Resting state EEG power spectrum and functional connectivity in autism: a cross-sectional analysis

Garcés, Pilar; Baumeister, Sarah; Mason, Luke; Chatham, Christopher H.; Holiga, Stefan; Dukart, Juergen; Jones, Emily J. H.; Banaschewski, Tobias; Baron-Cohen, Simon; Bölte, Sven; Buitelaar, Jan K.; Durston, Sarah; Oranje, Bob; Persico, Antonio M.; Beckmann, Christian F.; Bougeron, Thomas; Dell’Acqua, Flavio; Ecker, Christine; Moessnang, Carolin; Charman, Tony; Tillmann, Julian; Murphy, Declan G. M.; Johnson, Mark; Loth, Eva; Brandeis, Daniel; Hipp, Joerg F.

doi:10.1186/s13229-022-00500-x

Molecular Autism

Table 3 ASD vs NT cross-validation classification performance of PS and FC features (training dataset)

From: Resting state EEG power spectrum and functional connectivity in autism: a cross-sectional analysis

			Linear SVC				Elastic net				Boruta + rbf SVC
Age groups	Diagnosis	Measure	acc	sens	spec	p-val	acc	sens	spec	p-val	acc	sens	spec	p-val
All	ASD	PS	0.52	0.53	0.51	0.24	0.56	0.52	0.59	0.025	0.54	0.54	0.54	0.08
All	ASD	OrthPowCorr	0.48	0.50	0.45	0.81	0.50	0.51	0.49	0.45	0.53	0.52	0.53	0.18
All	ASD	wPLI	0.47	0.50	0.44	0.79	0.57	0.53	0.61	0.009	0.53	0.50	0.55	0.19
Build separate models for adults, adolescents and children
Adults	ASD	PS	0.50	0.48	0.51	0.50	0.48	0.45	0.52	0.63	0.49	0.50	0.48	0.49
Adolescents	ASD	PS	0.49	0.47	0.51	0.60	0.50	0.50	0.51	0.50	0.48	0.46	0.50	0.56
Children	ASD	PS	0.53	0.51	0.55	0.29	0.47	0.51	0.44	0.68	0.43	0.45	0.42	0.84
Adults	ASD	OrthPowCorr	0.56	0.63	0.49	0.09	0.50	0.48	0.53	0.44	0.47	0.50	0.45	0.76
Adolescents	ASD	OrthPowCorr	0.45	0.51	0.39	0.78	0.50	0.49	0.52	0.40	0.53	0.60	0.46	0.29
Children	ASD	OrthPowCorr	0.60	0.48	0.71	0.07	0.49	0.47	0.51	0.66	0.46	0.44	0.47	0.77
Adults	ASD	wPLI	0.54	0.63	0.45	0.30	0.50	0.45	0.55	0.52	0.52	0.53	0.52	0.31
Adolescents	ASD	wPLI	0.48	0.44	0.53	0.53	0.44	0.43	0.45	0.87	0.43	0.44	0.42	0.91
Children	ASD	wPLI	0.64	0.54	0.73	0.02	0.60	0.62	0.59	0.013	0.47	0.36	0.56	0.73
Use a narrower ASD definition
All	ASDn	PS	0.53	0.39	0.61	0.45	0.55	0.55	0.56	0.17	0.56	0.48	0.61	0.15
All	ASDn	OrthPowCorr	0.59	0.18	0.82	0.48	0.50	0.52	0.49	0.42	0.53	0.37	0.62	0.79
All	ASDn	wPLI	0.55	0.22	0.73	0.34	0.48	0.53	0.46	0.44	0.55	0.42	0.62	0.26
Evaluate other variations of EEG power spectral and functional connectivity algorithms
All	ASD	abs. PS	0.51	0.51	0.51	0.36	0.53	0.54	0.51	0.14	0.51	0.51	0.52	0.36
All	ASD	PowCorr	0.47	0.48	0.46	0.81	0.46	0.45	0.47	0.99	0.49	0.49	0.48	0.68
All	ASD	PLV	0.47	0.49	0.45	0.82	0.51	0.50	0.53	0.63	0.52	0.52	0.52	0.28
All	ASD	iCOH	0.48	0.53	0.42	0.81	0.46	0.45	0.48	0.87	0.49	0.51	0.47	0.65
All	ASD	COH	0.46	0.47	0.44	0.91	0.47	0.45	0.49	0.88	0.48	0.49	0.47	0.74

Accuracy (acc), sensitivity (sens), specificity (spec) and the p value derived from randomization tests are indicated. Classifiers with p < 0.05 are highlighted in bold. abs. PS refers to absolute power spectrum and was obtained in sensor space

Back to article page

ISSN: 2040-2392

Contact us

General enquiries: journalsubmissions@springernature.com