**3. Cortical activation during real walking and mental imagery of walking**

This section introduces two experiments in which we compared the cortical activation during real walking and that during mental imagery of walking.

#### **3.1 Experiment 1**

We compared the activation in the primary motor area and the primary somatosensory area of the brain during RW and WO by means of fNIRS. The possibility to quickening walking rehabilitation by mental imagery of walking is discussed based on the experiment results.

#### **3.1.1 Subjects**

Two male graduate students (SL and ZJ) of Kochi University of Technology participated in the experiment. Their ages were 27 to 28 years. All were right-handed and had no medical history of neurological or psychiatric disorders.

#### **3.1.2 fNIRS measurement**

As shown in Fig. 5, the optodes were fixed with a plastic shell and placed on the subject's left frontal areas according to the international 10-20 system, which is widely used in EEG measurement (Homan et al., 1987; Steinmetz et al., 1989; Okamoto et al., 2004). The numbers between the emitters and detectors were channel numbers. Channels were labelled from front to back as CH1-CH24. Data were measured with a sampling rate of 10 Hz. CH16 was placed on Cz. The areas measured by CH11, 12, 15, 18, 19 and CH13, 14, 17, 20, 21 covered the primary somatosensory cortex which processes the sensory information and the primary motor cortex which plans and executes movements related to walking.

#### **3.1.3 Stimuli**

The experiment scene is shown in Fig. 6. During the RW task, the subjects walked along a straight line at 0.3 m/s. The ETG-7100 system was pushed to follow the subject

Fig. 5. Schematic placement of the emitter and detector optodes on the subject's head

(a) Real walking (b) Walking observation

138 Infrared Spectroscopy – Life and Biomedical Sciences

changes independently. The distance between the detector optode and emitter optode was 30 mm, which enabled cerebral blood volume measurement at a 2 to 3 cm depth from the surface of cerebral cortex (Toronov et al., 2001). The midpoints of pairs of emitter-detector optodes were regarded as the points of measurement (channels). Data were measured with a sampling rate of 10 Hz. Light emitters and detectors were alternated at an equal distance of 3 cm to give one 4×4 optode probe sets (Fig. 4). All of the transmitted intensities of the two wavelengths that left the tissue were continuously recorded over 24 channels to

Data collection by this system is comfortable for subjects, since it requires less constrictive circumstances of measurements and fewer movement restrictions, yielding more ecologically valid conditions. The whole system is fixed on a platform installed with casters so that the system can move with the subject in moving tasks. In our study, the subjects walked in the experiment. Therefore fNIRS is more suitable for this study (Suzuki et al.,

**3. Cortical activation during real walking and mental imagery of walking** 

This section introduces two experiments in which we compared the cortical activation

We compared the activation in the primary motor area and the primary somatosensory area of the brain during RW and WO by means of fNIRS. The possibility to quickening walking rehabilitation by mental imagery of walking is discussed based on the experiment results.

Two male graduate students (SL and ZJ) of Kochi University of Technology participated in the experiment. Their ages were 27 to 28 years. All were right-handed and had no medical

As shown in Fig. 5, the optodes were fixed with a plastic shell and placed on the subject's left frontal areas according to the international 10-20 system, which is widely used in EEG measurement (Homan et al., 1987; Steinmetz et al., 1989; Okamoto et al., 2004). The numbers between the emitters and detectors were channel numbers. Channels were labelled from front to back as CH1-CH24. Data were measured with a sampling rate of 10 Hz. CH16 was placed on Cz. The areas measured by CH11, 12, 15, 18, 19 and CH13, 14, 17, 20, 21 covered the primary somatosensory cortex which processes the sensory information and the primary

The experiment scene is shown in Fig. 6. During the RW task, the subjects walked along a straight line at 0.3 m/s. The ETG-7100 system was pushed to follow the subject

motor cortex which plans and executes movements related to walking.

estimate changes in the concentrations of oxy-Hb and deoxy-Hb.

during real walking and that during mental imagery of walking.

history of neurological or psychiatric disorders.

