Investigation of Frontal Lobe Activation with fNIRS and Systemic Changes During Video Gaming

Frontal lobe activation caused by tasks such as videogames can be investigated using multichannel near-infrared spectroscopy (fNIRS), sometimes called optical topography. The aims of this study are to investigate the effects of video gaming (fighting and puzzle games) in the brain and the systemic physiology and to determine whether systemic responses during the gaming task are associated with the measurement of localised cerebral haemodynamic changes as measured by fNIRS. We used a continuous-wave 8-channel fNIRS system to measure the changes in concentration of oxy-haemoglobin (HbO2) and deoxy-haemoglobin (HHb) and changes in total haemoglobin (ΔtHb = ΔHbO2 + ΔHHb) over the frontal lobe in 30 healthy volunteers. The Portapres system was used to measure mean blood pressure (MBP) and heart rate (HR), and a laser Doppler was employed to measure the changes in scalp blood flow (or flux). Even though we observed significant changes in systemic variables during gaming, in particular in scalp flow, we also managed to see localised activation patterns over the frontal polar (FP1) region. However, in some channels over the frontal lobe, we also observed significant correlations between the HbO2 and systemic variables.


Introduction
Multichannel functional near-infrared spectroscopy (fNIRS), or optical topography (OT), is often employed to detect brain functional activation. fNIRS measures the changes in brain tissue concentrations of oxy -haemoglobin (HbO 2 ) and deoxyhaemoglobin (HHb) that occur secondary to the brain electrical activity changes due Chapter 13

Investigation of Frontal Lobe Activation with fNIRS and Systemic Changes During Video Gaming
to the activation task. The fNIRS haemodynamic changes should occur at specifi c locations that overlay the cortical activated areas and should be closely coupled to the task-related timing periods. This assumes that the functional haemodynamic task-related changes are occurring on top of an unchanged global systemic and brain resting state. However, in certain functional experiments, these assumptions are not accurate and can lead to false positives in fNIRS [ 1 ]. We have previously used fNIRS to monitor the frontal and prefrontal cortex during anagram-solving tasks and observed signifi cant task-related changes in mean blood pressure (MBP), heart rate (HR) and scalp blood fl ow (fl ux) that correlated with the fNIRS signals [ 2 , 3 ]. In a recent study using multichannel fNIRS to produce maps of the haemodynamic response during anagram solving while simultaneously monitoring the systemic physiology, we observed a number of fNIRS channels to be highly correlated with activation-related systemic changes leading to false-positive cortical activation locations [ 1 ]. In our latest study utilising both fNIRS and functional magnetic resonance imaging (fMRI) and angiography during frontal activation tasks, we observed a signifi cant correlation between the changes in HbO 2 and the systemic activation response of the deep scalp veins [ 4 ].
Several earlier studies used fNIRS to investigate the effect of videogames over the frontal and prefrontal lobe [ 5 , 6 ]; however, none of these studies investigated the systemic changes during this type of activation task. The main aims of this study are to determine whether there are signifi cant systemic changes during video gaming and if these changes are signifi cantly associated with the fNIRS haemodynamic measurements.

