Magnetoencephalography Recordings Reveal The Spatiotemporal Dynamics Of Recognition Memory For Complex Versus Simple Auditory Sequences

Auditory recognition is a crucial cognitive process that relies on the organization of single elements over time. However, little is known about the spatiotemporal dynamics underlying the conscious recognition of auditory sequences varying in complexity. To study this, we asked 71 participants to learn and recognize simple tonal musical sequences and matched complex atonal sequences while their brain activity was recorded using magnetoencephalography (MEG). 

Music has the unique and incredible power to turn us back in time and relive the wonderful memories of the past, or to immerse us deeply in the beauty of the present. More importantly, music can also improve our memory and enhance our learning and working abilities.

Research has found that musical sequences are closely linked to memory. Sometimes we find that we can easily remember the lyrics to a song because the song stimulates the language center in our brain and consolidates the memory through repeated playing. Likewise, when we are learning new information, we can remember it more easily by combining the information with melody and rhythm. Many medical experts and educationists recommend using music as a memory technique to enhance learning.

In addition to this, music can improve our emotional state and make us more relaxed and focused. When we feel tired or anxious, relaxing music can immediately relieve the situation. This kind of music is very effective in relaxing tense emotions and can also enhance our willpower and patience.

All in all, music is an integral part of our lives and can be very beneficial to our memory and learning abilities. So, whether you are working or studying, you might as well use some music to assist yourself. I believe this can help you complete your tasks better and also make your life better. It can be seen that we need to improve our immunity. Cistanche deserticola can significantly improve immunity because Cistanche deserticola is a traditional Chinese medicinal material with many unique effects, one of which is to improve memory. The efficacy of minced meat comes from the various active ingredients it contains, including acid, polysaccharides, flavonoids, etc. These ingredients can promote brain health in various ways.

Click Know Short-term Memory how to improve

Results reveal qualitative changes in neural activity dependent on stimulus complexity: recognition of tonal sequences engages hippocampal and cingulate areas, whereas recognition of atonal sequences mainly activates the auditory processing network. Our findings reveal the involvement of a cortico-subcortical brain network for auditory recognition and support the idea that stimulus complexity qualitatively alters the neural pathways of recognition memory.

Encoding and recognizing structurally complex sounds is a cognitive challenge relying on neural mechanisms that are not yet fully elucidated. To memorize complex sound sequences, we likely depend on the temporal organization of a stimulus’ components and memory functions.

Memory encoding takes place in the hippocampus2–4, whereas subsequent processes related to recognition memory are supported by a functional network of interconnected regions in the medial temporal lobe, including the hippocampus, insula, and inferior temporal cortex2,5,6. For memory consolidation, communication between hippocampal and neocortical areas is needed7–9. Much evidence comes from studies using static visual stimuli, such as pictures of objects, faces, or natural scenes10–12. However, information and meaning also unfold over time as the brain attempts to predict upcoming stimuli based on prior memory representations. Hence, to better understand memory recognition and its underlying fast brain dynamics, novel methods must be adopted that highlight the temporal properties of dynamic stimuli. This can be done by studying the neural activity underlying the processing of sound sequences that acquire meaning through their evolution over time, such as music13–15.

Few neuroscientific studies have explored the neural underpinnings of musical memory, as reviewed by Campo and Brattico16. For example, using functional resonance imaging (fMRI) and a naturalistic music listening paradigm, Alluri et al.17 investigated the neural correlates of music processing and reported activation of cognitive, motor, and limbic brain networks for the continuous processing of timbral, tonal, and rhythmic features. Subsequently, using the same stimuli, Burunat et al.18 reported the recruitment of memory-related and motor brain regions during the recognition of musical motifs. Despite their contributions, these studies fail to identify the fine-grained temporal mechanisms of sound encoding and memory processes.

More recently, we combined the high temporal resolution of magnetoencephalography (MEG) with the high spatial resolution of magnetic resonance imaging (MRI) to study music recognition. These studies accentuated the temporal involvement of a widespread cortico-subcortical brain network comprising the primary auditory cortex, superior temporal gyrus, frontal operculum, cingulate gyrus, orbitofrontal cortex, and hippocampus during recognition of auditory (musical) sequences19–21. Overall, these investigations have provided unique insight into the neural mechanisms underlying the recognition of temporal sequences. What remains to be addressed is how these mechanisms are modulated by stimulus complexity.

