of the disease, allowing distinction from aquaporin-4 (AQP4)-IgG-positive neuromyelitis optica spectrum disorder (AQP4-IgG+NMOSD) and multiple sclerosis (MS). The clinical manifestations of MOGAD are heterogeneous, ranging from isolated optic neuritis or myelitis to multifocal CNS demyelination often in the form of acute disseminated encephalomyelitis (ADEM), or cortical encephalitis. A relapsing course is observed in approximately 50% of patients. Characteristic MRI features have been described that increase the diagnostic suspicion (e.g., perineural optic nerve enhancement, spinal cord H-sign, T2-lesion resolution over time) and help discriminate from MS and AQP4+NMOSD, despite some overlap. The detection of MOG-IgG in the serum (and sometimes CSF) confirms the diagnosis in patients with compatible clinical-MRI phenotypes, but false positive results are occasionally encountered, especially with indiscriminate testing of large unselected populations. The type of cell-based assay used to evaluate for MOG-IgG (fixed vs. live) and antibody end-titer (low vs. high) can influence the likelihood of MOGAD diagnosis. International consensus diagnostic criteria for MOGAD are currently being compiled and will assist in clinical diagnosis and be useful for enrolment in clinical trials. Although randomized controlled trials are lacking, MOGAD acute attacks appear to be very responsive to high dose steroids and plasma exchange may be considered in refractory cases. Attack-prevention treatments also lack class-I data and empiric maintenance treatment is generally reserved for relapsing cases or patients with severe residual disability after the presenting attack. A variety of empiric steroid-sparing immunosuppressants can be considered and may be efficacious based on retrospective or prospective observational studies but prospective randomized placebo-controlled trials are needed to better guide treatment. In summary, this article will review our rapidly evolving understanding of MOGAD diagnosis and management.

Role of artificial intelligence in MS clinical practice

•

For medical applications, machine learning (including deep learning) are the most commonly used artificial intelligence (AI) approaches.

•

It can improve multiple sclerosis (MS) diagnosis, prognostication and treatment monitoring.

•

Thanks to AI, MRI and cognitive phenotypes of MS patients were identified.

•

AI can shorten MRI protocols for MS, allowing the application of advanced techniques.

•

It can reduce the human effort for MRI analysis, especially for lesion segmentation.

Machine learning (ML) and its subset, deep learning (DL), are branches of artificial intelligence (AI) showing promising findings in the medical field, especially when applied to imaging data. Given the substantial role of MRI in the diagnosis and management of patients with multiple sclerosis (MS), this disease is an ideal candidate for the application of AI techniques. In this narrative review, we are going to discuss the potential applications of AI for MS clinical practice, together with their limitations. Among their several advantages, ML algorithms are able to automate repetitive tasks, to analyze more data in less time and to achieve higher accuracy and reproducibility than the human counterpart. To date, these algorithms have been applied to MS diagnosis, prognosis, disease and treatment monitoring. Other fields of application have been improvement of MRI protocols as well as automated lesion and tissue segmentation. However, several challenges remain, including a better understanding of the information selected by AI algorithms, appropriate multicenter and longitudinal validations of results and practical aspects regarding hardware and software integration. Finally, one cannot overemphasize the paramount importance of human supervision, in order to optimize the use and take full advantage of the potential of AI approaches.

A Rare Case of Neuromyelitis Optica Spectrum Disorders With Unknown Fever and Subacute Cognitive Decline With Normal Images

We report the case of a 69-year-old Japanese man who came to our hospital with a chief complaint of fever and cognitive decline for three weeks. There were no neurological abnormalities other than the decreased level of consciousness. He developed urinary retention after admission, so we performed a lumbar puncture, although his head and neck magnetic resonance imaging (MRI) showed no abnormal findings. The cerebrospinal fluid (CSF) examination showed albuminocytologic dissociation and the anti-aquaporin 4 antibody was positive. Thus, we diagnosed him with neuromyelitis optica spectrum disorder (NMOSD). NMOSD is an autoimmune disease that causes demyelination. The clue to diagnosing NMOSD is demyelinating findings on MRI. Therefore, it is difficult to diagnose NMOSD if there are no abnormalities on the images. However, abnormal MRI findings are not necessary for the diagnosis of NMOSD. Thus, NMOSD cannot be ruled out even if MRI findings are normal and the real clue to diagnosing NMOSD is the anti-aquaporin 4 antibody.

