Brain Atrophy in Relapsing Optic Neuritis Is Associated With Crion Phenotype

FDA-approved machine learning algorithms in neuroradiology: A systematic review of the current evidence for approval

Over the past decade, machine learning (ML) and artificial intelligence (AI) have become increasingly prevalent in the medical field. In the United States, the Food and Drug Administration (FDA) is responsible for regulating AI algorithms as "medical devices" to ensure patient safety. However, recent work has shown that the FDA approval process may be deficient. In this study, we evaluate the evidence supporting FDA-approved neuroalgorithms, the subset of machine learning algorithms with applications in the central nervous system (CNS), through a systematic review of the primary literature. Articles covering the 53 FDA-approved algorithms with applications in the CNS published in PubMed, EMBASE, Google Scholar and Scopus between database inception and January 25, 2022 were queried. Initial searches identified 1505 studies, of which 92 articles met the criteria for extraction and inclusion. Studies were identified for 26 of the 53 neuroalgorithms, of which 10 algorithms had only a single peer-reviewed publication. Performance metrics were available for 15 algorithms, external validation studies were available for 24 algorithms, and studies exploring the use of algorithms in clinical practice were available for 7 algorithms. Papers studying the clinical utility of these algorithms focused on three domains: workflow efficiency, cost savings, and clinical outcomes. Our analysis suggests that there is a meaningful gap between the FDA approval of machine learning algorithms and their clinical utilization. There appears to be room for process improvement by implementation of the following recommendations: the provision of compelling evidence that algorithms perform as intended, mandating minimum sample sizes, reporting of a predefined set of performance metrics for all algorithms and clinical application of algorithms prior to widespread use. This work will serve as a baseline for future research into the ideal regulatory framework for AI applications worldwide.

Deviation From Normative Whole Brain and Deep Gray Matter Growth in Children With MOGAD, MS, and Monophasic Seronegative Demyelination

BACKGROUND AND OBJECTIVES

Pediatric-acquired demyelination of the CNS associated with antibodies directed against myelin oligodendrocyte glycoprotein (MOG; MOG antibody-associated disease [MOGAD]) occurs as a monophasic or relapsing disease and with variable but often extensive T2 lesions in the brain. The impact of MOGAD on brain growth during maturation is unknown. We quantified the effect of pediatric MOGAD on brain growth trajectories and compared this with the growth trajectories of age-matched and sex-matched healthy children and children with multiple sclerosis (MS, a chronic relapsing disease known to lead to failure of normal brain growth and to loss of brain volume) and monophasic seronegative demyelination.

METHODS

We included children enrolled at incident attack in the prospective longitudinal Canadian Pediatric Demyelinating Disease Study who were recruited at the 3 largest enrollment sites, underwent research brain MRI scans, and were tested for serum MOG-IgG. Children seropositive for MOG-IgG were diagnosed with MOGAD. MS was diagnosed per the 2017 McDonald criteria. Monophasic seronegative demyelination was confirmed in children with no clinical or MRI evidence of recurrent demyelination and negative results for MOG-IgG and aquaporin-4-IgG. Whole and regional brain volumes were computed through symmetric nonlinear registration to templates. We computed age-normalized and sex-normalized z scores for brain volume using a normative dataset of 813 brain MRI scans obtained from typically developing children and used mixed-effect models to assess potential deviation from brain growth trajectories.

RESULTS

We assessed brain volumes of 46 children with MOGAD, 26 with MS, and 51 with monophasic seronegative demyelinating syndrome. Children with MOGAD exhibited delayed (p < 0.001) age-expected and sex-expected growth of thalamus, caudate, and globus pallidus, normalized for the whole brain volume. Divergence from expected growth was particularly pronounced in the first year postonset and was detected even in children with monophasic MOGAD. Thalamic volume abnormalities were less pronounced in children with MOGAD compared with those in children with MS.

