OBJECTIVE:The interpretation of electrophysiological findings may lead to misdiagnosis in polyneuropathies. We investigated the electrodiagnostic accuracy of three supervised learning algorithms (SLAs): shrinkage discriminant analysis, multinomial logistic regression, and support vector machine (SVM), and three expert and three trainee neurophysiologists.METHODS:We enrolled 434 subjects with the following diagnoses: chronic inflammatory demyelinating polyneuropathy (99), Charcot-Marie-Tooth disease type 1A (124), hereditary neuropathy with liability to pressure palsy (46), diabetic polyneuropathy (67), and controls (98). In each diagnostic class, 90% of subjects were used as training set for SLAs to establish the best performing SLA by tenfold cross validation procedure and 10% of subjects were employed as test set. Performance indicators were accuracy, precision, sensitivity, and specificity.RESULTS:SVM showed the highest overall diagnostic accuracy both in training and test sets (90.5 and 93.2%) and ranked first in a multidimensional comparison analysis. Overall accuracy of neurophysiologists ranged from 54.5 to 81.8%.CONCLUSIONS:This proof of principle study shows that SVM provides a high electrodiagnostic accuracy in polyneuropathies. We suggest that the use of SLAs in electrodiagnosis should be exploited to possibly provide a diagnostic support system especially helpful for the less experienced practitioners.

Background: Despite recent progress in the multidisciplinary management of large middle cerebral artery infarcts, the neurological prognosis remains worrying in a non-negligible number of cases. The objective of this study is to analyze the contribution of optic nerve and perioptic sheath measurement on MRI to the acute phase of large middle cerebral artery infarcts. Methods: A retrospective case-control study between January 2008 and December 2019 in a single academic medical center was performed. Cases and controls were selected by interrogation of International Classification of Diseases (ICD), 10th edition, with ischemic stroke as criterion (code I64). Decompressive hemicraniectomy was a criterion for large middle cerebral artery infarcts (cases). Cases were matched with controls (1:3) based on age (+/- 5 years), sex, and year of hospitalization (+/- 2 years) The examinations were performed on 3T MRI (Siemens IRM 3T Magnetom).Optic nerve and perioptic sheath diameter was calculated using electronic calipers, 3 mm behind retina and in a perpendicular vector with reference to the orbit in axial 3D TOF sequence. Results: Of 2612 patients, 22 patients met all the criteria of large middle cerebral artery infarcts and they were paired with 44 controls. Patients were mainly women, mean age of 53.6 years. There is a significant difference in the size of the optic nerve and perioptic sheath diameter measured on MRI at patient's admission (right: 5.13 +/- 0.2 mm vs. 4.80 mm +/- 0.18, p <0. 0001, left: 5.16 +/- 0.17 vs 4.78 +/- 0.20, p<0.0001). The AUC of optic nerve and perioptic sheath diameter was 0.93 (95%IC [0.85-1.00]), for a threshold at 5.03 mm, the sensitivity was 0.82 (95%IC [0.6-0.93]), specificity 0.94 (95%IC [0.85-0.98]). The Odds Ratio of large middle cerebral artery infarcts was 46.4 for optic nerve and perioptic sheath diameter the (95%IC [6.15-350.1] p=0.0002). Conclusion: Optic nerve and perioptic sheath diameter in the first MRI can predict the risk of developing large middle cerebral artery infarcts requiring a decompressive hemicraniectomy.

Background: Hereditary neuropathy with liability to pressure palsies (HNPP) is a rare neuropathy with a heterogeneous clinical profile. Painless recurrent palsies are the usual presentation, but neuropathic pain could be predominant or inaugural. Browsing the medical literature, we only found two articles reffering to this important clinical feature. Whether there are differences between patients with or without pain is unclear. The main objective of this study was to compare the clinical and electrophysiological features of these patients and to evaluate the impact on their disability. Methods: All patients diagnosed with HNPP at the Limoges University Hospital Centre were included and separated into two groups according to the presence or absence of neuropathic pain. In each case, the clinical, genetic, electrodiagnostic, therapeutic features and the modified Rankin Scale (mRS) were evaluated. Results: Out of 23 patients, 52% presented with neuropathic pain. There was no difference between groups regarding to clinical and electrophysiological features, except for the amplitude of the ulnar sensory nerve (p < 0,003). The amplitudes of sensory nerve action potentials (SNAPs) seemed to be higher in patients with pain, but were below the lower limit of normal. Patients with pain had a higher mRS than patients without pain (p < 0,007). Conclusion: This study supports previous published results and highlights a trend for higher sensory amplitudes in HNPP patients with pain. We found a prevalence of neuropathic pain of 52% in patients with HNPP, underlining the need to systematically assess pain in such patients in order to improve their management.