Towards imaging criteria that best differentiate MS from NMOSD and MOGAD: Large multi-ethnic population and different clinical scenarios

BACKGROUND

The "1/3″ brain magnetic resonance imaging (MRI) criteria including 1) a lesion adjacent to the lateral ventricle and in the inferior temporal lobe, or 2) a juxtacortical lesion, or 3) a Dawson finger-type lesion were shown to distinguish multiple sclerosis (MS) from antibody-mediated conditions. In this large multicentre study, we aimed to assess how the criteria perform 1) in different onset phenotypes, 2) distinct ethnic groups, 3) when the absence of myelin oligodendrocyte glycoprotein antibody (MOG-Ab)-associated disease (MOGAD)-typical fluffy infratentorial (FIT) lesions and longitudinally extensive transverse myelitis (LETM) lesions are added as features ("2/4″ and 3/5″ criteria, respectively).

METHODS

577 patients with MS (n = 332), aquaporin-4 antibody (AQP4-Ab) neuromyelitis optica spectrum disorder (NMOSD) (n = 196) and MOGAD (n = 49) were recruited from 6 international centres (Buenos Aires, Sao Paolo, Maracaibo, Goyang, Oxford and Milan). Imaging scans were obtained at disease onset or relapse.

RESULTS

Adding the absence of FIT lesions increased the specificity of the "1/3″ criteria vs. AQP4-Ab NMOSD from 84.7% to 87.2% and vs. MOGAD from 85.7% to 93.9% without compromising their sensitivity (86%). In particular, for those presenting with brain/brainstem attacks "2/4″ had significantly higher specificity than "1/3″ (85% vs. 80% against AQP4-Ab NMOSD, 88.9% vs. 72.2% against MOGAD). Positive predictive values of the "1/3″ criteria for MS were lowest for Asian patients (84.8 vs. 99.1% for White) but were significantly increased by adding further criteria (94.1% for "3/5″).

CONCLUSION

The "1/3″ criteria perform well in discriminating MS from NMOSD and MOGAD regardless of ethnic background and clinical scenario. Adding the absence of FIT lesions increases the specificity in those presenting with brain/brainstem symptoms.

Elucidating distinct clinico-radiologic signatures in the borderland between neuromyelitis optica and multiple sclerosis

BACKGROUND

Separating antibody-negative neuromyelitis optica spectrum disorders (NMOSD) from multiple sclerosis (MS) in borderline cases is extremely challenging due to lack of biomarkers. Elucidating different pathologies within the likely heterogenous antibody-negative NMOSD/MS overlap syndrome is, therefore, a major unmet need which would help avoid disability from inappropriate treatment.

OBJECTIVE

In this study we aimed to identify distinct subgroups within the antibody-negative NMOSD/MS overlap syndrome.

METHODS

Twenty-five relapsing antibody-negative patients with NMOSD features underwent a prospective brain and spinal cord MRI. Subgroups were identified by an unsupervised algorithm based on pre-selected NMOSD/MS discriminators.

RESULTS

Four subgroups were identified. Patients from Group 1 termed "MS-like" (n = 6) often had central vein sign and cortical lesions (83% and 67%, respectively). All patients from Group 2 ("spinal MS-like", 8) had short-segment myelitis and no MS-like brain lesions. Group 3 ("classic NMO-like", 6) had high percentage of bilateral optic neuritis and longitudinally extensive transverse myelitis (LETM, 80% and 60%, respectively) and normal brain appearance (100%). Group 4 ("NMO-like with brain involvement", 5) typically had a history of NMOSD-like brain lesions and LETM. When compared with other groups, Group 4 had significantly decreased fractional anisotropy in non-lesioned tracts (0.46 vs. 0.49, p = 0.003) and decreased thalamus volume (0.84 vs. 0.98, p = 0.04).

CONCLUSIONS

NMOSD/MS cohort contains distinct subgroups likely corresponding to different pathologies and requiring tailored treatment. We propose that non-conventional MRI might help optimise diagnosis in these challenging patients.

Imaging features to distinguish AQP4-positive NMOSD and MS at disease onset: A retrospective analysis in a single-center cohort

PURPOSE

To compare the diagnostic performance of imaging criteria that differentiate AQP4⁺ Neuromyelitis Optica Spectrum Disorders (NMOSD) and Multiple Sclerosis (MS) at disease onset (DO) and follow-up (FU).

