Search

Manual scoring of sleep spindles can be very time-consuming, and achieving accurate manual scoring on a long-term recording requires high and sustained levels of vigilance, which makes it a highly demanding task with the associated risk of decreased diagnosis accuracy. Although automatic spindle detection would be attractive, most available algorithms are sensitive to variations in spindle amplitude and frequency that occur between both subjects and derivations, reducing their effectiveness. We propose here an algorithm that models the amplitude-frequency spindle distribution with a bivariate normal distribution (one normal model per derivation). Subsequently, spindles are detected when their amplitude-frequency characteristics are included within a given tolerance interval of the corresponding model. As a consequence, spindle detection is not directly based on amplitude and frequency thresholds, but instead on a spindle distribution model that is automatically adapted to each individual subject and derivation. The algorithm was first assessed against the scoring of one sleep scoring expert on EEG samples from seven healthy children. Afterward, a second study compared performance of two additional experts versus the algorithm on a dataset of six EEG samples from adult patients suffering from different pathologies, to submit the method to more challenging and clinically realistic conditions. Smaller and shorter spindles were more difficult to evaluate, as false positives and false negatives showed lower amplitude and smaller length than true positives. In both studies, normal modelling enhanced performance compared to fixed amplitude and frequency thresholds. Normal modelling is therefore attractive, as it enhances spindle detection quality., (Copyright © 2013 Elsevier B.V. All rights reserved.)

Visual quantification of interictal epileptiform activity is time consuming and requires a high level of expert's vigilance. This is especially true for overnight recordings of patient suffering from epileptic encephalopathy with continuous spike and waves during slow-wave sleep (CSWS) as they can show tens of thousands of spikes. Automatic spike detection would be attractive for this condition, but available algorithms have methodological limitations related to variation in spike morphology both between patients and within a single recording. We propose a fully automated method of interictal spike detection that adapts to interpatient and intrapatient variation in spike morphology. The algorithm works in five steps. (1) Spikes are detected using parameters suitable for highly sensitive detection. (2) Detected spikes are separated into clusters. (3) The number of clusters is automatically adjusted. (4) Centroids are used as templates for more specific spike detections, therefore adapting to the types of spike morphology. (5) Detected spikes are summed. The algorithm was evaluated on EEG samples from 20 children suffering from epilepsy with CSWS. When compared to the manual scoring of 3 EEG experts (3 records), the algorithm demonstrated similar performance since sensitivity and selectivity were 0.3% higher and 0.4% lower, respectively. The algorithm showed little difference compared to the manual scoring of another expert for the spike-and-wave index evaluation in 17 additional records (the mean absolute difference was 3.8%). This algorithm is therefore efficient for the count of interictal spikes and determination of a spike-and-wave index., (Copyright © 2012 Elsevier B.V. All rights reserved.)