Background: Electrodiagnosis is a mainstay in the evaluation of polyneuropathies. However the interpretation of electrophysiological findings may lead to misdiagnosis of the specific polyneuropathy and up to 45% of chronic inflammatory demyelination polyneuropathy (CIDP) patients received an alternative diagnosis at reevaluation of the data. Supervised learning algorithms (SLAs) can be helpful in predicting the electrodiagnostic classification of a subject, but have been rarely employed in this task.Methods: Three academic tertiary neuromuscular centers participated in the study and 434 subjects were enrolled with the following reference diagnoses: CIDP (99), Charcot-Marie-Tooth disease type 1A (124), hereditary neuropathy with liability to pressure palsy (46), diabetic polyneuropathy (67), and controls (98). We investigated the electrodiagnostic accuracy of three SLAs: shrinkage discriminant analysis, multinomial logistic regression, support vector machine (SVM), and of three expert and three trainee neurophysiologists. Four performance indicators were employed: accuracy, precision, sensitivity, and specificity.Findings: The final data set contained, for each subject, 27 electrophysiological parameters from three motor and three sensory nerves. SVM showed the highest (91%) overall diagnostic accuracy when compared with other SLAs (84·3 and 86·6%), expert (74·9-82%) and trainees (54·6-77%) neurophysiologists. SVM demonstrated also the best balance between sensitivity and specificity in all diagnostic classes, especially in CIDP (sensitivity=80·8%, specificity=95·2%), and ranked first in a multidimensional comparison analysis based on the performance indicators. The other two SLAs ranked second and third, and the first expert neurophysiologist ranked fourth.Interpretation: SVM is an ideal tool to increase the electrodiagnostic accuracy in polyneuropathies, particularly in CIDP. SVM, as other SLAs, has the advantage of assigning to each subject the probability of belonging to different diagnostic classes. SLAs are already available in statistical packages and electromyographic machines could be easily equipped with such algorithms to provide a robust diagnostic support system especially helpful for the less experienced practitioners. Funding Statement: The authors declare: "None."Declaration of Interests: The authors declare no conflicts of interest.Ethics Approval Statement: The study was carried out in accordance with the Declaration of Helsinki and approved by the Ethics Committee of the University Federico II of Naples and the respective institutional Medical Ethics Research Committee. All subjects signed a written informed consent that allowed the utilization of electrophysiological data for research purposes.

Introduction: Emergency departments play a key role in the diagnosis and treatment of transient ischemic attacks, but limited data are available about the early management of such patients in emergency wards. Therefore, we aimed to evaluate emergency physicians' management of transient ischemic attack and analyze variations factors. Methods: A multicenter survey among emergency physicians of the Grand Est region network (Est-RESCUE) was conducted from January 28th to March 28th, 2019. Medical and administrative data were collected by the same network and the national directory of medical resources. Results: Among 542 emergency physicians recipients, 78 answered (14%) and 71 were finally included, practicing in 25 public hospitals homogeneously distributed across the territory, including 3 university hospitals. A cerebral magnetic resonance imaging was obtained for 75%-100% of patients by 4.3% of responders, 36.4% of which were performed within more than 24 hours. A cardiac monitoring was prescribed in 75%-100% of cases by 32.4% of responders. A neurologic consultation was routinely requested by 84.6% of responders practicing in a university hospital and 36.8% of responders practicing in a community hospital (P = .02). Patients were hospitalized in a neurovascular unit in 75%-100% of cases by 17.4% of responders, which happened more likely in university hospitals (P < .001). Conclusions: Transient ischemic attack suffers from management disparities across territories, due to limited access to technical facilities and neurologic consultations. Therefore, international recommendations are too often not followed. Implementation of territorial neurovascular tracks may help to standardize the management of these patients.

L’horloge biologique circadienne a une activité rythmique endogène indépendante de l’environnement. Elle est aussi synchronisée sur le rythme des 24 heures et influencée par la lumière. L’horloge centrale est située au niveau des noyaux supra-chiasmatiques de l’hypothalamus. Une multitude d’horloges périphériques, situées dans presque tous les tissus de l’organisme.The circadian clock has an endogenous activity, independently of environmental cues. It is also synchronized on the 24 hours cycle specifically by light. The central biological clock is located in the supra-chiasmatic nuclei of the hypothalamus. Multiple peripheral biological clocks are also identified in almost all the systems.

Le lymphome endovasculaire est une hémopathie rare, difficile à diagnostiquer. S’il n’est pas rare de rencontrer des manifestations neurologiques centrales au cours de cette affection, les polyradiculonévrites sont peu décrites. Nous rapportons le cas d’un patient de 48 ans, sans antécédent, qui a présenté de façon soudaine des troubles de la marche associés à des troubles vésico-sphinctériens. Devant ce qui semblait être un syndrome de la queue de cheval, une imagerie par résonance magnétique médullaire est réalisée, n’expliquant pas la symptomatologie. En moins d’une semaine, sont apparus une paraplégie flasque et une parésie des membres supérieurs. L’électromyogramme et la ponction lombaire objectivaient respectivement une polyradiculonévrite démyélinisante et une dissociation albumino-cytologique. Les cures d’immunoglobulines n’ont pas permis d’améliorer le patient. Par la suite, la persistance d’un syndrome inflammatoire biologique, la dégradation clinique du patient, l’apparition de lésions punctiformes ischémiques cérébrales et d’un hypersignal médullaire faisaient évoquer le diagnostic de lymphome endovasculaire. C’est finalement la biopsie neuromusculaire avec recherche de clonalité qui a permis de confirmer le diagnostic et de débuter un traitement par cyclophosphamide et méthylprednisolone. Si cela a pu stopper l’aggravation, le handicap du patient est resté important. La confirmation anatomopathologique du lymphome endovasculaire est difficile à obtenir, souvent rendue en post-mortem, en raison de son pronostic sombre. La biopsie nerveuse et notamment la recherche de clonalité peut permettre de faire le diagnostic, d’où la nécessité de rechercher une atteinte neurologique périphérique clinique et électrophysiologique, même si elle semble moins fréquente que l’atteinte centrale. Du fait de son accessibilité, la biopsie neuromusculaire est un examen clé du diagnostic de lymphome endovasculaire, en cas d’atteintes du système nerveux périphérique, tels que les Pseudo-Guillain Barré.