Brain Imaging Analysis Can Identify Participants under Regular Mental Training

et al. (2012) Brain Imaging Analysis Can Identify Participants under Regular Mental Training. PLoS ONE 7(7): e39832. doi:10.1371/journal.pone.0039832 Brain Imaging Analysis Can Identify Participants under Regular Mental Training Joa o R. Sato 0 Elisa H. Kozasa 0 Tamara A. Russell 0 Joa o Radvany 0 Luiz E. A. M. Mello 0 Shirley S. Lacerda 0 Edson Amaro Jr 0 Dante R. Chialvo, National Research & Technology Council, Argentina 0 1 UFABC -Univ. Federal do ABC, Santo Andre , Brazil, 2 Instituto do Ce rebro, Hospital Israelita Albert Einstein, Sa o Paulo, Brazil, 3 Department of Psychobiology - UNIFESP - Univ. Federal De Sa o Paulo , Sa o Paulo , Brazil , 4 King's College London, Institute of Psychiatry, London, United Kingdom, 5 Department of Physiology, Univ. Federal de Sa o Paulo , Sa o Paulo , Brazil Multivariate pattern recognition approaches have become a prominent tool in neuroimaging data analysis. These methods enable the classification of groups of participants (e.g. controls and patients) on the basis of subtly different patterns across the whole brain. This study demonstrates that these methods can be used, in combination with automated morphometric analysis of structural MRI, to determine with great accuracy whether a single subject has been engaged in regular mental training or not. The proposed approach allowed us to identify with 94.87% accuracy (p,0.001) if a given participant is a regular meditator (from a sample of 19 regular meditators and 20 non-meditators). Neuroimaging has been a relevant tool for diagnosing neurological and psychiatric impairments. This study may suggest a novel step forward: the emergence of a new field in brain imaging applications, in which participants could be identified based on their mental experience. - Funding: Fundacao de Amparo a Pesquisa do Estado de Sao Paulo - FAPESP 2010/01394-4 Brazil and Instituto Israelita de Ensino e Pesquisa Albert Einstein. The funders had no role in study design, analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Pioneers in neuroscience studied patients with lesions and associated behavioural abnormalities, such as the classic case of Phineas Gage [1], in order to determine aspects of brain function. The advent of neuroimaging provided sufficient detail to enable the detection of brain damage in vivo, by the naked eye, and created the basis for neuroradiology [2]. Modern advances in neuroimaging, along with the use of computers, have resulted in more precise automated quantitative analysis. However, subtle differences in images were still difficult to identify accurately, until the application of Machine Learning methods for classification of brain images, such as Support Vector Machine (SVM [3]). These computational methods of pattern recognition have been used to aid discrimination of clinical brain pathologies associated with easily identifiable behavioural disorders [4,5]. Indeed, most studies focus on identifying participants with psychiatric or neurological conditions. However, less is known about the ability of these methods to classify the mental habits of a non-clinical population based only on information extracted from the brain. For example, suppose clinicians observe a group of subjects on a street market. It may not be too difficult to diagnose a person with autism. However, in the same scene it will be difficult to guess whether a person practices some form of mental training such as meditation. Previous research has revealed that meditation can be associated with changes in brain function and morphology. For example, Lutz et al. [6] demonstrated that long-term Buddhist meditation practitioners were able to self-induce sustained electroencephalographic high-amplitude gamma-band oscillations and phase-synchrony during meditation. This was particularly apparent at lateral frontoparietal electrodes. Kozasa et al. [7] compared the neural activity of non-meditators and meditators during a task which assessed attention (the Stroop Word-Color Task). Non-meditators showed greater activity than meditators in the right medial frontal, middle temporal, precentral and postcentral gyri and the lentiform nucleus. There were no regions with greater activity in meditators relative to non-meditators. Therefore, non-meditators required greater neural activation compared to regular meditators to achieve equivalent behavioural performance. This supports the hypothesis that meditation training results in greater efficiency via improved sustained attention and impulse control. In addition, there is evidence that long-term meditation practice is associated with increased cortical thickness. Lazar et al. [8] reported that prefrontal cortex and right anterior insula were thicker in meditators compared to matched controls. These areas are thought to be involved in attention, interoception and sensory processing. Alternatively, Ho lzel et al. [9] compared Vipassana meditators with non-meditators and found greater grey matter concentration in the right anterior insula, left inferior temporal gyrus and right hippocampus. The current study looks to build on this previous research by asking: is it possible to determine whether a person regularly meditates using only their structural brain image? We set out to explore this question by classifying participants by their expertise in meditation and then attempting to identify subtle differences between participants engaged in regular meditation and those who do not meditate. A pattern recognition approach based on SVM and feature selection was applied as a tool for automated classification. Materials and Methods This project was approved by the Ethics Committee of the Instituto Israelita de Ensino e Pesquisa Albert Einstein - Brazil (no. 07/762). Participants taking part in the study were given adequate information before participating and freely signed a consent form. Participants Participants were recruited from mailing lists and were split into regular meditators (19 subjects) or non-meditators (20 subjects) dependent on their responses. Regular meditators were considered to be those who practised meditation three times a week, and had been practising for at least three years. Non-meditators were those who reported practising less than once a week, or not at all. The groups were matched for age (meditators: 45.4769.47; non-meditators: 43.8069.35), gender (meditators: 8M/11F; nonmeditators: 9M/11F) and education level (meditators: 78% undergraduate degree, 22% post-graduate; non-meditators: 65% undergraduate degree, 25% post-graduate, 10% secondary school). There was no statistically significant difference between groups on any of these factors. On average, the meditator group had been regularly meditating for 8.564.1 years. The styles of meditation used in this group were: zazen (N = 4), mantra meditation (N = 2) mindfulness of breathing (N = 6), kr (...truncated)