Here, we used melodic sequences, where meaning emerged from the sequential combination of individual tones over time21, and varied the tone distribution to obtain new, complex musical sequences. In this scenario, encoding and recognition of the musical sequences largely depend on the sequential order of the tones that comprise it. We first selected musical sequences based on the rules of tonality, which is the dominant musical system in Western popular music22. Second, by modifying the tone intervals (i.e., the distances between pitches) while keeping all other variables (e.g., rhythm, tempo, timbre) constant, we generated matched atonal musical sequences. The stimulus manipulation was based on previous literature, which reported that tonal rather than atonal musical sequences are overall easier to process23–27 and more appreciated by non-expert listeners27–29. Unlike tonal music, atonal music is characterized by the absence of a clear tonal center and hierarchical stability, which significantly reduces its predictive value and gives rise to increased prediction errors23–26,30. Thus, we expected that altering tonal intervals would reduce the predictability of the atonal sequences, leading to increased difficulty in recognizing them.

To summarize, in the current study we used MEG and a musical recognition task19–21 while participants listened to and recognized auditory (musical) sequences of varying complexity and aimed at describing the fine-grained spatiotemporal dynamics of auditory recognition memory. Following the previous studies17–21, we expected that the recognition of auditory sequences would activate a widespread brain network that includes both auditory (e.g., primary auditory cortex, superior temporal gyrus, Heschl’s gyrus, planum temporale, insula) and memory processing areas (e.g., hippocampus, cingulate gyrus). 

We further hypothesized that neural activity would be distributed along two main frequency bands that reflected the occurrence of two different cognitive processes: a slow frequency band related to the recognition of the full musical sequence in memory processing areas, and a faster frequency band associated with the processing of each tone of the musical sequence in auditory regions. More importantly, we hypothesized that, based on stimulus complexity, tonal music would be more efficiently processed than atonal music, which would be reflected in different behavioral responses and distinct neural pathways during the recognition of tonal and atonal sequences.

Our behavioral and neural results showed clear differences between the recognition of tonal and atonal sequences. Source reconstruction analyses indicated different activation clusters for tonal and atonal sequences. Overall, the neural activity was stronger in memory processing areas for memorized tonal sequences and in auditory processing regions for memorized atonal sequences.

Results

Behavioral data. Participants performed an old/new auditory recognition task (Fig. 1). They first listened to a full musical piece (encoding) and subsequently identified which musical sequences were memorized or novel. During recognition, the participants' response accuracy and reaction time were recorded using a joystick. These behavioral data were statistically analyzed to examine the differences between the four experimental conditions (memorized tonal sequences, novel tonal sequences, memorized atonal sequences, and novel atonal sequences).

A one-way analysis of variance (ANOVA) showed that the differences in response accuracy were statistically significant, F(3, 280) = 6.87, p = 0.002. Post-hoc analyses indicated that the average number of correct responses was significantly lower for memorized atonal sequences (M = 30.98, SD = 5.46) than for novel atonal (M = 34.51, SD = 4.26, p < 0.001), memorized tonal (M = 34.34, SD = 5.95, p = 0.002) and novel tonal sequences (M = 34.41, SD = 6.04, p = 0.001). Figure 2a shows the differences between conditions using raincloud plots31.

Regarding the mean reaction time, there was a statistically significant difference between conditions as determined by one-way ANOVA, F(3, 280) = 4.94, p = 0.002. Post-hoc analyses revealed that the average reaction time was significantly lower for memorized tonal sequences (M = 1735.17, SD = 259.91) compared to memorized atonal (M = 1879.44, SD = 259.34, p = 0.005) and novel atonal sequences (M = 1873.78, SD = 250.48, p = 0.007), but not compared to novel tonal sequences (M = 1799.52, SD = 267.14, p = .450). Figure 2b displays the differences between conditions.

Numerical source data for behavioral responses is provided in Supplementary Data 1.

