Emotion-centered imagery training is seen as a powerful tool and is widely used in modern cognitive psychotherapy. But almost all the knowledge achieved during its six decades of history is associated with guided imagery, meanwhile research exploring the efficacy of self-guided is still very limited. Estimation of the potential of self-guided imagery, for example- if the regular practice of imagery containing a positive scenario, created in accordance with current needs can have a beneficial effect on the emotional condition—is a question of practical importance. If the answer is positive, this might offer new strategies to combat negative emotions in our daily lives. Here we investigated the impact of self-guided positive imagery training on emotions and brain functions.

As originally stated in the Lang's bio-informational theory of emotional imagery (Lang, 1977, 1979), mental imagery differs from verbal thought in that only mental imagery has the capacity to activate physiological and behavioral response system (Lang, 1987). It has been shown significant heart rate acceleration, relative to baseline during mental imagery, but not during verbal repetition of fearful scenarios (Vrana et al., 1986; Cuthbert et al., 2003), suggesting the imagery has easier access to our emotional system than the information provided by other sources, such as verbal (Holmes and Mathews, 2005). Thus, imagery can easily provoke an emotional disturbance. On the other hand, all mental disorders are accompanied by any dysfunction in image generation, such as: A deficit or excess in generation of images, an imbalance in the generation of positive and negative images, image distortion. A classical example of a severe violation in the imagery system is the post-traumatic stress disorder, where the presence of post-traumatic flashbacks is considered hallmark of the disorder (Ehlers et al., 2002). The potential of imagery systems is less effective when one is concerned, rather than when there is positive thinking (Hirsch et al., 2012). This deficiency is more pronounced in people with generalized anxiety (GAD) than in a control group (Hirsch et al., 2012). According to a study of Morina et al. (2011), the presence of anxiety was associated with a higher ability to generate bright images for potential negative scenarios and a deterioration of the ability to generate positive images (also true for depressive individuals) compared with control group. Likewise, as shown by Holmes et al. (2008), the imagery abilities were also related to the different personality traits, such as high dysphoria (compared to low) was associated with worsening ability to vividly imagine positive but not negative future events. On the other hand, distortions in imagery may be related to the facts of the past and can affect interpersonal relationships through altered judgments about past events (Garry et al., 1996; Sharman et al., 2004). In brief, the negative emotions and disturbances in the imagery system form a vicious circle thus supporting the fact of emotional imbalance.

Guided cognitive imagery interventions demonstrated the ability to break this cycle, and therefore are widely used in modern cognitive psychotherapy in the treatment of phobias, anxiety, depression, etc. (Wolpe, 1958; Anderson and Borkovec, 1980; Hackmann et al., 2011; Ji et al., 2016). Furthermore, the emotion-centered imagery training have demonstrated the ability to increase optimism (Blackwell et al., 2013), to improve interpersonal relationships (Meleady et al., 2013), as well as in the promotion of empathy and pro-social behavior (Gaesser, 2012) in healthy individuals. Although there is a vast knowledge about the beneficial impact of guided emotion-centered imagery on the emotional life, the effectiveness of self-guided training has yet to be explored. One of the obstacles for these types of assessments is to have only limited information (Huang et al., 2014) in relation to “how” emotional imagery affects the brain, as this information may allow an objective assessment of the impact of training. Therefore, here we explored the effect of self-guided emotion-centered imagery, trying to answer not only the question “whether”, according to the subjects, there was an effect on their state of emotional well-being, but also “how” functioning of the brain was affected by training. Objective effects on brain function were tested comparing the electroencephalograms (EEGs) recorded at rest with eyes closed, collected at baseline and after 12 weeks of practicing self-guided emotion-centered positive imagery training. The analysis of EEG obtained in this state can provide information about the functioning of the structures included in the default mode network (DMN) and the functional connectivity of the network, which are reported to be modified in the course of various mind-body practices (Brewer et al., 2011; Berkovich-Ohana et al., 2014; Garrison et al., 2015). The analysis of EEG resting state and the resting state networks was implemented through the use of standardized and exact low resolution brain electromagnetic tomography (sLORETA, eLORETA) (Pascual-Marqui, 2002, 2007a; Pascual-Marqui et al., 2011). LORETA neuroimaging software was used previously by other authors for the study of emotions (Saletu et al., 2010), in addition, this approach has been applied to investigate DMN in relation to practice different mind-body techniques (Lehmann et al., 2012; Berkovich-Ohana et al., 2014).

Here we hypothesized, that if efficiently, then self-guided positive emotion-focused imagery training will demonstrate a beneficial effect on the emotional state of the trainees called out by psychological testing (first hypothesis), and will be linked with EEG changes, explainable by the training (second hypothesis), namely: (a) from a topographical point of view, it could be expected changes in the structures/regions participating in the imagery or/and emotional processing; (b) anticipated increase in EEG connectivity [in accordance with previously reported data on the practice of visual imagery (Sviderskaya et al., 2006)]; (c) since the imagery training by itself is a creative work, some of the changes in EEG are supposed to occur in theta range, which is one of the most frequently reported as involved in the creative process (Petsche, 1996).

