Recalibrating Your Internal Metronome, With a Little Imagination

Tick…tock…tick…tock… We are all familiar with the unceasing sound of a clock ticking, or the regular beat of a metronome. This simple beat has big implications, though: repetitive auditory patterns form the basis of therapies such as rhythmic auditory stimulation (RAS) that are helping patients with motor impairments such as those induced by Parkinson's disease (PD) to move again [1]. There is a network of brain regions involved in the timing of our movements - an “internal metronome” - which enables us to complete rhythmic actions such as walking. In PD patients, however, parts of the network are not functioning correctly and movement is disrupted as a result. With RAS treatment, a regular beat is played for PD patients to help recalibrate their disrupted internal metronome, as they practice synchronized walking in step to the beat. But in order to further develop these therapies, we need to understand just how the brain is keeping in beat with the music - to uncover the mechanisms behind interactions between the auditory and motor brain regions, and what this means for internal timing.

Neurons in the brain, when activated in groups, produce electromagnetic fields strong enough to be detected and measured. Takako Fujioka and colleagues from the Rotman Research Institute in Ontario, Canada have suggested that the changes in electromagnetic brain activity within the “beta band” frequency (15 – 30 Hz) over time could provide insight into coordination between the auditory and movement regions during timed events. Using a technique called magnetoencephalography, they measured changes in brain activity within the beta band range as participants heard a regular sounded pulse, much like a metronome, at a number of different speeds [2]. The researchers used signal processing analysis to approximate what was going on in the auditory cortex, and found the beta band activity decreases about 200 milliseconds (ms) after a tone is heard. This decrease in activity is known as Event Related Desynchronization (ERD). The beta band activity then rebounds back to the original level in anticipation for the onset of the next beat, such that for faster tempos, the rebound is faster than for slower tempos. A whole-brain analysis suggested that this pattern of beta-ERD and rebound could provide functional coordination between auditory and motor systems.

But the thing is – in real life, music is not just a metronome; there is a hierarchy, a structure of downbeats and upbeats that gives a song a groove. For example, march- and waltz-style music have different patterns. Here, the downbeats are capitalized (ONE):

- March: ONE two ONE two… the accent pattern of a military-style song, like “Seven Nation Army” by The White Stripes;
- Waltz: ONE two three ONE …the pattern of “A Thousand Years” by Christina Perri.

Rhythmic auditory therapies to date use either simple marching music or a metronome. No study has directly compared effectiveness of treatment of music compared with a metronome within in a PD sample [1] and little research has been done on the difference between up- and downbeats, particularly in the context of differing hierarchical structures. So in their most recent study at Stanford University, Fujioka and colleagues were interested in whether participants listening to a rhythmic pattern of downbeats and upbeats (in these studies, downbeats were accented with a louder tone) would show the same beta-ERD when hearing two different patterns (march and waltz) [3]. Furthermore, would beta-ERD change when participants only imagined the downbeat accents on top of hearing a continuous beep track? This is an important consideration, as the auditory imagery literature suggests that imagining a sound involves many of the same auditory and motor regions [4] in which beta band activity is observed when hearing the beats.

Figure 1: The green sections of Sam and Morgan's beta-ERD graphs look similar. Thus the authors saw that imagining an accent on the downbeat showed similar changes in beta band activity as physically hearing the accent. All participants of the study completed both conditions. Figure design by Malika Kumar. Beta-ERD graphs are reproduced from Fujioka et al. [2].

Participants listened to a steady pulse every 390 ms, with an accented beat playing in two different patterns: either march or waltz (“Sam” in Figure 1; "Hearing a REAL Downbeat" video below). In a second condition, participants also listened to a steady 390 ms pulse, but had to imagine the accent on the downbeat of these different meters as instructed (“Morgan” in Figure 1; "IMAGINING the Downbeat" video below).

Using this paradigm Fujioka et al. measured the beta-ERD from the auditory regions of the brain and found that the amount of beta-ERD at 200 ms after the beat was larger for a downbeat than the following beat in both march and waltz patterns, regardless of whether the accent on the downbeat was physically present or simply imagined. That is, even when the participant was hearing a simple click track and only imagining the accent on the downbeat (ONE), the beta-ERD after 200 ms was greater than ERD after an upbeat (two).

Estimating the brain activity from only the auditory cortices enabled the researchers to measure beta-ERD at those locations, but didn’t give any information about other brain regions that contribute to changes in beta-ERD. To investigate, they used a whole-brain analysis of changes in the beta band frequency and identified a greater number of brain areas involved in imagining the downbeat compared to how many were involved when actually hearing it. This is to be expected given the more effortful task of imagining rather than simply hearing the accents. In addition, compared to the simple March pattern, the more complex Waltz rhythm was associated with changes in the beta band in a greater number of brain auditory and motor regions.

These findings are remarkable because the event-related brain response (calculated by averaging together all of the signals, not just beta band frequencies, from the whole brain on the accented beat) is much larger when hearing an accented beat than imagining an accent. However, when we tune into the beta frequency only, the modulations of beta after downbeats and upbeats in the March and Waltz patterns is actually similar across both perception and imagery – with greater beta-ERD after the downbeat than the upbeat. This novel finding supports the idea that beta band activity plays a role in setting our internal metronome, helping us to predict and anticipate timed events. Regardless of whether we hear or just imagine an accent, our brains are following along in the background, keeping track of the downbeat.

Fujioka et al. conclude that the observed beta band activities reflect the translation of timing information to auditory-motor coordination. Since the internal metronome in many motor-impaired patients is aided by use of an external metronome in therapies such as RAS, beta band activity may be useful as a measure of rehabilitation progress. So next time you hear a clock, take time to appreciate your own beta band internal metronome that is anticipating that next: tick…tock…tick…tock.

- Rebecca Gelding
Guest Contributor

Rebecca Gelding (@RebeccaGelding) is a part time PhD student within the Department of Cognitive Science and the ARC Centre of Excellence in Cognition and Its Disorders at Macquarie University, Sydney. She researches what happens in the brain as people imagine music. You can read more about her and her work on her blog.

[1] Ashoori A, Eagleman DM, Jankovic J. Effects of auditory rhythm and music on gait disturbances in Parkinson’s disease. Frontiers in Neurology. 2015;6. doi: 10.3389/fneur.2015.00234.
[2] Fujioka T, Trainor LJ, Large EW, Ross B. Internalized Timing of Isochronous Sounds Is Represented in Neuromagnetic Beta Oscillations. J Neurosci. 2012;32(5):1791-802. doi: 10.1523/jneurosci.4107-11.2012.
[3] Fujioka T, Ross B, Trainor LJ. Beta-Band Oscillations Represent Auditory Beat and Its Metrical Hierarchy in Perception and Imagery. J Neurosci. 2015;35(45):15187-98. doi: 10.1523/jneurosci.2397-15.2015.
[4] Hubbard TL. Auditory imagery: empirical findings. Psychol Bull. 2010;136(2):302-29. PubMed PMID: 00006823-201003000-00014.