Individual alpha neurofeedback training effect on short term memory

Abstract

Memory performance has been reported to be associated with electroencephalogram (EEG) alpha activity. This study aimed to improve short term memory performance by individual alpha neurofeedback training (NFT). With appropriate protocol designed for NFT, the experimental results showed that the participants were able to learn to increase the relative amplitude in individual alpha band during NFT and short term memory performance was significantly enhanced by 20 sessions of NFT. More importantly, further analysis revealed that the improvement of short term memory was positively correlated with the increase of the relative amplitude in the individual upper alpha band during training. In addition, effective strategies for individual alpha training varied among individuals and the most successful mental strategies were related to positive thinking.

Introduction

The relation between memory performance and electroencephalogram (EEG) alpha activity has been reported by several studies in the last two decades (Klimesch, 1996, Klimesch, 1999, Klimesch et al., 2006). It has been observed that memory performance is positively related to resting power in the alpha band (Klimesch, 1999). One interesting question from the association between memory and alpha is, whether short term memory performance can be improved through the enhancement of alpha power by means of neurofeedback training (NFT).

NFT is an operant conditioning procedure in which individuals learn to self-regulate their brain activity. During NFT, the EEG is recorded from one or more electrodes placed on the scalp and the relevant components are extracted and fed back using an online feedback loop in the form of audio, visual or combined audio-visual information (Vernon, 2005). A number of studies have demonstrated the positive effects of NFT on treatment of psychological disorders such as attention deficit hyperactivity disorder (Friel, 2007, Heinrich et al., 2007), substance use disorder (Sokhadze et al., 2008), epilepsy (Sterman and Egner, 2006, Egner and Sterman, 2006) and autistic spectrum disorder (Coben et al., 2010). Besides clinical applications, the benefits of NFT have also been reported in enhancement of cognitive and artistic performance in healthy individuals (Egner and Gruzelier, 2003, Gruzelier et al., 2006, Gruzelier, 2009). For instance, semantic working memory can be enhanced by sensorimotor rhythm (SMR) NFT (Vernon et al., 2003) or by individual upper alpha NFT (Escolano et al., 2011). For mental rotation ability, it can be improved by NFT using SMR (Doppelmayr and Weber, 2011) or using individual upper alpha (Hanslmayr et al., 2005, Zoefel et al., 2011). In addition, increasing peak alpha frequency (PAF) by NFT can improve the cognitive processing speed and executive function in the elderly individuals (Angelakis et al., 2007).

Even though an increasing amount of studies reported positive effects of NFT in some fields, there is still a lack of successful NFT for the improvement in short term memory which refers to the temporary storage of information without the attention component of working memory (Conway et al., 2002, Stipacek et al., 2003). The early work by Bauer (1976) investigated the effect of fixed alpha frequency band (8.5–12.5 Hz) NFT on short term memory in young adults. The participants were asked to produce alpha as long as possible during the training. Short term memory performance was assessed by a verbal free recall task and a digit span task. Although a significant increase in the percentage alpha was achieved by the end of the four sessions, neither digit span task nor verbal free recall task showed improvement. Vernon (2005) stated that this failure could be explained by the use of a fixed frequency band instead of individual alpha frequency range as proposed by Klimesch (1999). Due to the large inter-individual differences in alpha frequency band, the personalization of alpha band is the only way to observe and interpret any findings that involve alpha frequency band (Klimesch, 1999, Klimesch et al., 2003). Another possible reason for the failure in Bauer (1976) is that the feedback parameter was the percentage alpha in terms of time instead of the alpha power that is related to memory performance (Klimesch, 1999).

Moreover, the mental strategies used during the NFT have not been provided with sufficient details in the literature. Angelakis et al. (2007) only described two subjects' self-report during alpha amplitude NFT. Suggestions on mental strategies to be used during NFT shall contribute greatly to increase the rate of successful NFT learning.

Considering the relation between memory performance and alpha power as well as the importance of personalization of alpha band (Klimesch, 1999), this study aimed to improve short term memory performance by individual alpha NFT. More specifically, we attempted to answer three questions, namely, (a) whether NFT can increase the relative amplitude in individual alpha band, (b) whether this increase in individual alpha band is correlated with short term memory improvement, and (c) which mental strategies are more effective for the training.