After 12 weeks of positive imagery training, the participants perceived themselves in a more positive light and had a feel for more meaning in their lives. In the end, they were more satisfied with their own lives, which confirmed our first hypothesis. The observed dynamic in the emotional state of the trainees, as pointed out in previous studies on the guided mental imagery, proved, that the methods of eliminating troublesome images and creating healthy alternatives (Hackmann and Holmes, 2004), as well as techniques to boost the positive image of the future (Blackwell et al., 2013) have a beneficial impact on the emotional state. Furthermore, the present study found that a positive emotional transformation may also be obtained, when the visualization is a self-guided exercise and thematically adapted to the current needs.

The analysis of EEG data confirmed the second hypothesis; it was validated that the regular practice of positive imagery cause changes in brain oscillatory activity and there are some arguments to indicate that the EEG changes can be seen in reference to the training. Firstly, the 3-D topographic analysis using sLORETA showed an increase in the current source density (CSD) in the right mPFC (BA10) after the training. Given that this region is involved in imagination of pleasant scenes (Costa et al., 2010), and also plays a role in self-reflection and self-awareness (Johnson et al., 2002), the observed changes in this area can be seen as being related to the training program. The fact of increased CSD in BA10 after training from other point of view is in line with the results of psychological tests, since this region is involved in the regulation of emotion (Liotti et al., 2002), as well as, contributes for the degree of satisfaction with life (Kong et al., 2015), parameters that were significantly changed after training. Our findings are consistent with data from meta-analytic reviews of neuroimaging studies in healthy participants, according to which mPFC is one of the structures, along with the anterior cingulate cortex, amigdala (Phan et al., 2002; Murphy et al., 2003) and the Insular cortex (Craig, 2009) most consistently associated with emotional processing.

Secondly, lagged coherence analysis demonstrated increased (as originally hypothesized) value of EEG coherence after a workout. One of the observations was increased inter-hemispheric coherence between the inferior temporal gyruses of both hemispheres. As these regions that are involved in the processing of images (Mellet et al., 1996), enhancing the coherence between them may indicate that the training caused an improvement of coordination of relevant networks located in temporal gyruses. Increased EEG coherence has been reported in connection with the visual imagery of simple graphical elements (topographic distribution and bands of observed changes have been different, depending on previous experience) (Sviderskaya et al., 2006), so it can be assumed that increased inter-hemispheric lagged coherence between inferior temporal gyruses is associated with the image forming process in general, but do not reflect the content of the images. Further, it can be supposed that the increase in the functional connectivity between the PCC and the right inferior temporal gyrus is associated with recurrent self-image processing in the course of the training, as among the various functions, PCC also plays a role in the self-referential process (Garrison et al., 2015). Moreover, self-referential process has been previously associated with a group of theta frequencies (Mu and Han, 2010), the same frequency range, where the observed changes described here.

Another hypothesis to consider in connection with the observed changes in the EEG, is the possible contribution of enhancement activity of GABA (gamma-aminobutyric acid) -ergić system, which is well-known for its anti-anxiety and antidepressant properties (Möhler, 2012). The first argument in favor of this hypothesis due to the fact that LORETA changes were found in beta range, which appears to be modulated by GABAergic system (Yoto et al., 2012). At the same time, it has been shown that GABA enhancement leads to increased interhemispheric EEG coherence (Fingelkurts et al., 2004; Sampaio et al., 2007). Thus, the possible contribution of GABA (gain at the end of training, compared to baseline) can be seen in agreement with the results of both analyzes. Indeed, it has been shown that practice of mind-body techniques, such as yoga, can lead to enhancement of GABA (Streeter et al., 2007).

One can argue about the need for the control group, performing only relaxation (in order to distinguish the effect of relaxation and imagery training) or other image script, for example, to play football (in order to distinguish the effect of the imagery scenario). In regard to this, we believe that the relaxation itself can't explain the observed results, not only because of the time constraints of relaxing exercises, but also because the results in this study are consistent with those observed during the imagery and contrast changes seen in other studies throughout the relaxation: EEG studies on relaxation techniques reported in the main decrease in coherence in the DMN (Lehmann et al., 2012; Berkovich-Ohana et al., 2014), which is opposite of the observed here and in other studies dedicated on imagery in which increased coherence was demonstrated (Sviderskaya et al., 2006). With regard to the option “control group, imaging different scenario,” we believe that such an option is incorrect, because it gives equal task to all of the participants. Consequently this makes the training of guided instead of self-guided, so unsuitable for further comparison because one and the same scenario can have different emotional content for the participants, which breaks down the main principle that the images must have a positive content for the person.

The knowledge gained from this study show that the self-guided imagery training (after adequate guided training) may be useful as a part of self-development programs to improve the emotional well-being in healthy subjects, and has the potential to be cost-effective method of intervention for subthreshold depression.

The project was reviewed and approved by Regional Ethics Committee South East Norway (D 2016/921).

SV analyzed the EEG data and wrote the article. HS recorded EEG, participated in the discussion and interpretation of data. BN analyzed the data from psychological tests and participated in the discussion and interpretation of data.

This study was partially supported by the Nord University, Norway.