Methods
We used two commercially available videogames for the Game Boy Advance SP (Nintendo Corp. Japan), a 'fi ghting' game (Final Fight One, Capcom) and a 'puzzle' game (Polarium Advance, Nintendo). The former is an arcade game where the player can choose a hero and fi ght against different enemies in order to complete specifi c missions. The latter is a very simple puzzle game, where the player has to fl ip black or white tiles on a square board in order to create horizontal rows of one colour and erase all the tiles in a single stroke to clear the board.
We studied two groups of healthy young volunteers, most of whom had some previous experience in video gaming. The fi rst group ( n = 17, mean age 24 years) did the puzzle game and the second group ( n = 13, mean age 24 years) did the fi ghting game. These studies were approved by the Research Ethics Committee of UCL.
In order to become familiar with the experimental environment, the volunteers were given the game to practise for about 10 min. Following that, each subject sat in front of a desk on which a computer monitor was placed to alert the subjects via a visual stimulus when to rest and when to start playing the game. The protocol involved a single block of 5 min playing the game continuously (activation period) with a 2-min period of rest before and after the activation block.
A continuous-wave (CW) 8-channel fNIRS system, the Oxymon Mk III (Artinis Medical Systems BV, The Netherlands), was used. This system measures the changes in light attenuation at two wavelengths, 764 nm and 858 nm, and utilises the modifi ed Beer-Lambert law with an age-dependent differential path length factor (DPF) to resolve the concentration changes in oxy (HbO 2 )-and deoxy (HHb)haemoglobin and calculate the changes in total haemoglobin (tHb) which is the summation of ΔHbO 2 and ΔHHb. The optode (source-detector fi bre) confi guration used in this study was the 8-channel split, which allows eight channel recordings with an inter-optode distance of 40 mm. The optode template was placed on the volunteer's forehead, using the international 10/20 system of electrode placement [ 7 ], such that (i) FP1 region was covered between light-emitting fi bres (Tx2, Tx4a) and light-receiving fi bre (Rx2), corresponding to channels 7 and 8, and (ii) FP2 region was covered between light-emitting fi bres (Tx2, Tx3b) and light-receiving fi bre (Rx1), corresponding to channels 5 and 6 ( Fig. 13.1 ).
The Portapres system (Finapres Medical Systems) was employed, with the infl atable cuff placed on the index fi nger of the left hand, in order to measure the mean blood pressure (MBP) and the heart rate (HR). The laser Doppler (Moor Instruments) was used to measure the scalp blood fl ow (fl ux) with the laser probe placed on the forehead.
In order to locate the activation channels, the activation period was split in three separate epochs each one having a duration of 10 s. The fi rst epoch was immediately after the beginning of the stimuli (120-129 s); the second was in the middle of the activation period (270-279 s) and the third was at the end of the activation (411-420 s). For each epoch, we calculate a mean value for all measurements and subtracted that from a 10-s mean calculated at the beginning of the rest period (1-10 s). The difference was then compared to zero using a Student's t -test to assess the signifi cance ( p ≤ 0.05). We then defi ned activation as a signifi cant increase in HbO 2 , a signifi cant decrease or no change in HHb and a signifi cant increase in tHb [ 3 ]. Further, we estimated the correlation between the fNIRS and systemic signals to assess the relation of the brain haemodynamic and systemic signals. No correlation was defi ned as 0.25 > r > −0.25.   (Table 13.1 ).

Results
Correlation analysis revealed signifi cant correlations between the HbO 2 signal and systemic variables (Table 13.2 ). For the fi ghting game task, a signifi cant correlation was seen at 40 % of channels between HbO 2 and MBP, 42 % of channels between HbO 2 and HR and 46 % of channels between HbO 2 and fl ux. For the puzzle game task, a signifi cant correlation was seen at 57 % of channels between HbO 2 and MBP, 41 % of channels between HbO 2 and HR and 70 % of channels between HbO 2 and fl ux (Table 13.2 ).

Discussion
We found signifi cant localised changes in HbO 2 and HHb measured over the frontal lobe during the gaming tasks. In addition, during the fi ghting game, our group analysis revealed signifi cant changes in MBP and HR occurring at the middle and later periods of the activation block. We observed during both the fi ghting and puzzle 4.4( ± 5.5)* 7.9 ± (9.2)* 7 ± (9.1)* 7.5( ± 11.3)* 12.5( ± 19)* 11.4( ± 17)* games signifi cant changes in the fl ux signal throughout the activation period. Correlation analysis between the HbO 2 and HHb with the systemic measurements revealed some individual and channel variability with most of the HbO 2 measurements correlating positively with fl ux in both the fi ghting and the puzzles game.
There was a variation in the results among the volunteers between the two games. More haemodynamic activation patterns were seen during the puzzle game studies in the third epoch of the activation block compared to the other epochs, whereas in fi ghting game studies, these patterns were seen in the fi rst period, just after the start of the activation block. In addition, during the fi ghting game studies, the percentage of activation patterns gradually decreases over time with the third epoch demonstrating the lowest percentage of volunteers showing activation. Previous studies reported that playing videogames signifi cantly increased systolic and diastolic pressure, heart rate and oxygen consumption in adolescents [ 8 ], in particular during the fi ghting game, which is a very dynamic game (the characters are moving, jumping and punching) that can possibly cause more areas in the brain to be activated (such as the motor cortex). The action involved in such a game can also cause emotional reactions in the subject. In a study exploring if playing a game that contains violence causes any sympathetic or parasympathetic reactions, the investigators concluded that a violent game causes different autonomic responses and affects heart rate and heart rate variability, which is a measure of stress reactivity [ 9 ]. In our studies, the high correlation coeffi cients found between fNIRS channels and systemic variables give an indication that some changes in fNIRS signals are due to changes in systemic variables and not due to haemodynamic changes originating from specifi c regions of the frontal lobe. Therefore, during analysis of brain activation during tasks such as video gaming, the contribution of the systemic changes should be taken into consideration.