The basic principle of cortical oscillations-based neurofeedback training (NFT) is to visually and/or acoustically feed back cortical oscillations, which are typically recorded by EEG. NFT has previously been applied in a range of clinical and non-clinical conditions. For example, it has been reported that NFT is an effective tool to reduce the symptomatology in attention deficit hyperactivity disorder (Arns et al. 2009), the frequency of seizures in epileptic patients (Tan et al. 2009) and the number of awakenings in insomniac patients (Schabus et al. 2013). Furthermore, NFT has been shown to enhance music, acting and dance performance (Gruzelier 2013a), as well as microsurgical skills (Ros et al. 2009). There are also a few studies where NFT was applied to enhance athletic performance. Landers et al. (1991) examined whether EEG biofeedback training for approximately 60 min could improve archery performance, as well as self-reported measures of concentration and self-confidence. For this purpose, 24 experienced pre-elite archers were randomly assigned to one of three treatment conditions: (a) correct feedback (i.e., reduced left temporal activation), (b) incorrect feedback (i.e., reduced right temporal activation) and (c) control without NFT. The rationale for the NFT protocol was derived from previous studies suggesting that reduced activation of the left relative to the right hemisphere is associated with superior performance (see Sect. 3.1). Results indicated that only the correct feedback group significantly improved performance, while no changes were observed in the control group. In the incorrect feedback group, shooting accuracy was even worse after treatment (Landers et al. 1991). In a more recent study, Rostami et al. (2012) investigated the effect of fifteen 60-min sessions NFT on performance in expert rifle shooters. Participants in the intervention group received two different NFT protocols within each session. One protocol focused on simultaneous enhancement of the sensorimotor rhythm (SMR; 13–15 Hz) and inhibition of high beta activity (20–30 Hz). The other focused on modulation of the alpha (8–12 Hz)/theta (4–8 Hz) ratio combined with inhibition of high beta activity (20–30 Hz). These (or similar) protocols have been successfully applied in numerous NFT studies outside of the sporting arena and therefore, they represent rather standard NFT protocols. Compared to an untreated control group, shooting performance significantly improved in the NFT group (Rostami et al. 2012). Arns et al. (2008) studied a new method for golf performance enhancement employing personalised event-locked EEG neurofeedback during putting. Six amateur golfers received three real-life NFT sessions. Each session consisted of four series of 80 putts. However, feedback was provided only in the second and fourth series whereas series one and three served as a control condition. Target frequency bands and amplitudes for the individualised NFT were derived from a prior assessment comparing successful and unsuccessful putts. The overall percentage of successful putts was significantly larger in those series with feedback compared to the control (no feedback) series. The average improvement in performance with feedback on the personalised EEG profile was about 25 % (Arns et al. 2008). The results of this study may suggest that individual rather than generalised EEG profiles should be used for NFT. However, the subjects were amateur and not expert golfers. Their task-related EEG profiles may not only be different from those of experts but also, they may show larger variability/less stability. Therefore, it could be hypothesised that individualised NFT protocols may be particularly useful in less skilled athletes, whereas for experts there might be a higher chance to identify common EEG features of peak performance. Consistent with this, a preliminary study conducted by Muangjaroen and Wongsawat (2012) indicates that high-frequency alpha power at electrode C4, theta power at Fz, as well as theta and high alpha power at Pz are sufficient to calculate an index to predict successful golf putting Muangjaroen and Wongsawat (2012). However, caution is recommended before suggesting a particular NFT protocol for a particular sport and level of expertise. As indicated by the study of Landers et al. (1991), incorrect feedback (i.e. wrong NFT protocol) may even deteriorate performance.

The results presented in this chapter indicate that superior sport performance is associated with changes in cortical oscillation patterns. However, these changes depend on the specific demands of the sport. A good example of this is the relationship between high-frequency alpha power and performance in golf versus shooting. High-frequency alpha power at electrodes overlying sensorimotor cortical areas was reduced during successful compared to unsuccessful putting in expert golfers (Babiloni et al. 2007) but increased during high compared to low shooting scores in expert pistol shooters (Del Percio et al. 2009). Alpha desynchronisation prior to putting could be interpreted to reflect an increase in cortical activation in order to facilitate movement execution (Pfurtscheller et al. 1996), whereas alpha synchronisation during aiming prior to the shot may indicate inhibitory processes (Hummel et al. 2002) to avoid aim point fluctuations. Future experiments may manipulate conditions (e.g. distance from the target, target size, time to complete the task, dual tasking) to gain a better understanding of the processes underlying changed EEG profiles during superior performance. Furthermore, more longitudinal studies are needed and tasks other than golf putting and shooting should be studied. It is also recommended to consider multimodal approaches for a more comprehensive view.

Skilled athletic performance often requires a substantial amount of action monitoring and adequate error detection on a subsecond timescale. It is suggested this is processed in the theta frequency band. Consistent with this, studies revealed that theta oscillations are linked to balance control (Slobounov et al. 2009, 2013), as well as performance in golf (Baumeister et al. 2008; Kao et al. 2013) and pistol shooting (Doppelmayr et al. 2008). Future research on balance control may identify characteristic changes in brain activity that are related to the difficulty of balance tasks and to balance performance. Furthermore, adaptations to different balance training programs on a behavioral and cortical level may help to better identify optimal balance exercises to improve performance in sports.

To this point of time, only few studies are published that applied EEG neurofeedback to enhance athletic performance. These studies revealed that NFT has the potential to improve the performance of archers (Landers et al. 1991), golfers (Arns et al. 2008) or shooters (Rostami et al. 2012). Beyond that, NFT has been associated with positive effects on cognitive and affective outcome measures (Gruzelier 2013b), creativity (Gruzelier 2013a), fine motor skills (Ros et al. 2009), reaction time (Doppelmayr and Weber 2011), as well as sleep quality and memory (Schabus et al. 2013). These results suggest that NFT could be a useful tool to directly or indirectly (e.g. via improved sleep) enhance athletic performance.