METHODS

We retrospectively analyzed MRI scans at DO (defined as the first 60 days of patient-reported symptom onset) in 10 AQP4⁺NMOSD and 25 (time to MRI matched) relapsing-remitting MS patients from a monocentric cohort.

RESULTS

The Matthews criteria were met in 20% of AQP4⁺NMOSD patients at DO vs. 33% at FU, and in 96% of RRMS patients vs.100% at FU. Specificity (SP) and sensitivity (SE) were thus high at both time-points: SP-DO: 96%; SP-FU:100%; and SE-DO: 80%; SE-FU: 67%, with similar area under the curve (AUC) values at DO: 88% [95% CI 74%-100%] and FU: 83% [95% CI 67%-100%]. The Cacciaguerra criteria were met in 90% of AQP4⁺NMOSD patients at DO vs. 88.9% at FU and in 24% of RRMS patients vs. 14% at FU; SP-DO: 87%; SP-FU: 86%; and SE-DO: 90%; SE-FU: 89%, with similar AUC values at DO: 88% [95% CI 76%-98%] and FU: 87% [95% CI 74%-98%].

CONCLUSIONS

While diagnostic MRI criteria were developed on data acquired years after disease onset, our study demonstrates their high applicability at the earliest disease stages, thus emphasising their valuable use in clinical practice.

MRI Patterns Distinguish AQP4 Antibody Positive Neuromyelitis Optica Spectrum Disorder From Multiple Sclerosis

Neuromyelitis optica spectrum disorder (NMOSD) and multiple sclerosis (MS) are inflammatory diseases of the CNS. Overlap in the clinical and MRI features of NMOSD and MS means that distinguishing these conditions can be difficult. With the aim of evaluating the diagnostic utility of MRI features in distinguishing NMOSD from MS, we have conducted a cross-sectional analysis of imaging data and developed predictive models to distinguish the two conditions. NMOSD and MS MRI lesions were identified and defined through a literature search. Aquaporin-4 (AQP4) antibody positive NMOSD cases and age- and sex-matched MS cases were collected. MRI of orbits, brain and spine were reported by at least two blinded reviewers. MRI brain or spine was available for 166/168 (99%) of cases. Longitudinally extensive (OR = 203), "bright spotty" (OR = 93.8), whole (axial; OR = 57.8) or gadolinium (Gd) enhancing (OR = 28.6) spinal cord lesions, bilateral (OR = 31.3) or Gd-enhancing (OR = 15.4) optic nerve lesions, and nucleus tractus solitarius (OR = 19.2), periaqueductal (OR = 16.8) or hypothalamic (OR = 7.2) brain lesions were associated with NMOSD. Ovoid (OR = 0.029), Dawson's fingers (OR = 0.031), pyramidal corpus callosum (OR = 0.058), periventricular (OR = 0.136), temporal lobe (OR = 0.137) and T1 black holes (OR = 0.154) brain lesions were associated with MS. A score-based algorithm and a decision tree determined by machine learning accurately predicted more than 85% of both diagnoses using first available imaging alone. We have confirmed NMOSD and MS specific MRI features and combined these in predictive models that can accurately identify more than 85% of cases as either AQP4 seropositive NMOSD or MS.

Retinal optical coherence tomography and magnetic resonance imaging in neuromyelitis optica spectrum disorders and MOG-antibody associated disorders: an updated review

INTRODUCTION

Neuromyelitis optica spectrum disorders (NMOSD) and myelin oligodendrocyte glycoprotein IgG antibody-associated disorders (MOGAD) comprise two groups of rare neuroinflammatory diseases that cause attack-related damage to the central nervous system (CNS). Clinical attacks are often characterized by optic neuritis, transverse myelitis, and to a lesser extent, brainstem encephalitis/area postrema syndrome. Retinal optical coherence tomography (OCT) is a non-invasive technique that allows for in vivo thickness quantification of the retinal layers. Apart from OCT, magnetic resonance imaging (MRI) plays an increasingly important role in NMOSD and MOGAD diagnosis based on the current international diagnostic criteria. Retinal OCT and brain/spinal cord/optic nerve MRI can help to distinguish NMOSD and MOGAD from other neuroinflammatory diseases, particularly from multiple sclerosis, and to monitor disease-associated CNS-damage.

AREAS COVERED

This article summarizes the current status of imaging research in NMOSD and MOGAD, and reviews the clinical relevance of OCT, MRI and other relevant imaging techniques for differential diagnosis, screening and monitoring of the disease course.

