*2.4. Surgical Preparation and Hemodynamic Measurements*

The animals were initially sedated with intramuscular injections of 5 mg/kg of tiletamine hydrochloride and zolazepam hydrochloride (Zoletil, Virbac, France) and 2 mg/kg of Xylazine (Rompun, Bayer, Korea), followed by inhaled isoflurane at a dose of 1–1.5%. Endotracheal intubation was performed on the sedated animals, and a capnography (Capstar-100, CWE Inc., Ardmore, PA, USA) was installed. Mechanical ventilation was initiated. To continue the anesthesia, a tidal volume of 12 mL/kg, respiratory rate of 10 breaths/min, partial pressure of arterial carbon dioxide at approximately 40 mmHg, and partial pressure of arterial oxygen over 80 mmHg were maintained.

An implantable perivascular probe (MA2PSB, Transonic Systems, Ithaca, NY, USA) combined with a perivascular flowmeter (T420, Transonic Systems, Ithaca, NY, USA) was placed on the internal carotid artery to measure the CBF. A pressure catheter (Mikro-tip pressure catheter, Millar, Houston, TX, USA) was inserted into the left femoral artery and placed in the descending thoracic aorta to measure the arterial blood pressure. Another Mikro-tip pressure catheter was inserted into the right atrium to measure the right atrial pressure. The ECG and saturation of percutaneous oxygen were also measured. All signals except EEG were gathered and saved in a data acquisition system (PowerLab 16/35, ADInstruments, Dunedin, New Zealand) simultaneously.

A pace-making wire was inserted into the right ventricle through the central vein catheter. Isoflurane was stopped before inducing VF to recover EEG signal. EEG started to recover, and appeared similar to the recording obtained before the injections. Then, a direct-current shock was applied to induce VF. Mechanical ventilation was halted, and the animals were left without assistance for 1 min. Thereafter, CPR and defibrillation attempts were executed, and manual ventilation using a resuscitator bag (Ambu Resuscitators, Ambu A/S, Ballerup, Denmark) was initiated to provide positive pressure ventilation to the animal at a rate of once per 10 compressions.

## *2.5. EEG Measurement*

A portable single-channel digital electroencephalograph and disposable surface electrodes (MT100, Kendall Healthcare, Toronto, Ontario, Canada) were attached to measure the scalp EEG under the referential montage. Reference and ground electrodes were attached on either side of the mastoid. Active electrodes connected to the device were placed on the forehead (Figure 1b). The raw EEG signal was bandpass filtered with a frequency range of 0.5–47 Hz and amplified with a gain of 12,000 *v*/*v*. The amplified signal with a low noise level under ±3 µVp-p was digitized and transmitted to the laptop via Bluetooth communication at a rate of 250 Hz. The data acquisition software in the laptop receives and saves the EEG data.

## *2.6. Data Processing*

All data were processed using MATLAB (MATLAB R2017b, Mathworks, Natick, MA, USA). The EEG and hemodynamic data were synchronized. Approximately 3-s-long pauses right before the defibrillation shocks were selected for analysis. The selected EEG was segmented into three 2-s-long sub-epochs with 1.5-s overlaps to reduce variation; 0–2 s, 0.5–2.5 s, and 1–3 s period. The representative EEG parameters were obtained from the average of three sub-epochs. Segmenting the EEG and obtaining parameters is similar to the signal processing technique for the BIS monitor [24]. Time and frequency domain parameters and entropy indices were obtained in this manner. All parameters considered are listed in Table 1.



Abbreviation: Pa–b Hz, the sum of spectral power from a–b Hz; Ba–b Hz, the sum of bispectral activity from a–b Hz; *p*(*xi*), probability distribution function of signal *x<sup>i</sup>* ; α of Rényi entropy was 0.5.

#### *2.7. Data Analysis*

First, CBF recovery during CPR were analyzed to investigate their relationship with resuscitation rates. The recovery rate was defined as a relative scale of each hemodynamic parameter with respect to the baseline value in the pre-VF state. Second, the EEG waveforms were scrutinized according to the test scenario. EEG activity was evaluated, along with the recovery of CBF.

Pearson correlation coefficients between each EEG parameter and the recovery rates of CBF for all experiments were obtained to inspect whether EEG parameters show similar changes with the CBF. In addition, the recovery rates of CBF were categorized into four quartile groups: group 1 (<25%); group 2 (25–50%); group 3 (50–75%), and group 4 (>75%). Averages of each EEG parameter among groups were evaluated through one-way analysis of variance (ANOVA). Significance was considered at a level of *p* < 0.05. Receiver operating characteristic (ROC) curve analysis was also performed to measure the optimal cut-off values of EEG parameters, to discriminate between the higher and the lower group of the CBF recovery based on the median value, which was approximately 30%. These tests were performed with SPSS (SPSS Statistics 23, IBM SPSS Statistics, New York, NY, USA).

#### **3. Results**
