Table 1 Characteristics of fast spindles (frequency > 13 Hz) detected by the SUMOv2 model in the BD recordings for healthy controls (HC) and patients with bipolar disorder (BP).

From: From sleep staging to spindle detection: a case study on end-to-end automated sleep analysis

 

Channel

HC

BP

p-value

Density

F3-A2

3.43 (1.70)

2.15 (1.46)

0.008

F4-A1

3.39 (1.68)

2.16 (1.39)

0.009

C3-A2

4.37 (2.01)

2.80 (1.98)

0.011

C4-A1

4.37 (2.00)

2.77 (1.91)

0.008

Frequency

F3-A2

13.91 (0.14)

13.91 (0.18)

0.990

F4-A1

13.92 (0.13)

13.92 (0.12)

0.983

C3-A2

13.98 (0.22)

13.90 (0.18)

0.184

C4-A1

13.98 (0.21)

13.88 (0.17)

0.089

Duration

F3-A2

0.87 (0.09)

0.82 (0.12)

0.125

F4-A1

0.87 (0.09)

0.82 (0.12)

0.150

C3-A2

0.88 (0.09)

0.83 (0.13)

0.154

C4-A1

0.88 (0.10)

0.83 (0.13)

0.203

Amplitude

F3-A2

6.95 (1.49)

6.99 (1.43)

0.910\(^\dag\)

F4-A1

7.12 (1.40)

7.06 (1.39)

0.992\(^\dag\)

C3-A2

8.36 (1.73)

8.32 (2.13)

0.703\(^\dag\)

C4-A1

8.41 (1.88)

8.47 (1.61)

0.900

  1. While spindles were aggregated across all channels, we calculated individual spindle characteristics separately for each channel using only that channel’s EEG signal. P-values were calculated between the characteristics of the two cohorts with Welch’s two-sided independent t-tests unless indicated otherwise. We use \(\dag\) to indicate that a Wilcoxon two-sided rank-sum test was used.