EXPERT OPINION

Retinal OCT and MRI can visualize and quantify CNS damage in vivo, improving our understanding of NMOSD and MOGAD pathology. Further efforts on the standardization of these imaging techniques are essential for implementation into clinical practice and as outcome parameters in clinical trials.

Comparative analysis of clinical and imaging data between patients with myelin oligodendrocyte glycoprotein antibody disease and patients with aquaporin 4 antibody-positive neuromyelitis optica spectrum disorder

BACKGROUND

We aimed to compare the clinical data, laboratory findings, and imaging characteristics of myelin oligodendrocyte glycoprotein antibody disease (MOGAD) and aquaporin 4 antibody (AQP4)-positive neuromyelitis optica spectrum disorder (NMOSD), as detailed comparative analyses of laboratory data for both diseases are rare.

METHODS

Our retrospective study compared the clinical data, laboratory findings, and imaging characteristics of 118 AQP4-positive patients with first-episode NMOSD and 25 patients with first-episode MOGAD. Logistic regression was used to determine the factors that differentiated MOGAD and AQP4-positive NMOSD.

RESULTS

There were significant differences in age, symptoms, recurrence rate, laboratory indicators, and imaging examinations between patients with MOGAD and patients with AQP4-positive NMOSD. Patients with MOGAD were younger and had higher levels of uric acid than those with AQP4-positive NMOSD. The proportion of cortical gray matter/juxtacortical white matter lesions was significantly higher in the MOGAD group than in the NMOSD group. Logistic regression revealed that young age [odds ratio (OR) = 0.947, 95% confidence interval (CI) = 0.905-0.99], high uric acid level (OR = 1.016, 95% CI = 1.006-1.027), and cortical gray matter/juxtacortical white matter involvement (OR = 3.889, 95% CI = 1.048-14.442) were significantly related to MOGAD.

CONCLUSION

The multivariate analysis of the present study demonstrated that age, uric acid level, and the presence of lesions in the cortical gray matter/juxtacortical white matter can aid in distinguishing patients with AQP4-positive NMOSD from those with MOGAD. These factors may also aid in determining which patients should be tested for antibodies.

Comparisons of clinical phenotype, radiological and laboratory features, and therapy of neuromyelitis optica spectrum disorder by regions: update and challenges

Neuromyelitis optica spectrum disorder (NMOSD) is an inflammatory demyelinating disease of the central nervous system (CNS) associated with autoantibody (ab) to aquaporin-4 (AQP4). There is obvious variation between regions and countries in the epidemiology, clinical features and management in NMOSD. Based on published population-based observation and cohort studies, the different clinical pattern of NMOSD has been seen in several geographical regions and some of these patients with NMOSD-like features do not fully meet the current diagnostic criteria, which is needed to consider the value of recently revised diagnostic criteria. At present, all treatments applied in NMOSD have made great progress, however, these treatments failed in AQP4 ab negative and refractory patients. Therefore, it is necessary to turn into an innovative idea and to open a new era of NMOSD treatment to develop novel and diverse targets and effective therapeutic drugs in NMOSD and to conduct the trails in large clinical samples and case-control studies to confirm their therapeutic effects on NMOSD in the future, which still remain a challenge.

Application of deep-learning to the seronegative side of the NMO spectrum

OBJECTIVES

To apply a deep-learning algorithm to brain MRIs of seronegative patients with neuromyelitis optica spectrum disorders (NMOSD) and NMOSD-like manifestations and assess whether their structural features are similar to aquaporin-4-seropositive NMOSD or multiple sclerosis (MS) patients.

PATIENTS AND METHODS

We analyzed 228 T2- and T1-weighted brain MRIs acquired from aquaporin-4-seropositive NMOSD (n = 85), MS (n = 95), aquaporin-4-seronegative NMOSD [n = 11, three with anti-myelin oligodendrocyte glycoprotein antibodies (MOG)], and aquaporin-4-seronegative patients with NMOSD-like manifestations (idiopathic recurrent optic neuritis and myelitis, n = 37), who were recruited from February 2010 to December 2019. Seventy-three percent of aquaporin-4-seronegative patients with NMOSD-like manifestations also had a clinical follow-up (median duration of 4 years). The deep-learning neural network architecture was based on four 3D convolutional layers. It was trained and validated on MRI scans of aquaporin-4-seropositive NMOSD and MS patients and was then applied to aquaporin-4-seronegative NMOSD and NMOSD-like manifestations. Assignment of unclassified aquaporin-4-seronegative patients was compared with their clinical follow-up.

