GROND: automated characterization of periodic and rhythmic EEG patterns
5 affiliations
- 1Department of Neurology, Clinical Data Animation Center, Beth Israel Deaconess Medical Center/Harvard Medical School, Boston, MA, United States of America
- 2Stanford University, Palo Alto, CA, United States of America
- 3Department of Neurology, Massachusetts General Hospital/Harvard Medical School, Boston, MA, United States of America
- 4Department of Neurology, Yale University School of Medicine, New Haven, CT, United States of America
- 5Department of Neurology, Washington University in St. Louis, St. Louis, MO, United States of America
Abstract
Objective. Periodic discharges (PDs) and rhythmic delta activity (RDA) are common electroencephalographic patterns in critically ill patients that require detailed characterization—including lateralization, spatial localization, and frequency estimation—according to the American Clinical Neurophysiology Society 2021 standardized terminology. Manual characterization is subjective, time-consuming, and exhibits substantial inter-rater variability. This study presents a comprehensive automated system for characterizing all four major subtypes: lateralized PDs (LPDs), generalized PDs (GPD), lateralized RDA (LRDA), and generalized RDA (GRDA). Approach. Two complementary pipelines were developed: the PD-Profiler and the RDA-Profiler. The PD-Profiler combines a per-channel convolutional neural network (Channelpd-Net) with hemisphere-specific learned evidence traces (HemiCET-UNet), dynamic programming with an approximately-periodic prior, and discharge-locked topographic localization. The RDA-Profiler uses iterative narrowband Hilbert refinement for frequency and lateralization, with phase-locking value analysis for spatial extent. Both pipelines were trained and evaluated on 12 425 EEG segments from 11 729 unique patients. Main results. Four electroencephalographers—three (SZ, TZ, AS) not involved in algorithm development, plus the primary annotator (MW) — independently scored 200 stratified segments per pattern subtype, with analysis restricted to segments accepted by a majority of the four raters. The algorithm exceeded expert–expert (EE) inter-rater agreement on LPD frequency (mean expert–algorithm (EA) intraclass correlation coefficient (ICC) 0.931 vs. EE 0.897; paired segment-bootstrap Δ = +0.034Δ=+0.034, 95% confidence interval [+0.016, +0.055][+0.016,+0.055], p < 0.001p<0.001), and was statistically indistinguishable from EE on every other attribute tested: GPD frequency (EA ICC 0.980 vs. EE 0.973; Δ = +0.007Δ=+0.007, [-0.000, +0.017][−0.000,+0.017], p = 0.061p=0.061), GRDA frequency (Δ = -0.020Δ=−0.020, p = 0.056p=0.056), LRDA frequency (Δ = -0.055Δ=−0.055, p = 0.435p=0.435), LPD laterality (EA κκ 0.945 vs. EE 0.974; Δ = -0.030Δ=−0.030, p = 0.079p=0.079) and LRDA laterality (EA κκ 0.961 vs. EE 0.993; Δ = -0.032Δ=−0.032, [-0.087, +0.004][−0.087,+0.004], p = 0.269p=0.269). No attribute fell significantly below the EE ceiling. PD discharge timing achieved an F1 of 0.817 at single-sample precision (mean absolute timing error 3.1 ms at 200 Hz; median error 0 samples). Pattern lateralization area under the receiver operating characteristic curve was 0.989 for PD hemisphere (n = 1276n=1276), 0.909 for LPD vs. GPD (n = 7037n=7037), and 0.837 for RDA (n = 4253n=4253). PD spatial localization reached 97.3% of expert inter-rater Jaccard agreement; RDA spatial extent matched EE ICC (0.371 vs. 0.373), at a ceiling set by the inter-rater noise of verbal region-counting labels. Significance. This work presents the first system to reach expert-level inter-rater reliability across lateralization, spatial localization, discharge timing and frequency for both periodic and rhythmic EEG patterns: it exceeded the EE ceiling on LPD frequency and was statistically indistinguishable from that ceiling on every other attribute tested, with none falling significantly below it. Agreement remains lowest, though not significantly below the ceiling, on LRDA laterality and LRDA frequency, which we identify as the priority targets for further work. In post-hoc review of discordant cases, experts judged the algorithm’s frequency estimates more accurate than the original expert labels in 94% of cases. Automated characterization can now both substitute for and improve manual annotation in critical-care EEG.
