Prediction of patient survival following postanoxic coma using EEG data and clinical features.

Electroencephalography (EEG) is an effective and non-invasive technique commonly used to monitor brain activity and assist in outcome prediction for comatose patients post cardiac arrest. EEG data may demonstrate patterns associated with poor neurological outcome for patients with hypoxic injury. Thus, both quantitative EEG (qEEG) and clinical data contain prognostic information for patient outcome. In this study we use machine learning (ML) techniques, random forest (RF) and support vector machine (SVM) to classify patient outcome post cardiac arrest using qEEG and clinical feature sets, individually and combined. Our ML experiments show RF and SVM perform better using the joint feature set. In addition, we extend our work by implementing a convolutional neural network (CNN) based on time-frequency images derived from EEG to compare with our qEEG ML models. The results demonstrate significant performance improvement in outcome prediction using non-feature based CNN compared to our feature based ML models. Implementation of ML and DL methods in clinical practice have the potential to improve reliability of traditional qualitative assessments for postanoxic coma patients.

Developing Normative Reference Values for Nerve Conduction Studies of Commonly Tested Nerves among a Sample Pakistani Population.

INTRODUCTION: Most neurophysiology departments around the world establish their own normative data. However, ethnic differences are not taken into account. Our aim was to establish normal nerve conduction studies (NCS) data for routinely tested nerves in individuals of Pakistani (South Asian) origin and to compare with Western published data. MATERIALS AND METHODS: One hundred healthy adults' nerves were assessed, using standardized techniques. Individuals were grouped into age groups. Gender differences were assessed. RESULTS: Of the 100 volunteers, 49 were female and 51 were male. Their mean age was 39.8 years. Findings showed statistically significant prolongation of median distal motor latency (DML) and F-wave latency with age and reduction of median, ulnar, and sural sensory amplitudes as age increased. Gender differences showed consistent difference in the normal values for median, ulnar, and peroneal DMLs and respective F-wave latencies, which were significantly shorter in females. Sensory amplitudes of tested upper extremity nerves were significantly lower in males. Comparing with available data, our findings are similar to the Saudi population but significantly different from the American and multiethnic Malaysian populations. Pakistani individuals generally have significantly higher amplitudes and faster conduction velocities with similarities to South Asian studies. CONCLUSIONS: We recommend normative NCS parameters for commonly tested nerves for the Pakistani population, using standardized techniques to ensure highest quality testing and outcomes.

Non-convulsive seizures and electroencephalography findings as predictors of clinical outcomes at a tertiary intensive care unit in Saudi Arabia.

OBJECTIVE: Electroencephalography (EEG) in the intensive care unit (ICU) is often done to detect non-convulsive seizures (NCS). The outcome of ICU patients with NCS strongly depends on the underlying etiology. The implication of NCS and other EEG findings on clinical outcome independent from their etiology is not well understood and our aim to investigate it. PATIENTS AND METHODS: We retrospectively identified all adult patients in the ICU who underwent EEG monitoring between January 2008 and December 2011. The main goals were to define the rate of NCS or non-convulsive status epilepticus (NCSE) occurrence in our center among patients who underwent EEG monitoring and to examine if NCS/NCSE are associated with poor outcome [defined as death or dependence] with and without adjustment for underlying etiology. The rate of poor outcome among different EEG categories were also investigated. RESULTS: During the study period, 177 patients underwent EEG monitoring in our ICU. The overall outcome was poor in 62.7% of those undergoing EEG. The rate of occurrence of NCS/NCSE was 8.5% and was associated with poor outcome in 86.7% with an odds ratio (OR) of 5.1 (95% confidence interval [CI] 1.09-23.8). This association was maintained after adjusting for underlying etiologies with OR 5.6 (95% CI 1.05-29.6). The rate of poor outcome was high in the presence of periodic discharges and sharp and slow waves of 75% and 61.5%, respectively. CONCLUSIONS: Our cohort of ICU patients undergoing EEGs had a poor outcome. Those who developed NCS/NCSE experienced an even worse outcome regardless of the underlying etiology.

Subacute sclerosing panencephalitis presenting with unilateral periodic myoclonic jerks.

BACKGROUND: Subacute sclerosing panencephalitis (SSPE) is a rare complication of measles virus infection. The disease is characterized by behavioural abnormalities, intellectual deterioration, motor weakness, and generalized myoclonic jerks progressing to coma and death in one to two years in 80% of the cases. The myoclonic jerks are associated with characteristic generalized slow periodic complexes on electroencephalography (EEG). The symptoms and signs of SSPE are frequently quite variable. The clinical course is equally variable and difficult to predict. The characteristic periodic myoclonus can rarely occur unilaterally particularly in the early stages of the disease. As well, the periodic EEG complexes have been reported unilaterally in up to 3% of cases. CASE REPORT: A 12-year-old boy, who was seen at a later stage with atypical manifestation of myoclonic body jerks confined entirely unilaterally, combined with contralateral periodic EEG complexes. One could assume clinically that the more diseased hemisphere was responsible for generating the jerks. However, brain magnetic resonance imaging revealed asymmetric hemispheric changes suggesting that the less neurologically damaged hemisphere is responsible for generating the unilateral myoclonic jerks. This has led to the interpretation that the more severely damaged hemisphere has lost the neuronal connectivity required to generate these periodic myoclonic jerks. CONCLUSIONS: Subacute sclerosing panencephalitis may have asymmetric hemispheric involvement, not only early, but also in the advanced stages of the disease, which can result in unilateral periodic myoclonic jerks.

Topiramate for the treatment of infants with early myoclonic encephalopathy.

OBJECTIVE: Early myoclonic encephalopathy (EME) is a rare epileptic syndrome characterized by neonatal onset of severe recurrent seizures of multiple types often resistant to antiepileptic drugs (AEDs). Topiramate (TPM) is a new AED, which has a wide spectrum of antiepileptic activities suggesting a potentially valuable therapeutic profile. There is limited clinical data available on TPM use in infants and our aim is to report our experience with TPM for the treatment of infants with intractable seizures secondary to EME. METHODS: Prospective, open label, add on trial of TPM in treating a series of infants with EME at King Faisal Specialist Hospital and Research Centre and King Abdul-Aziz University Hospital, Jeddah, Kingdom of Saudi Arabia between June 1999 and March 2002. Topiramate was started at 12.5 mg/day and was increased by doubling the dose every week until the minimum effective dose was reached (seizure free outcome) or up to a maximum of 10 mg/kg/day. RESULTS: Four consecutive infants (2 males and 2 females) were included. In addition to daily seizures, they all had global hypotonia, developmental delay, and progressive microcephaly. The syndrome was cryptogenic in 3, and one had nonketotic hyperglycinemia. Initial electroencephalograms showed generalized epileptic burst suppression pattern. Infants were tried on multiple AEDs (6-11, mean 7.5) with no success. Topiramate was added at age 5-12 months (mean 9) reaching a maximum dose of 5.5-10 mg/kg/day (mean 7.6). The infants were then followed for up to 19 months (mean 13.5). After introducing TPM, one infant became completely seizure free and 3 had significant (>50%) seizure reduction. Electroencephalograms in 3 infants showed marked improvement. One infant had weight loss that resulted in discontinuing the drug after 6 months. Follow-up renal ultrasound findings were normal in all infants. CONCLUSION: Topiramate is effective and safe in treating infants with intractable epilepsy secondary to EME.