Gamma2Patterns: Deep Cognitive Attention Region Identification and Gamma–Alpha Pattern Analysis
Sobhana Jahan 1 , Saydul Akbar Murad 2 , Nick Rahimi 2 , Noorbakhsh Amiri Golilarz 1 · 2026
Read Gamma2Patterns: Deep Cognitive Attention Region Identification and Gamma–Alpha Pattern Analysis by Sobhana Jahan 1 , Saydul Akbar Murad 2 , Nick Rahimi 2 , Noorbakhsh Amiri Golilarz 1 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-06257.txt.
Introduction
Deep focus has a brain signature, and Sobhana Jahan and 3 co-authors map it in a 2026 paper. Sustained attention involves strong gamma brain waves plus coordinated eye behavior. Few computational studies combine these signals or pinpoint which regions drive deep cognitive attention.
Their Gamma2Patterns framework joins gamma and alpha band EEG recordings with eye-tracking data from the SEED-IV dataset. Features cover spectral power, burst timing, and fixation, saccade, and pupil signals across 62 electrodes. Frontopolar, temporal, anterior frontal, and parieto-occipital regions showed the strongest gamma activity in high-focus states. Eye signals pointed to overlapping frontal areas. Gamma power and burst duration marked deep focus better than alpha power alone, giving a foundation for attention decoding.
Catalog record
| CREATOR | Sobhana Jahan 1 , Saydul Akbar Murad 2 , Nick Rahimi 2 , Noorbakhsh Amiri Golilarz 1 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-06257
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Gamma2Patterns: Deep Cognitive Attention Region Identification and Gamma–Alpha Pattern Analysis Sobhana Jahan 1 , Saydul Akbar Murad 2 , Nick Rahimi 2 , Noorbakhsh Amiri Golilarz 1 Abstract Deep cognitive attention is characterized by heightened gamma oscillations and coordinated visual behavior. Despite the physiological importance of these mechanisms, computational studies rarely synthesize these modalities or identify the neural regions most responsible for sustained focus. To address this gap, this work introduces Gamma2Patterns, a multimodal framework that characterizes deep cognitive attention by leveraging complementary Gamma- and Alpha-band EEG activity alongside Eye-tracking measurements. Using the SEED-IV dataset [ 41 ] , we extract spectral power, burst-based temporal dynamics, and fixation–saccade–pupil signals across 62 channels or electrodes to analyze how neural activation differs between high-focus (Gamma-dominant) and low-focus (Alpha-dominant) states. Our findings reveal that frontopolar, temporal, anterior frontal, and parieto-occipital regions exhibit the strongest Gamma power and burst rates, indicating their dominant role in deep attentional engagement, while Eye-tracking signals confirm complementary contributions from frontal, frontopolar, and frontotemporal regions. Furthermore, we show that Gamma power and burst duration provide more discriminative markers of deep focus than Alpha power alone, demonstrating their value for attention decoding. Collectively, these results establish a multimodal, evidence-based map of cortical regions and oscillatory signatures underlying deep focus, providing a neurophysiological foundation for future brain-inspired attention mechanisms in AI systems. I Introduction Deep cognitive attention refers to a form of sustained attention where the relevant information to the task is being processed, and distractions are kept to a minimum [ 12 ] . It requires various neural activities and different oscillations that correspond to different attentional states, and it has been researched using EEG techniques, which are essential to studies related to neuroscience and human-computer interaction [ 25 ] . Deep attention models and theories are highly important for applications in neuroscience, human-computer interaction, education, cognitive health, and cognitive intelligence-inspired artificial intelligence systems. Deep attention can facilitate more complex cognitive tasks like learning and decision-making. Attention deficits could cause cognitive exhaustion and different neurological disorders [ 17 , 20 ] . Therefore, it is important to provide appropriate biological indices or brain regions of deep attention for more effective attention-tracking systems. Gamma oscillations (31-50 Hz) [ 8 ] are neural signatures of cognitive processing and attention, arising from the organized firing of neurons. There is a notable enhancement of Gamma oscillations in relevant brain regions during deep attention, which facilitates communication among diverse neural assemblies, thereby supporting working memory and perceptual binding. Increases in Gamma power and bursts features reflect cognitive demand and attention [ 3 ] . On the other hand, Alpha oscillations (8-14 Hz) [ 16 ] are associated with lower brain excitability and are linked to inhibitory control, dominating periods of relaxed and low-cognitive states. Gamma oscillations indicate that there is active processing of information, while Alpha oscillations indicate that they are functionally disengaged from other stimuli. Therefore, the Gamma and Alpha bands play