One outstanding challenge for machine learning in diagnostic biomedical imaging is algorithm interpretability. A key application is the identification of subtle epileptogenic focal cortical dysplasias (FCDs) from structural MRI. FCDs are difficult to visualize on structural MRI but are often amenable to surgical resection. We aimed to develop an open-source, interpretable, surface-based machine-learning algorithm to automatically identify FCDs on heterogeneous structural MRI data from epilepsy surgery centres worldwide. The Multi-centre Epilepsy Lesion Detection (MELD) Project collated and harmonized a retrospective MRI cohort of 1015 participants, 618 patients with focal FCD-related epilepsy and 397 controls, from 22 epilepsy centres worldwide. We created a neural network for FCD detection based on 33 surface-based features. The network was trained and cross-validated on 50% of the total cohort and tested on the remaining 50% as well as on 2 independent test sites. Multidimensional feature analysis and integrated gradient saliencies were used to interrogate network performance. Our pipeline outputs individual patient reports, which identify the location of predicted lesions, alongside their imaging features and relative saliency to the classifier. On a restricted ‘gold-standard’ subcohort of seizure-free patients with FCD type IIB who had T1 and fluid-attenuated inversion recovery MRI data, the MELD FCD surface-based algorithm had a sensitivity of 85%. Across the entire withheld test cohort the sensitivity was 59% and specificity was 54%. After including a border zone around lesions, to account for uncertainty around the borders of manually delineated lesion masks, the sensitivity was 67%. This multicentre, multinational study with open access protocols and code has developed a robust and interpretable machine-learning algorithm for automated detection of focal cortical dysplasias, giving physicians greater confidence in the identification of subtle MRI lesions in individuals with epilepsy.

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

Acute cerebellar ataxia is the most common cause of acute ataxia in children and it usually runs a self-limiting and ultimately benign clinical course. A small proportion of children have evidence of inflammatory swelling in the cerebellum. Many of these children suffer more severe and potentially life-threatening forms of cerebellar ataxia and may need more intensive treatments including urgent neurosurgical treatments. This more severe form of acute cerebellar ataxia is often termed acute cerebellitis. Many children with acute cerebellitis have long-term neurological sequela and evidence of structural cerebellar changes on follow-up imaging. Several patterns of cerebellar inflammation have been described. The authors describe the variabilities in the clinical and radiological patterns of disease in the cases that have been described in the literature.

Near infrared spectroscopy (NIRS) is a noninvasive method for bedside measurement of cerebral oxygenation (SaO(2)). The purpose of this study was to establish differences in SaO(2)for complex partial seizures (CPS) and rapidly secondarily generalized CPS (RCPS). We studied eight adults with medically refractory epilepsy undergoing evaluation for temporal lobectomy. We continually measured cerebral SaO(2)via a Somanetic Invos 3100a cerebral oximeter, pre-ictal (5 minutes), ictal, immediate (30 seconds) post-ictal, and late post-ictal (5 minutes after ictus). Seventeen seizures (12 CPS, four RCPS and one subclinical) were recorded in eight patients. The percentage change in cerebral SaO(2)from pre-ictal to ictal periods was derived. Cerebral SaO(2)increased (percentage change, mean: 16.6, SD: 13.9) for CPS and decreased (percentage change, mean: 51.1, SD: 18.1) for RCPS. No change in cerebral oximetry was recorded for the subclinical seizure. Post-ictal (immediate and late) increase in cerebral SaO(2)was seen for 11 of the 17 seizures (nine CPS and two RCPS). Peripheral SaO(2)rose greater than 93% for all CPS and the subclinical seizure, but decreased between 78 and 84% during RCPS. These results suggest NIRS distinguishes cerebral SaO(2)patterns between CPS and RCPS. The decrease in peripheral SaO(2), however, may account for the decrease in cerebral SaO(2)seen in generalized seizures.

Synapse loss correlates strongly with cognitive decline in Alzheimer's disease, but the 20 mechanisms underpinning this phenomenon remain unclear. Recent evidence from mouse models points to microglial cells as mediators of synapse removal, and human genetic evidence implicates microglia in disease risk. Here we demonstrate that microglia from human postmortem brain contain synaptic proteins and that greater amounts are observed in microglia from Alzheimer's compared to non-diseased brain tissue. Further, we observe that primary human 25 adult microglia phagocytose synapses isolated from human brain, and that AD brain-derived synapses are phagocytosed more rapidly and abundantly than controls. Together, these data show that synapses in the human AD brain are more prone to ingestion by microglia. Our findings provide evidence from human tissue implicating altered microglial-mediated synaptic uptake in AD pathobiology. 30One Sentence Summary: AD alters synapse ingestion by microglia