DISCUSSION

The onset of MOGAD during childhood adversely affects the expected trajectory of growth of deep gray matter structures, with accelerated changes in the months after an acute attack. Further studies are required to better determine the relative impact of monophasic vs relapsing MOGAD and whether relapsing MOGAD with attacks isolated to the optic nerves or spinal cord affects brain volume over time.

Updated Review of the Evidence Supporting the Medical and Legal Use of NeuroQuant® and NeuroGage® in Patients With Traumatic Brain Injury

Over 40 years of research have shown that traumatic brain injury affects brain volume. However, technical and practical limitations made it difficult to detect brain volume abnormalities in patients suffering from chronic effects of mild or moderate traumatic brain injury. This situation improved in 2006 with the FDA clearance of NeuroQuant®, a commercially available, computer-automated software program for measuring MRI brain volume in human subjects. More recent strides were made with the introduction of NeuroGage®, commercially available software that is based on NeuroQuant® and extends its utility in several ways. Studies using these and similar methods have found that most patients with chronic mild or moderate traumatic brain injury have brain volume abnormalities, and several of these studies found-surprisingly-more abnormal enlargement than atrophy. More generally, 102 peer-reviewed studies have supported the reliability and validity of NeuroQuant® and NeuroGage®. Furthermore, this updated version of a previous review addresses whether NeuroQuant® and NeuroGage® meet the Daubert standard for admissibility in court. It concludes that NeuroQuant® and NeuroGage® meet the Daubert standard based on their reliability, validity, and objectivity. Due to the improvements in technology over the years, these brain volumetric techniques are practical and readily available for clinical or forensic use, and thus they are important tools for detecting signs of brain injury.

Evidence for and against subclinical disease activity and progressive disease in MOG antibody disease and neuromyelitis optica spectrum disorder

Myelin oligodendrocyte glycoprotein antibody disease (MOGAD) and aquaporin-4 IgG seropositive neuromyelitis optica spectrum disorder (AQP4-IgG+ NMOSD) are generally considered to be relapsing disorders, without clinical progression or subclinical disease activity outside of clinical relapses, in contrast to multiple sclerosis (MS). With advances in the diagnosis and treatment of these conditions, prolonged periods of remission without relapses can be achieved, and the question of whether progressive disease courses can occur has re-emerged. In this review, we focus on studies exploring evidence for and against relapse-independent clinical progression and/or subclinical disease activity in patients with MOGAD and AQP4-IgG+ NMOSD.

Altered Small-World Functional Network Topology in Patients with Optic Neuritis: A Resting-State fMRI Study

Aim

This study investigated changes in small-world topology and brain functional connectivity in patients with optic neuritis (ON) by resting-state functional magnetic resonance imaging (rs-fMRI) and based on graph theory.

Methods

A total of 21 patients with ON (8 males and 13 females) and 21 matched healthy control subjects (8 males and 13 females) were enrolled and underwent rs-fMRI. Data were preprocessed and the brain was divided into 116 regions of interest. Small-world network parameters and area under the integral curve (AUC) were calculated from pairwise brain interval correlation coefficients. Differences in brain network parameter AUCs between the 2 groups were evaluated with the independent sample t-test, and changes in brain connection strength between ON patients and control subjects were assessed by network-based statistical analysis.

Results

In the sparsity range from 0.08 to 0.48, both groups exhibited small-world attributes. Compared to the control group, global network efficiency, normalized clustering coefficient, and small-world value were higher whereas the clustering coefficient value was lower in ON patients. There were no differences in characteristic path length, local network efficiency, and normalized characteristic path length between groups. In addition, ON patients had lower brain functional connectivity strength among the rolandic operculum, medial superior frontal gyrus, insula, median cingulate and paracingulate gyri, amygdala, superior parietal gyrus, inferior parietal gyrus, supramarginal gyrus, angular gyrus, lenticular nucleus, pallidum, superior temporal gyrus, and cerebellum compared to the control group (P < 0.05).

Conclusion

Patients with ON show typical "small world" topology that differed from that detected in HC brain networks. The brain network in ON has a small-world attribute but shows reduced and abnormal connectivity compared to normal subjects and likely causes symptoms of cognitive impairment.