distinct roles in the regulation of attention and cognitive processes. Eye-tracking features provide additional behavioral data that can enhance EEG analysis of deep cognitive attention. Fixation duration indicates ongoing processing and gaze duration, saccade movement provides insight into fixation point stability, and pupil dilation indicates the level of cognitive load. Together with EEG, especially Gamma-band data, these metrics can contribute to the verification of deep attention engagement. Gamma power and burst features, which correlate with extended fixation points, moderate saccadic movement, and higher pupil dilation; can thus serve as evidence of deep attention engagement [ 18 ] . To date, neural oscillations are considered an integral part of attentional regulation and information processing, especially in the Alpha and Gamma bands. Gamma activity is highly associated with high-level cognitive effort and sustained attention [ 4 ] . In contrast, Alpha activity reflects inhibitory control and reduced engagement. Eye-tracking features, such as fixation, saccades, and pupil dilation, also provide informative behavioral markers on attentional states [ 19 ] . Together, these neural and behavioral modalities offer a rich foundation for studying deep cognitive focus. However, despite their relevance, the relationship between EEG Alpha–Gamma patterns and Eye-movement behavior during deep focus has not been adequately investigated. While past studies have often emphasized spectral properties of EEG signals or the behavioral markers from Eye movements, they have predominantly done so in isolation, providing fragmented rather than holistic insights into attentional processes [ 30 ] . Consequently, how oscillatory activity, especially from higher frequency bands reflects sustained focus when synchronized with Eye-movement behavior has only begun to be explored [ 13 ] . Paralleling the shift toward multimodal diagnostics in healthcare, there is significant potential that the integration of neural oscillations with Eye-tracking measures may provide a more complete representation of attentional states [ 42 ] . Therefore, the neurophysiological basis of sustained focus, particularly regarding localized Gamma activation and burst characteristics, remains unclear. Thus, a critical need exists for a multimodal understanding of these brain regions and oscillatory signals. Recognizing the necessity of analyzing brain activity, researchers have increasingly employed diverse computational frameworks to classify cognitive states. Deep learning approaches, such as Convolutional Neural Networks (CNNs) [ 27 ] and Bidirectional Long Short-Term Memory (BiLSTMs) [ 35 ] , have been utilized to detect high-Gamma activity and temporal patterns with promising accuracy, yet these methods are often limited by an exclusive focus on single frequency bands or susceptibility to subject variability [ 2 , 36 ] . Shifting toward connectivity, other frameworks have investigated functional network features and Phase Locking Values (PLV) to assess emotional states and cognitive loads; while these studies successfully demonstrate that functional connectivity varies with workload, they frequently lack comprehensive temporal dynamics or face challenges regarding ecological validity [ 40 , 28 ] . Furthermore, general machine learning models have been developed to classify attention levels using extensive time-frequency features, but despite good generalization, these approaches are often hindered by imbalanced class distributions [ 24 ] . Despite promising progress, review of existing literature reveals that most prior studies have focused primarily on EEG-based classification [ 5 ] , such as emotion recognition or frequency band discrimination, rather than understanding the underlying neural mechanisms of attention. Many works also suffer from notable limitations, including imbalanced datasets, limited robustness, limited to laboratory conditions [ 15 ] , and the absence of multimodal assessment [ 38 ] . Moreover, studies that attempt to visualize regional activity often report subject-specific topography maps for only a few participants, without providing a generalized representation of the cortical regions involved [ 2 ] . These studies also tend to lack clear identification or interpretation of the brain areas most relevant to cognitive attention. As a result, there remains a meaningful gap in the literature: the field lacks a comprehensive, multimodal framework capable of determining which brain regions dominate during deep cognitive focus and how gamma-dominant patterns differ from Alpha-dominant activity. Addressing this gap provides the motivation for the present study. To overcome these limitations, we introduce Gamma2Patterns, the first multimodal framework that leverages EEG spectral power and burst characteristics alongside Eye-tracking measures to characterize attention-related neural patterns. The approach identifies influential regions of the brain, quantifies differences between Alpha and Gamma Bands, and extracts Gamma power and burst