RESULTS

The final algorithm differentiated aquaporin-4-seropositive NMOSD and MS patients with an accuracy of 0.95. All aquaporin-4-seronegative NMOSD and 36/37 aquaporin-4-seronegative patients with NMOSD-like manifestations were classified as NMOSD. Anti-MOG patients had a similar probability of being NMOSD or MS. At clinical follow-up, one unclassified aquaporin-4-seronegative patient evolved to MS, three developed NMOSD, and the others did not change phenotype.

CONCLUSIONS

Our findings support the inclusion of aquaporin4-seronegative patients into NMOSD and suggest a possible expansion to aquaporin-4-seronegative unclassified patients with NMOSD-like manifestations. Anti-MOG patients are likely to have intermediate brain features between NMOSD and MS.

Distinct patterns of MRI lesions in MOG antibody disease and AQP4 NMOSD: a systematic review and meta-analysis

BACKGROUND

the distinct MRI features of MOG-antibody disease (MOG-AD) and AQP4-NMOSD are still poorly defined. We performed a systematic review and meta-analysis to identify specific patterns of MRI abnormalities able to discriminate between MOG-AD and AQP4-NMOSD.

METHODS

fourteen case-series (1028 patients) were included. Outcomes were MRI lesion patterns in optic nerve (ON), brain and spinal cord (SC) that were selected after a systematic literature review and analysed separately as the event rate for individual MRI lesions in MOG-AD (experimental group) and AQP4-NMOSD (control group) by using a random effect model.

RESULTS

MOG-AD showed a higher number of MRI lesions than AQP4-NMOSD patients in the retrobulbar ON (OR=5.67; 95%CI=2.11-15.24; p=0.0006) with ON head swelling (OR=8.20; 95%CI=4.13-16.28; p<0.00001), corpus callosum (OR=2.30; 95%CI=1.11-4.76; p=0.02), pons (OR=2.87; 95%CI=1.45-5.67; p=0.002), and lumbar/conus SC (OR=3.47; 95%CI=1.66-7.24; p=0.0009). Conversely, lesions in the canalicular (OR=0.42; 95%CI=0.18-0.98; p=0.05) and intracranial ON (OR=0.30; 95%CI=0.11=0.84; p=0.02), area postrema (OR=0.12; 95%CI=0.02-0.61; p=0.01), medulla (OR=0.40; 95%CI=0.20-0.78; p=0.007), and cervical SC (OR=0.29; 95%CI=0.09-0.92; p=0.04) were prominent in patients with AQP4-NMOSD. Participants' age was found to be a source of heterogeneity across studies.

CONCLUSION

our study provides further evidence that MOG-AD and AQP4-NMOSD have distinct MRI features that may help clinicians for an early differential diagnosis.

A window into the future? MRI for evaluation of neuromyelitis optica spectrum disorder throughout the disease course

Neuromyelitis optica spectrum disorder (NMOSD) is a relapsing, inflammatory disease of the central nervous system marked by relapses often associated with poor recovery and long-term disability. Magnetic resonance imaging (MRI) is recognized as an important tool for timely diagnosis of NMOSD as, in combination with serologic testing, it aids in distinguishing NMOSD from possible mimics. Although the role of MRI for disease monitoring after diagnosis is not as well established, MRI may provide important prognostic information and help differentiate between relapses and pseudorelapses. Increasing evidence of subclinical disease activity and the emergence of newly approved, highly effective immunotherapies for NMOSD adjure us to re-evaluate MRI as a tool to guide optimal treatment selection and escalation throughout the disease course. In this article we review the role of MRI in NMOSD diagnosis, prognostication, disease monitoring, and treatment selection.

Deep Learning on Conventional Magnetic Resonance Imaging Improves the Diagnosis of Multiple Sclerosis Mimics

OBJECTIVES

The aims of this study were to present a deep learning approach for the automated classification of multiple sclerosis and its mimics and compare model performance with that of 2 expert neuroradiologists.

MATERIALS AND METHODS

A total of 268 T2-weighted and T1-weighted brain magnetic resonance imagin scans were retrospectively collected from patients with migraine (n = 56), multiple sclerosis (n = 70), neuromyelitis optica spectrum disorders (n = 91), and central nervous system vasculitis (n = 51). The neural network architecture, trained on 178 scans, was based on a cascade of 4 three-dimensional convolutional layers, followed by a fully dense layer after feature extraction. The ability of the final algorithm to correctly classify the diseases in an independent test set of 90 scans was compared with that of the neuroradiologists.

