Monitoring Neurological Injury by qEEG - Quantitative EEG Analysis Methods and Clinical Applications - page 174

I II III IV

10

0

− 10

0 2000 4000 6000 8000

2.5

Shannon

2

1.5

0 2000 4000 6000 8000

1.4

Tsallis (q=1.5)

1.2

1

0 2000 4000 6000 8000

0.5

Tsallis (q=3)

0.48

0.46

0 2000 4000 6000 8000

0.26

Tsallis (q=5)

0.24

0 2000 4000 6000 8000

Figure 7.2 An EEG trace composed by sequencing (I…IV), going from (I) Gaussian, (II) uniform dis-
tributed noises, and segments of EEG from experiments: baseline (III) and early recovery EEG after
brain injury (IV and V). The Shannon and Tsallis entropy for different q index values ( q 1.5, 3, 5) are

shown. Tsallis entropy is sensitive enough to distinguish between signals with different probability
distribution functions and to differentiate the baseline EEG from early recovery EEG. ( From: [15]. ©
2003. Biomedical Engineering Society. Reprinted with permission.)

=

Figure 7.3 Block diagram of EEG signal processing using the conventional Shannon entropy (SE)
measure and the proposed information quantity (IQ) measure.