We report a case of an acute HHV-7 encephalitis involving the nucleus of the VI cranial nerve in an immunocompetent host. The patient was an adult male admitted to our Clinic with headache, diplopia, fever, nausea, vertigo, asthenia and general malaise. PCR for viral and bacterial genomes was run on both serum and cerebral spinal fluid (CSF) after performing lumbar puncture, resulting positive only for HHV-7 DNA on CSF. MRI showed hyperintensity in FLAIR signal in the dorsal pons, in the area of the VI cranial nerve nucleus. Empirical therapy with Acyclovir and Dexamethasone was started at the time of admission and was continued after the microbiology results. After three days of therapy diplopia, fever and other previous clinical manifestations improved and the patient recovered normal sight. Our case report contributes to a better understanding of the presentation, diagnosis and treatment of HHV-7 encephalitis in immunocompetent patients due to reactivation in adult age.

Together with hippocampus, the amygdala is important in the epileptogenic network of patients with temporal lobe epilepsy. Recently, an increase in amygdala volumes (i.e., amygdala enlargement) has been proposed as morphological biomarker of a subtype of temporal lobe epilepsy patients without MRI abnormalities, although other data suggest that this finding might be unspecific and not exclusive to temporal lobe epilepsy. In these studies, the amygdala is treated as a single entity, while instead it is composed of different nuclei, each with peculiar function and connection. By adopting a recently developed methodology of amygdala’s subnuclei parcellation based of high-resolution T1-weighted image, this study aims to map specific amygdalar subnuclei participation in temporal lobe epilepsy due to hippocampal sclerosis (n = 24) and non-lesional temporal lobe epilepsy (n = 24) with respect to patients with focal extra temporal lobe epilepsies (n = 20) and healthy controls (n = 30). The volumes of amygdala subnuclei were compared between groups adopting multivariate analyses of covariance and correlated with clinical variables. Additionally, a logistic regression analysis on the nuclei resulting statistically different across groups was performed. Compared to other populations, temporal lobe epilepsy with hippocampal sclerosis showed a significant atrophy of the whole amygdala (pBonferroni = .040), particularly the basolateral complex (pBonferroni = .033), while the non-lesional temporal lobe epilepsy group demonstrated an isolated hypertrophy of the medial nucleus (pBonferroni = .012). In both scenarios, the involved amygdala was ipsilateral to the epileptic focus. The medial nucleus demonstrated a volume increase even in extra temporal lobe epilepsies although contralateral to the seizure onset hemisphere (pBonferroni = .037). Non-lesional patients with psychiatric comorbidities showed a larger ipsilateral lateral nucleus compared to those without psychiatric disorders. This exploratory study corroborates the involvement of the amygdala in temporal lobe epilepsy, particularly in mesial temporal lobe epilepsy and suggests a different amygdala subnuclei engagement depending on the etiology and lateralization of epilepsy. Furthermore, the logistic regression analysis indicated that the basolateral complex and the medial nucleus of amygdala can be helpful to differentiate temporal lobe epilepsy with hippocampal sclerosis and with MRI negative, respectively, versus controls with a consequent potential clinical yield. Finally, the present results contribute to the literature about the amygdala enlargement in temporal lobe epilepsy, suggesting that the increased volume of amygdala can be regarded as epilepsy-related structural changes common across different syndromes whose meaning should be clarified.

Research on segmentation of the hippocampus in magnetic resonance images through deep learning convolutional neural networks (CNNs) shows promising results, suggesting that these methods can identify small structural abnormalities of the hippocampus, which are among the earliest and most frequent brain changes associated with Alzheimer disease (AD). However, CNNs typically achieve the highest accuracy on datasets acquired from the same domain as the training dataset. Transfer learning allows domain adaptation through further training on a limited dataset. In this study, we applied transfer learning on a network called spatial warping network segmentation (SWANS), developed and trained in a previous study. We used MR images of patients with clinical diagnoses of mild cognitive impairment (MCI) and AD, segmented by two different raters. By using transfer learning techniques, we developed four new models, using different training methods. Testing was performed using 26% of the original dataset, which was excluded from training as a hold‐out test set. In addition, 10% of the overall training dataset was used as a hold‐out validation set. Results showed that all the new models achieved better hippocampal segmentation quality than the baseline SWANS model ( p s < .001), with high similarity to the manual segmentations (mean dice [best model] = 0.878 ± 0.003). The best model was chosen based on visual assessment and volume percentage error (VPE). The increased precision in estimating hippocampal volumes allows the detection of small hippocampal abnormalities already present in the MCI phase ( SD = [3.9 ± 0.6]%), which may be crucial for early diagnosis.

No abstract

N