RESULTS

The interrater agreement was 84.9% (Cohen κ = 0.78, P < 0.001). In the test set, deep learning and expert raters reached the highest diagnostic accuracy in multiple sclerosis (98.8% vs 72.8%, P < 0.001, for rater 1; and 81.8%, P < 0.001, for rater 2) and the lowest in neuromyelitis optica spectrum disorders (88.6% vs 4.4%, P < 0.001, for both raters), whereas they achieved intermediate values for migraine (92.2% vs 53%, P = 0.03, for rater 1; and 64.8%, P = 0.01, for rater 2) and vasculitis (92.1% vs 54.6%, P = 0.3, for rater 1; and 45.5%, P = 0.2, for rater 2). The overall performance of the automated method exceeded that of expert raters, with the worst misdiagnosis when discriminating between neuromyelitis optica spectrum disorders and vasculitis or migraine.

CONCLUSIONS

A neural network performed better than expert raters in terms of accuracy in classifying white matter disorders from magnetic resonance imaging and may help in their diagnostic work-up.

Novel insights into pathophysiology and therapeutic possibilities reveal further differences between AQP4-IgG- and MOG-IgG-associated diseases

PURPOSE OF REVIEW

This review summarizes recent insights into the pathogenesis and therapeutic options for patients with MOG- or AQP4-antibodies.

RECENT FINDINGS

Although AQP4-IgG are linked to NMOSD, MOG-IgG-associated diseases (MOGAD) include a broader clinical spectrum of autoimmune diseases of the central nervous system (CNS). Details of membrane assembly of AQP4-IgG required for complement activation have been uncovered. Affinity-purified MOG-IgG from patients were shown to be pathogenic by induction of demyelination when the blood--brain barrier (BBB) was breached and by enhancement of activation of cognate T cells. A high-affinity AQP4-IgG, given peripherally, could induce NMOSD-like lesions in rats in the absence of BBB breach. Circulating AQP4-specific and MOG-specific B cells were identified and suggest differences in origin of MOG-antibodies or AQP4-antibodies. Patients with MOG-IgG show a dichotomy concerning circulating MOG-specific B cells; whether this is related to differences in clinical response of anti-CD20 therapy remains to be analyzed. Clinical trials of AQP4-IgG-positive NMOSD patients showed success with eculizumab (preventing cleavage of complement factor C5, thereby blocking formation of chemotactic C5a and membrane attack complex C9neo), inebilizumab (depleting CD19 + B cells), and satralizumab (anti-IL-6R blocking IL-6 actions).

SUMMARY

New insights into pathological mechanisms and therapeutic responses argue to consider NMOSD with AQP4-IgG and MOGAD as separate disease entities.

3D Compressed Convolutional Neural Network Differentiates Neuromyelitis Optical Spectrum Disorders From Multiple Sclerosis Using Automated White Matter Hyperintensities Segmentations

Background

Magnetic resonance imaging (MRI) has a wide range of applications in medical imaging. Recently, studies based on deep learning algorithms have demonstrated powerful processing capabilities for medical imaging data. Previous studies have mostly focused on common diseases that usually have large scales of datasets and centralized the lesions in the brain. In this paper, we used deep learning models to process MRI images to differentiate the rare neuromyelitis optical spectrum disorder (NMOSD) from multiple sclerosis (MS) automatically, which are characterized by scattered and overlapping lesions.

Methods

We proposed a novel model structure to capture 3D MRI images’ essential information and converted them into lower dimensions. To empirically prove the efficiency of our model, firstly, we used a conventional 3-dimensional (3D) model to classify the T2-weighted fluid-attenuated inversion recovery (T2-FLAIR) images and proved that the traditional 3D convolutional neural network (CNN) models lack the learning capacity to distinguish between NMOSD and MS. Then, we compressed the 3D T2-FLAIR images by a two-view compression block to apply two different depths (18 and 34 layers) of 2D models for disease diagnosis and also applied transfer learning by pre-training our model on ImageNet dataset.

Results

We found that our models possess superior performance when our models were pre-trained on ImageNet dataset, in which the models’ average accuracies of 34 layers model and 18 layers model were 0.75 and 0.725, sensitivities were 0.707 and 0.708, and specificities were 0.759 and 0.719, respectively. Meanwhile, the traditional 3D CNN models lacked the learning capacity to distinguish between NMOSD and MS.

