Search

Objective Methods Results Significance The automated interpretation of clinical electroencephalograms (EEGs) using artificial intelligence (AI) holds the potential to bridge the treatment gap in resource‐limited settings and reduce the workload at specialized centers. However, to facilitate broad clinical implementation, it is essential to establish generalizability across diverse patient populations and equipment. We assessed whether SCORE‐AI demonstrates diagnostic accuracy comparable to that of experts when applied to a geographically different patient population, recorded with distinct EEG equipment and technical settings.We assessed the diagnostic accuracy of a “fixed‐and‐frozen” AI model, using an independent dataset and external gold standard, and benchmarked it against three experts blinded to all other data. The dataset comprised 50% normal and 50% abnormal routine EEGs, equally distributed among the four major classes of EEG abnormalities (focal epileptiform, generalized epileptiform, focal nonepileptiform, and diffuse nonepileptiform). To assess diagnostic accuracy, we computed sensitivity, specificity, and accuracy of the AI model and the experts against the external gold standard.We analyzed EEGs from 104 patients (64 females, median age = 38.6 [range = 16–91] years). SCORE‐AI performed equally well compared to the experts, with an overall accuracy of 92% (95% confidence interval [CI] = 90%–94%) versus 94% (95% CI = 92%–96%). There was no significant difference between SCORE‐AI and the experts for any metric or category. SCORE‐AI performed well independently of the vigilance state (false classification during awake: 5/41 [12.2%], false classification during sleep: 2/11 [18.2%]; p = .63) and normal variants (false classification in presence of normal variants: 4/14 [28.6%], false classification in absence of normal variants: 3/38 [7.9%]; p = .07).SCORE‐AI achieved diagnostic performance equal to human experts in an EEG dataset independent of the development dataset, in a geographically distinct patient population, recorded with different equipment and technical settings than the development dataset. [ABSTRACT FROM AUTHOR]

Our goal was to assess the interrater agreement (IRA) of photoparoxysmal response (PPR) using the classification proposed by a task force of the International League Against Epilepsy (ILAE), and a simplified classification system proposed by our group. In addition, we evaluated IRA of epileptiform discharges (EDs) and the diagnostic significance of the electroencephalographic (EEG) abnormalities. We used EEG recordings from the European Reference Network (EpiCARE) and Standardized Computer‐based Organized Reporting of EEG (SCORE). Six raters independently scored EEG recordings from 30 patients. We calculated the agreement coefficient (AC) for each feature. IRA of PPR using the classification proposed by the ILAE task force was only fair (AC = 0.38). This improved to a moderate agreement by using the simplified classification (AC = 0.56; P =.004). IRA of EDs was almost perfect (AC = 0.98), and IRA of scoring the diagnostic significance was moderate (AC = 0.51). Our results suggest that the simplified classification of the PPR is suitable for implementation in clinical practice. [ABSTRACT FROM AUTHOR]

• We characterized photoparoxysmal response (PPR) in a SCORE database of 10,671 EEG recordings. • Using standardized intermittent photic stimulation protocol and definitions, the prevalence of PPR was 4%. • The most common PPR type was activation of preexisting epileptogenic area. To characterize photoparoxysmal EEG response (PPR) using a standardized protocol of intermittent photic stimulation (IPS) and standardized definitions for PPR, classified into six types. Using the SCORE system (Standardized Computer-Based Organized Reporting of EEG) we prospectively built a large database of standardized EEG annotations. In this study, we extracted the features related to PPR from the structured dataset consisting of 10,671 EEG recordings with IPS, from 7,188 patients. The standardized IPS protocol elicited PPR in 375 recordings (3.5%), in 288 patients (4%), with a preponderance among young (11–20 years) and female patients (67%). PPR was persistent in patients with multiple recordings. The most frequent type of PPR was activation of preexisting epileptogenic area (58%), followed by generalized-PPR limited to the stimulus train (22%). We could not find any recording with self-sustained posterior response. Seizures were elicited in 27% of patients with PPR, most often myoclonic seizures and absences, in patients with self-sustained generalized PPR. The most common type of PPR was accentuation of preexisting epileptogenic area. Self-sustained posterior response could not be documented. Self-sustained generalized-PPR had the highest association with seizures. Using standardized stimulation protocol and definitions for PPR types, IPS provides high diagnostic yield. [ABSTRACT FROM AUTHOR]