This Neuroscientist Decoded the Brain Patterns of Meditators

Recent neuroscientific research results illuminate the way a meditator's internal brain processes can flow and fluctuate during meditation.

Feodora/Adobe Stock

When we practice meditation, what’s really happening inside our brains? Neuroscientist and clinician Dr. Helen Weng is not only finding answers to this question but also pioneering a novel approach to the neuroscience of mindfulness.

Instead of treating meditation as a uniform and unchanging process that occurs more or less the same way in everyone’s brain, Weng wanted to accurately portray how internal states ebb and flow during meditation.

Weng is an Assistant Professor of Psychiatry and Behavioral Sciences at the University of California, San Francisco, core faculty of the Osher Center for Integrative Medicine, and affiliate faculty member of the Neuroscape Center. In August 2020, she and her team published the first-ever peer-reviewed study to decode brain networks during meditation, using machine learning. 

Instead of treating meditation as a uniform and unchanging process that occurs more or less the same way in everyone’s brain, Weng wanted to accurately portray how internal states ebb and flow during meditation. Her recent paper published results from a machine-learning experiment on brain network dynamics during meditation. Weng and her team used a new framework they developed, the EMBODY framework (Evaluating Multivariate Maps of Body Awareness), which measures and shows the fluctuation of body awareness during the moment-to-moment unfolding of meditation.

Inside the Minds of Meditators  

Weng and her team were interested in revealing what our internal attention states are doing in every second of a meditation session. 

First, they “taught” a machine-learning program to recognize brain patterns related to different meditative states. Simply put, machine learning involves feeding large amounts of data (in this case, brain imaging data of active brain networks) into a program, repeatedly “teaching” the program, and then testing how accurate it is at recognizing different brain networks. This is similar to a common way machine learning is applied to recognize faces, where algorithms learn to recognize a face after being given many images to learn from and then can use that information to recognize that face in new pictures.

In this case, the researchers taught the machine-learning program how to recognize three internal attention brain networks, specifically during these three situations:

  1. Breath Attention (‘Pay attention to your breath’)
  2. Mind Wandering (‘You may stop paying attention to your breath now’)
  3. Self-Referential Processing (or thinking about one’s life with the instruction ‘Think about the past week’).

Weng’s team recorded a whopping 2,160 brain patterns per person, from eight experienced and eight novice meditators, to find the most accurate and individualized brain network model for each of these three internal states. These brain patterns were specific to each person’s unique brain structure and function, and were adapted for each participant (like adapting to recognize each person’s unique face or fingerprint).

The researchers were able to estimate new metrics of attention during meditation, such as the percentage of time focused on the breath or engaging in self-referential processing, that were not previously available.

Second, they used the machine-learning program (now “well-learned” in recognizing the three brain networks) to decode the brain activity of each person during a 10-minute meditation session. In other words, individuals’ brain data was recorded during a 10-minute meditation where they were invited to focus their attention on their breath. As meditators know, our attention often fluctuates on and off the breath during meditation. To determine what each participant was focused on moment-by-moment, the machine-learning program then determined which of the brain networks —“Breath Attention”, “Mind Wandering,” or “Self-Referential Processing” —was being activated in each second. 

Finally, the researchers counted up the different meditative states that occurred, second by second, and with this data they produced a “storyline” of what each individual’s brain was doing during those 10 minutes. This data could then be used to estimate the percentage of time each person was attending to their breath during meditation. Simply put, this is the first time scientists have been able to “read the mind” during meditation at this level of detail.

Each participant’s unique brain pattern for attending to either the breath, mind wandering (MW), or self-referential processing (thinking about one’s life), identified by machine learning algorithms. Figure 3 from Weng et al., 2020. 

Based on each meditators’ unique brain signals, the machine-learning algorithms decoded or identified the mental states occurring second-by-second during 10 minutes of meditation. This generated a “storyline” for each meditator, showing when their attention was focused on their breath, when their minds were wandering, and when they were thinking about their lives. Figure 4 from Weng et al., 2020.

Decoding the Results

So what does this study tell us? First, the researchers showed that the machine-learning program was able to recognize different modes of internal attention that occur during meditation (attending to the breath or not), even when participants’ eyes were closed during the experiment. Second, they found that this method of detecting brain networks (of internal attention states) was possible for a large majority of participants (87.5%), which means this approach to using brain signals to study meditation can be applied to both experienced and brand-new meditators. Third, the researchers were able to estimate new metrics of attention during meditation, such as the percentage of time focused on the breath or engaging in self-referential processing, that were not previously available. They also found using these new metrics that on average, participants were able to focus longer on the breath than other mental states during meditation. 

These novel approaches will help the field to better measure how attention changes through meditation practice, and future research may help us better tailor meditation practices to improve outcomes and treat symptoms. 

The Science of Meeting People Where They Are

As a meditator, a clinician, and a scientist, Dr. Helen Weng holds a powerful combination of skills that support the aim of mindfulness research—to gain a deeper scientific understanding of how and why people may benefit from practices and therapies rooted in mindfulness. Her guiding spirit in both clinical and scientific work is grounded in her own practice of compassion and mindfulness, as well as her own life experiences as a female neuroscientist of Taiwanese descent. 

Weng’s broad field of expertise is intertwined, too, with the development of her own meditation and clinical practices. “Learning compassion meditation, I found I was having trouble repeating the phrases, and would get confused or forget the words,” she said. “After four years of practice, I gave myself permission to let the words go and focus on the strongest response in my body—a warmth in my chest which is my embodied feeling of love and compassion.” 

And once she had given herself this permission, Weng said, she found it also became easier to adapt meditation instructions for her psychotherapy clients, to suit their diverse backgrounds, minds, and ways of understanding the world. “The more I tailored meditation exercises to their unique minds, the more quickly they could access their own healing process. After witnessing these inspiring transformations, I wanted to bring this flexibility and diversity into my neuroscience research,” explained Weng. (You can access meditation tools for cultivating compassion, developed by Dr. Weng, through The Center for Healthy Minds at the University of Wisconsin-Madison.)

The EMBODY study opens new avenues to study the fluid and dynamic internal states of the meditating brain, in a way that more accurately reflects what the experience of meditation really is like. As we grow in our meditation experience, we recognize that moments in meditation are ever-changing. The goal of mindfulness practice is not to attain a static, rigid state of attention, but to train the skill of adapting to, and meeting moment-to-moment, the dynamic experiences of the mind, heart, and body. This study, one piece of ongoing research to include diverse experiences and meditators in contemplative research, recognizes that no two individuals’ brain processes are identical, and no two moments in meditation are identical. With these new methods, researchers can measure and study a fuller range of diverse experiences during meditation.

read more