Real-time assessment of hypnotic depth, using an EEG-based brain-computer interface: a preliminary study

Obukhov, Nikita V.; Naish, Peter L.N.; Solnyshkina, Irina E.; Siourdaki, Tatiana G.; Martynov, Ilya A.

doi:10.1186/s13104-023-06553-2

BMC Research Notes

Table 1 Deep hypnosis and wakefulness classification accuracy according to the band used to train the models

From: Real-time assessment of hypnotic depth, using an EEG-based brain-computer interface: a preliminary study

Patient	Session	Accuracy for the corresponding band, %
Patient	Session	1.5–45 Hz	1.5-8 Hz	1.5–14 Hz	4–15 Hz
(A) The 10-fold cross-validation test results, M ± SD
A	1	96 ± 8	86 ± 18	98 ± 6	94 ± 9.17
	2	93.5 ± 10.01	95.5 ± 9.07	86.5 ± 14.33	72.5 ± 26.48
	3	88 ± 18.33	72 ± 31.24	86 ± 25.38	86 ± 25.38
	4	95 ± 15	83 ± 15.84	95 ± 10	95 ± 10
	5	94 ± 9.17	89.67 ± 17.16	92.33 ± 13	96.33 ± 7.37
	6	98 ± 6	87 ± 15.52	92 ± 13.27	96 ± 8
	7	84.17 ± 21.87	93.33 ± 13.33	83.33 ± 22.36	86.67 ± 16.33
	M ± SD	92.67 ± 4.86	86.64 ± 7.75	90.45 ± 5.32	89.5 ± 8.65
E	1	97.5 ± 7.5	82 ± 15.68	88.5 ± 16.13	95 ± 10
	2	100 ± 0	90.5 ± 11.72	93 ± 15.52	86.5 ± 19.11
	3	100 ± 0	91 ± 11.14	100 ± 0	100 ± 0
	4	95 ± 10	93 ± 15.52	91 ± 15.78	92 ± 24
	6*	100 ± 0	96 ± 8	94 ± 18	96 ± 8
	M ± SD	98.5 ± 2.24	90.5 ± 5.22	93.3 ± 4.3	93.9 ± 5.03
G	1	94.33 ± 8.7	96 ± 8	98.33 ± 5	96 ± 12
	2	98.33 ± 5	92.33 ± 18.14	94.67 ± 11.08	93 ± 15.52
	3	100 ± 0	100 ± 0	100 ± 0	96.33 ± 7.37
	4	98 ± 6	98 ± 6	98 ± 6	98 ± 6
	M ± SD	97.67 ± 2.39	96.58 ± 3.27	97.75 ± 2.23	95.83 ± 2.08
S	1	100 ± 0	100 ± 0	97.5 ± 7.5	95.5 ± 9.07
	2	97.5 ± 7.5	100 ± 0	100 ± 0	100 ± 0
	3	97.5 ± 7.5	97.5 ± 7.5	97.5 ± 7.5	97.5 ± 7.5
	M ± SD	98.33 ± 1.44	99.17 ± 1.44	98.33 ± 1.44	97.67 ± 2.25
O	1	100 ± 0	100 ± 0	97.5 ± 7.5	97.5 ± 7.5
	2	100 ± 0	100 ± 0	97.5 ± 7.5	87.5 ± 16.78
	M ± SD	100 ± 0	100 ± 0	97.5 ± 0	92.5 ± 7.07
N	1	96 ± 8	98 ± 6	94 ± 18	98 ± 6
	2	100 ± 0	100 ± 0	100 ± 0	100 ± 0
	M ± SD	98 ± 2.83	99 ± 1.41	97 ± 4.24	99 ± 1.41
V	1	91 ± 15.78	93.5 ± 10.01	98 ± 6	95.5 ± 9.07
	2	100 ± 0	98 ± 6	100 ± 0	100 ± 0
	M ± SD	95.5 ± 6.36	95.75 ± 3.18	99 ± 1.41	97.75 ± 3.18
C	1	100 ± 0	95 ± 15	81.67 ± 22.91	96.67 ± 10
	2	100 ± 0	77.5 ± 29.36	93.33 ± 20	93.33 ± 20
	M ± SD	100 ± 0	86.25 ± 12.37	87.5 ± 8.24	95 ± 2.36
(B) Results based on the data from the second (and subsequent) sessions
A	2	74.31	52.08	74.65	66.32
	3	80.0	75.34	91.51	89.86
	4	99.48	64.83	87.66	88.98
	5	69.09	66.94	94.62	91.94
	6	90.85	69.82	90.24	87.8
	7	79.73	66.32	80.76	73.54
	M ± SD	82.24 ± 11.12	65.89 ± 7.72	86.57 ± 7.48	83.07 ± 10.52
E	2	54.75 (98.42**)	38.29	70.89	80.38
	3	54.18 (72.76***)	65.63	53.56	83.9
	4	71.82	69.7	81.52	84.55
	6*	70.06	44.19	59.01	84.01
	M ± SD	62.7 ± 9.54	54.45 ± 15.53	66.24 ± 12.49	83.21 ± 1.9
G	2	87.10	91.13	91.13	73.92
	3	88.98	97.31	90.05	81.72
	4	75.78	94.3	89.46	87.75
	M ± SD	83.95 ± 7.14	94.25 ± 3.09	90.21 ± 0.85	81.13 ± 6.93
S	2	53.03	93.64	95.76	53.64
	3	94.3	87.03	78.16	75.95
	M ± SD	73.67 ± 29.18	90.34 ± 4.67	86.96 ± 12.45	64.8 ± 15.78
O	2	61.46	89.58	89.24	88.89
N	2	97.09	93.6	96.8	96.51
V	2	95.06	90.41	88.08	94.48
C	2	71.92	61.15	66.15	82.31
The average for all sessions, M ± SD		77.32 ± 14.89	74.28 ± 17.96	82.59 ± 12.53	82.44 ± 10.35

(A) The results of the 10-fold cross-validation test are shown as the average classification accuracy among the ten partitions of the marked file and the corresponding standard deviation
(B) This accuracy is calculated based on the data from the second and subsequent sessions as a percentage of coincidence between the states predicted by the model for different EEG epochs and the actual states related to these epochs
* Session 5 has been excluded from the analysis
**Value after a drift correction with a subtraction of 1.5 from the original feature vector (for details, see “Supplements D”)
*** Value after a drift correction with a subtraction of 2.5 from the original feature vector (for details, see “Supplements D”)

Back to article page

ISSN: 1756-0500

Contact us

Submission enquiries: bmcresearchnotes@biomedcentral.com
General enquiries: ORSupport@springernature.com