2004; Miyai et al., 2001).

**3.1 Experiment 1** 

**3.1.1 Subjects** 

**3.1.3 Stimuli** 

**3.1.2 fNIRS measurement** 

in the task. During the WO task, the subjects were shown a video (Fig. 7) in which a person walks along a line at the same speed (0.3 m/s). The subjects were instructed to imagine that they were walking the same as the person in the video, especially to match their gaits to the video, while keep standing on the ground. The video was shot in a corridor against the wall.

### **3.1.4 Task paradigm**

The experiment procedure is shown in Fig. 8. During the experiment, a 20 s real walking task and a 20 s imaginary walking task were performed, with a 60 s rest period before and after each task. In order to avoid the influence by the order of the tasks, two procedures were conducted. In procedure 1, the real walking task was before the imaginary walking task while in procedure 2 the order was reversed.

Fig. 6. Experiment scene

Comparison of Cortical Activation During Real Walking and Mental Imagery of

and CH13, 14, 17, 20, 21 between the RW task and the WO task.

0 9 29 48 41 Time (s)

0 9 29 48 41 Time (s)

(a) Imaginary walking (b) Real walking

**3.1.6 Results and discussions** 


Fig. 10. Results of procedure 1


Fig. 11. Results of procedure 2

walking was implied.




0

0

0.2

0.2

Hb (mmol-mm)

0.4

0.4




0

0

0.2

0.2

Hb (mmol-mm)

0.4

0.4

Walking – The Possibility of Quickening Walking Rehabilitation by Mental Imaginary of Walking 141

was applied to the data between these two baselines. The data in the statistical period, which was made up of task and recovery periods, were considered as the activation signals which were used to analyze the cerebral blood volume difference between real walking and imaginary walking. In the analysis we compared the average oxy-Hb of CH11, 12, 15, 18, 19

(a) Real walking (b) Imaginary walking

The average oxy-Hb of subject SL and ZJ during the statistical period of procedure 1 and 2 are shown in Fig. 10 and Fig. 11, respectively. In Fig. 10 and Fig. 11, the horizontal axis is the time shown in Fig. 9. Both oxy-Hb and deoxy-Hb, and their sum total-Hb are shown. It can be seen from the experiment results that both in procedure 1 and procedure 2, oxy-Hb increased significantly no matter the subject really walked or just imagined walking. The experiment results suggest that the cortical areas related to walking is activated by both RW and WO. The possibility of brain exercise for walking rehabilitation by mental imagery of





0

0

0.2

0.2

Hb (mmol-mm)

0.4

0.4



0 9 29 48 41 Time (s)

0 9 29 48 41 Time (s)



0

0

0.2

0.2

Hb (mmol-mm)

0.4

0.4

Fig. 7. Walking video shown to the subjects during WO task

Procedure 1 Procedure 2

Fig. 8. Walking video shown to the subjects during WO task

#### **3.1.5 Data analysis**

According to previous studies, oxy-Hb is more useful for analysis of cortical activation than deoxy-Hb because of its higher reproducibility (Plichta 2006; Sato 2005), lower inter-subject variability (Sato 2006), and higher correlation with fMRI signals. Therefore, in the present study, oxy-Hb was measured as an indicator of changes in blood volume. The measurements of the subjects were checked visually for artifacts due to body movements. Obtained data were analyzed using the "integral mode" in the ETG-7100 software. We defined a 48 s measurement block made up of a 9 s pretask baseline period before task period (20 s), a 12s recovery period and a 7 s posttask baseline period (Fig. 9). Linear fitting

Fig. 9. Parameters of the integral mode in the ETG-7100 analysis software

was applied to the data between these two baselines. The data in the statistical period, which was made up of task and recovery periods, were considered as the activation signals which were used to analyze the cerebral blood volume difference between real walking and imaginary walking. In the analysis we compared the average oxy-Hb of CH11, 12, 15, 18, 19 and CH13, 14, 17, 20, 21 between the RW task and the WO task.