Related
- Categorical rhythmic patterns do not predict song preference in zebra finches
- EEG Entrainment to Non-Speech Rhythmic Patterns Associated with Dimensions of Language Performance
- Characterizing rhythmic wheel-turning behavioral patterns in cockroaches Rhyparobia maderae using machine learning
- Cortex-wide neuronal dynamics underpin quasi-periodic patterns in resting-state fMRI
- Automated acoustic classifiers provide insights into calling patterns of cicadas in tropical Australia
BibTeX
@article{jing2026grond,
title = {{GROND: automated characterization of periodic and rhythmic EEG patterns}},
author = {Jin Jing and Chenxi Sun and Tianyu Zhang and Matthew Byrd and Alexandra–Maria Tăuţan and Lara Basovic and Peter N Hadar and Marta P Fernandes and Daniel Goldenholz and Jennifer Kim and Aaron F Struck and Sahar F Zafar and M Brandon Westover},
journal = {Journal of Neural Engineering},
year = {2026},
month = nov,
doi = {10.1088/1741-2552/aeb331},
url = {https://doi.org/10.1088/1741-2552/aeb331},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - GROND: automated characterization of periodic and rhythmic EEG patterns
AU - Jin Jing
AU - Chenxi Sun
AU - Tianyu Zhang
AU - Matthew Byrd
AU - Alexandra–Maria Tăuţan
AU - Lara Basovic
AU - Peter N Hadar
AU - Marta P Fernandes
AU - Daniel Goldenholz
AU - Jennifer Kim
AU - Aaron F Struck
AU - Sahar F Zafar
AU - M Brandon Westover
PY - 2026
DA - 2026/11/03
JO - Journal of Neural Engineering
DO - 10.1088/1741-2552/aeb331
UR - https://doi.org/10.1088/1741-2552/aeb331
AB - Objective. Periodic discharges (PDs) and rhythmic delta activity (RDA) are common electroencephalographic patterns in critically ill patients that require detailed characterization—including lateralization, spatial localization, and frequency estimation—according to the American Clinical Neurophysiology Society 2021 standardized terminology. Manual characterization is subjective, time-consuming, and exhibits substantial inter-rater variability. This study presents a comprehensive automated system for characterizing all four major subtypes: lateralized PDs (LPDs), generalized PDs (GPD), lateralized RDA (LRDA), and generalized RDA (GRDA). Approach. Two complementary pipelines were developed: the PD-Profiler and the RDA-Profiler. The PD-Profiler combines a per-channel convolutional neural network (Channelpd-Net) with hemisphere-specific learned evidence traces (HemiCET-UNet), dynamic programming with an approximately-periodic prior, and discharge-locked topographic localization. The RDA-Profiler uses iterative narrowband Hilbert refinement for frequency and lateralization, with phase-locking value analysis for spatial extent. Both pipelines were trained and evaluated on 12 425 EEG segments from 11 729 unique patients. Main results. Four electroencephalographers—three (SZ, TZ, AS) not involved in algorithm development, plus the primary annotator (MW) — independently scored 200 stratified segments per pattern subtype, with analysis restricted to segments accepted by a majority of the four raters. The algorithm exceeded expert–expert (EE) inter-rater agreement on LPD frequency (mean expert–algorithm (EA) intraclass correlation coefficient (ICC) 0.931 vs. EE 0.897; paired segment-bootstrap Δ = +0.034Δ=+0.034, 95% confidence interval [+0.016, +0.055][+0.016,+0.055], p < 0.001p<0.001), and was statistically indistinguishable from EE on every other attribute tested: GPD frequency (EA ICC 0.980 vs. EE 0.973; Δ = +0.007Δ=+0.007, [-0.000, +0.017][−0.000,+0.017], p = 0.061p=0.061), GRDA frequency (Δ = -0.020Δ=−0.020, p = 0.056p=0.056), LRDA frequency (Δ = -0.055Δ=−0.055, p = 0.435p=0.435), LPD laterality (EA κκ 0.945 vs. EE 0.974; Δ = -0.030Δ=−0.030, p = 0.079p=0.079) and LRDA laterality (EA κκ 0.961 vs. EE 0.993; Δ = -0.032Δ=−0.032, [-0.087, +0.004][−0.087,+0.004], p = 0.269p=0.269). No attribute fell significantly below the EE ceiling. PD discharge timing achieved an F1 of 0.817 at single-sample precision (mean absolute timing error 3.1 ms at 200 Hz; median error 0 samples). Pattern lateralization area under the receiver operating characteristic curve was 0.989 for PD hemisphere (n = 1276n=1276), 0.909 for LPD vs. GPD (n = 7037n=7037), and 0.837 for RDA (n = 4253n=4253). PD spatial localization reached 97.3% of expert inter-rater Jaccard agreement; RDA spatial extent matched EE ICC (0.371 vs. 0.373), at a ceiling set by the inter-rater noise of verbal region-counting labels. Significance. This work presents the first system to reach expert-level inter-rater reliability across lateralization, spatial localization, discharge timing and frequency for both periodic and rhythmic EEG patterns: it exceeded the EE ceiling on LPD frequency and was statistically indistinguishable from that ceiling on every other attribute tested, with none falling significantly below it. Agreement remains lowest, though not significantly below the ceiling, on LRDA laterality and LRDA frequency, which we identify as the priority targets for further work. In post-hoc review of discordant cases, experts judged the algorithm’s frequency estimates more accurate than the original expert labels in 94% of cases. Automated characterization can now both substitute for and improve manual annotation in critical-care EEG.
SN - 1741-2560
SN - 1741-2552
ER -