MEG sensor data. The MEG data (204 planar gradiometers and 102 magnetometers) were analyzed at the MEG sensor level, using the broadband signal. Although the emphasis of the study lies in identifying the brain areas involved in recognizing tonal versus atonal musical sequences, the MEG sensor data were examined to assess whether the neural signal was significantly different for memorized than for novel trials and thus would corroborate the results of previous studies19,20.

After averaging the epoched data of correct trials for each experimental condition and combining the planar gradiometers, paired-sample t-tests were performed to identify which condition (memorized or novel) generated a stronger neural signal for each time sample and MEG sensor. Cluster-based MCS were then calculated to correct for multiple comparisons. This was performed independently for both tonal and atonal data (see Methods for details).

First, paired-samples t-tests (α = 0.01) were calculated for the tonal data in the time interval 0–2500 ms (from the onset of the Regarding the atonal data, paired-samples t-tests (α = 0.01) were calculated in the same time interval (0–2500 ms) using combined planar gradiometers. Next, multiple comparisons were corrected for using MCS on the significant t-tests’ results (α = 0.001, 1000 permutations). This procedure identified three main significant clusters of activation when contrasting memorized versus novel excerpts (Table 2, Supplementary Fig. 3, and Supplementary Data 2). In the case of the novel versus memorized contrast, three main significant clusters of activity were found (Table 2, Supplementary Fig. 4, and Supplementary Data 2).

Source reconstruction. After examining the strength of the neural signals at the MEG sensor level, we focused on the main aim of the study, namely to investigate the neural differences underlying the recognition of tonal versus atonal musical sequences in MEG reconstructed source space. To perform this analysis, we localized the brain sources of the event-related fields recorded by the MEG channels. This was performed for both the tonal and atonal sequences and for two frequency bands (slow [0.1–1 Hz] and faster [2–8 Hz]) that were used in Bonetti et al.19,20 and linked to the processing of single components (slow) relative to a whole sequence (fast). Additional analyses were computed to contrast the brain activity between the slow and fast frequency bands and to examine the pattern of activity at the standard delta (1–4 Hz) and theta (5–8 Hz) frequency bands.

Slow frequency band (0.1–1 Hz). The neural sources were calculated using a beamformer approach. First, a forward model was computed by considering each brain source as an active dipole and calculating its strength across the MEG sensors. Second, a beamforming algorithm was used as an inverse model to estimate the brain sources of neural activity based on the MEG recordings.

After computing the neural sources, a GLM was calculated at each timepoint and dipole location. A series of t-tests (α = 0.05) was carried out at the first and group levels to estimate the main effect of memorized and novel conditions and their contrast for both the tonal and atonal data independently. Cluster-based MCS (α = 0.001, 1000 permutations) were computed to correct for multiple comparisons and to determine the brain activity underlying the development of the musical sequences. These analyses were carried out for five specific time intervals that corresponded to each of the tones comprising the sequences: first tone (0–250 ms), second tone (251–500 ms), third tone (501–750 ms), fourth tone (751–1000 ms), and fifth tone (1001–1250 ms). This was estimated for the memorized versus novel contrast for both tonal and atonal sequences independently and for memorized tonal versus memorized atonal sequences.

Significant clusters of activity (p < 0.001) were located across several brain voxels (k) for each tone of the tonal sequences, as reported in Supplementary Data 3. For memorized tonal sequences, the neural activity was overall stronger for the third (k = 69), fourth (k = 266), and fifth tones (k = 229). 

The largest differences were localized in the middle cingulate gyrus, right supplementary motor area, precuneus, and left lingual gyrus for the third tone; the left amygdala left parahippocampal gyrus, left lingual gyrus, left hippocampus, and middle cingulate gyrus for the fourth tone, and the anterior and middle cingulate gyrus and left lingual gyrus for the last tone. For novel tonal sequences, the brain activity was stronger for the first (k = 54) and second tones (k = 29). In particular, the difference between novel and memorized sequences was strongest in the left calcarine fissure, left lingual gyrus, left hippocampus, left precuneus, and left superior temporal gyrus for the first tone, and the right fusiform gyrus, right lingual gyrus, and right inferior occipital gyrus for the second tone. The contrast between memorized and novel tonal sequences for the slow band is depicted in Fig. 3a.

For more information:1950477648nn@gmail.com