Conclusion

The novel CNN model we proposed could automatically differentiate the rare NMOSD from MS, especially, our model showed better performance than traditional3D CNN models. It indicated that our 3D compressed CNN models are applicable in handling diseases with small-scale datasets and possess overlapping and scattered lesions.

Update on neuromyelitis optica spectrum disorder

PURPOSE OF REVIEW

Neuromyelitis optica spectrum disorder is an autoimmune disease that causes optic neuritis and transverse myelitis. Attacks can cause severe neurological damage leading to blindness and paralysis. Understanding of the immunopathogenesis of this disease has led to major breakthroughs in diagnosis and treatment. In the past 18 months, three successful phase 3 clinical trials have been published using targeted approaches to preventing relapses.

RECENT FINDINGS

Updates in epidemiology, imaging, quality of life and treatment for acute relapse and prevention have been published in the past 18 months. Epidemiology studies are distinguishing patients based on their antigen specificity for aquaporin-4 and myelin oligodendrocyte glycoprotein, which are increasingly recognized as separate immunological conditions. Imaging by MRI and optical coherence tomography continue to be developed as tools to distinguish neuromyelitis optica spectrum disorders (NMOSD) from other diseases. This is especially relevant as the recent clinical trials showed differences in response between aquaporin-4 seropositive and seronegative patients. The three drugs that were tested for prevention of NMOSD relapses were eculizumab, inebilizumab, and satralizumab. All of the trials were worldwide, placebo-controlled, double-masked studies that demonstrated a clear benefit with each approach.

SUMMARY

Recent research in NMOSD has resulted in improved diagnosis and approved treatments.

Brain MRI characteristics in neuromyelitis optica spectrum disorders: A large multi-center retrospective study in China

PURPOSE

To investigate the brain MRI features in neuromyelitis optica spectrum disorders (NMOSD) and its clinical relevance in a large multi-center cohort in China.

METHODS

270 NMOSD patients were recruited from seven centers. The brain MRI were classified as normal, NMOSD-specific lesions, multiple sclerosis-like, nonspecific white matter changes. Brain volumes including whole brain, white, gray matter, cortex and subcortex gray matter volume were measured. The relationship between MRI measures, clinical disability and cognitive impairment were investigated.

RESULTS

98 patients (36.3%) had normal brain MRI; 48 patients (17.7%) had NMOSD-specific lesions located in dorsal brainstem, corticospinal tract corpus, callosum and periependymal lesions surrounding the ventricular system; 16 patients (6%) had multiple sclerosis-like lesions; and 108 patients (40%) had nonspecific white matter changes. NMOSD patients with brain lesions had a trend of lower subcortex gray matter volume compared to patients without lesions. 52.5% patients with normal brain MRI and 50.8% patients with abnormal brain MRI showed cognitive impairment. No significant differences were identified in brain volume between cognitive impairment and cognitive preserved groups.

CONCLUSION

In this large multicenter NMOSD cohort, nonspecific white matter changes were the most common findings (40%). NMOSD patients with brain lesions demonstrated a trend of having lower brain volume than patients without lesions. Approximately 50% NMOSD patients presented cognitive impairment independent of brain lesions.

Spinal Cord Atrophy in Neuromyelitis Optica Spectrum Disorders Is Spatially Related to Cord Lesions and Disability

Background The spinal cord is commonly involved in patients with neuromyelitis optica spectrum disorders (NMOSDs). However, the relationship between inflammation and atrophy remains unclear. Purpose To characterize the spatial distribution of T1-hypointense lesions in the spinal cord at MRI, its association with cord atrophy, and its correlation with disability in participants with NMOSDs. Materials and Methods This prospective study evaluated three-dimensional T1-weighted spinal cord MRI scans in seropositive participants with NMOSDs and in age-matched healthy control participants acquired between February 2010 and July 2018. Binary masks of T1-hypointense lesions and lesion probability maps were produced. Cross-sectional area of the cervical and upper thoracic cord (down to T3 level) was calculated with the active-surface method. Full factorial models were used to assess cord atrophy in participants with NMOSDs. Correlations between cord atrophy and clinical and brain MRI measures were investigated with multiple regression models. Results A total of 52 participants with NMOSDs (mean age ± standard deviation, 44 years ± 15; 45 women) and 28 age-matched healthy control participants (mean age, 44 years ± 13; 16 women) were evaluated. Thirty-eight of 52 (73%) participants with NMOSDs had T1-hypointense cord lesions. No cord lesions were detected in the healthy control participants. Lesion probability maps showed a predominant involvement of the upper cervical (C2-C4) and upper thoracic (T1-T3 level) cord. The greater involvement of C1-C4 survived Bonferroni correction (P value range, .007-.04), with a higher percentage lesion extent in the gray matter (P < .001). Atrophy colocalized with focal cord lesions and correlated with pyramidal subscore (r ranging from -0.53 to -0.40; P < .001) and sensitive subscore (r ranging from -0.48 to -0.46; P = .001) of the Expanded Disability Status Scale. Participants without cord lesions had no cord atrophy. Conclusion In participants with neuromyelitis optica spectrum disorders, focal areas of spinal cord atrophy at MRI were topographically associated with lesions and correlated to motor and sensory disability. Participants without visible cord lesions had no atrophy. © RSNA, 2020 Online supplemental material is available for this article.

Evaluation of brain and spinal cord lesion distribution criteria at disease onset in distinguishing NMOSD from MS and MOG antibody-associated disorder

OBJECTIVE

To validate the recently proposed imaging criteria in distinguishing aquaporin-4 antibody (AQP4-ab)-seropositive neuromyelitis optica spectrum disorder (NMOSD) from multiple sclerosis (MS) and myelin oligodendrocyte glycoprotein antibody-associated disorder (MOG-AD) at disease onset in a Chinese population.

METHODS

We enrolled 241 patients in this retrospective study, including 143 AQP4-ab-seropositive NMOSD, 73 MS, and 25 MOG-AD. Cacciaguerra's criteria were described as fulfillment of at least 2/5 conditions including the absence of the combined juxtacortical/cortical lesions, the presence of longitudinal extensive transverse myelitis (LETM) lesions, the presence of periependymal-lateral ventricles lesions, the absence of Dawson's fingers lesions, and the absence of periventricular lesions.

RESULTS

Fulfillment of at least 3/5 conditions was able to differentiate NMOSD from MS with a good diagnostic performance (accuracy = 0.92, sensitivity = 0.91, specificity = 0.93), yet failed to differentiate NMOSD from MOG-AD. LETM lesions showed the highest accuracy (0.78), sensitivity (0.70), and specificity (0.97) for NMSOD.

CONCLUSION

Our research suggested the utility of Cacciaguerra's criteria in a Chinese population at disease onset. A better diagnostic performance in NMOSD could be attained with at least 3/5 conditions fulfilled. Yet their utility in distinguishing NMOSD from MOG-AD was limited.

Mapping white matter damage distribution in neuromyelitis optica spectrum disorders with a multimodal MRI approach

BACKGROUND

The pathogenetic mechanisms sustaining neuroinflammatory disorders may originate from the cerebrospinal fluid.

OBJECTIVE

To evaluate white matter damage with diffusion tensor imaging and T1/T2-weighted ratio at progressive distances from the ventricular system in neuromyelitis optica spectrum disorders and multiple sclerosis.

METHODS

Fractional anisotropy, mean, axial, and radial diffusivity and T1/T2-weighted ratio maps were obtained from patients with seropositive neuromyelitis optica spectrum disorders, multiple sclerosis, and healthy controls (n = 20 each group). White matter damage was assessed as function of ventricular distance within progressive concentric bands.

RESULTS

Compared to healthy controls, neuromyelitis optica spectrum disorders patients had similar fractional anisotropy, radial and axial diffusivity, increased mean diffusivity (p = 0.009-0.013) and reduced T1/T2-weighted ratio (p = 0.024-0.037) in all bands. In multiple sclerosis, gradient of percentage lesion volume and intra-lesional mean and axial diffusivity were higher in periventricular bands. Compared to healthy controls, multiple sclerosis patients had reduced fractional anisotropy (p = 0.001-0.043) in periventricular bands, increased mean (p < 0.001), radial (p < 0.001-0.004), and axial diffusivity (p = 0.002-0.008) and preserved T1/T2-weighted ratio in all bands.

CONCLUSION

White matter damage is higher at periventricular level in multiple sclerosis and diffuse in neuromyelitis optica spectrum disorders. Fractional anisotropy preservation, associated with increased mean diffusivity and reduced T1/T2-weighted ratio may reflect astrocyte damage.