EEG-based driving intuition and collision anticipation using joint temporal-frequency multi-layer dynamic brain network

Liang, Jialong; Wang, Zhe; Han, Jinghang; Zhang, Lihua · 2024 · Crossref

DOI: 10.3389/fnins.2024.1421010

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the neural mechanisms underlying driving intuition and collision anticipation, addressing the gap in understanding how the brain rapidly processes risk in high-stakes scenarios. While intuition is critical for traffic safety, existing research often relies on static EEG signal analysis, neglecting the dynamic, interconnected nature of brain activity. To address this, the authors introduce the Joint Temporal-Frequency Multi-layer Dynamic Brain Network (JTF-MDBN), a novel method that captures both temporal evolution and frequency-specific connectivity. The research aims to characterize brain network dynamics during the initial and advanced phases of intuition training and to identify biomarkers capable of predicting imminent vehicle collisions. The study utilized the Simulated Car Crash Anticipation EEG Dataset, comprising recordings from 40 participants performing driving simulation tasks. Participants engaged in two conditions: a Non-Alert State (NAS) with no collisions and an Alert State (AS) where collisions occurred randomly. EEG data were preprocessed using Independent Component Analysis and filtered into theta, alpha, and beta bands. Functional connectivity was quantified using the Phase Lag Index (PLI) within sliding time windows to construct multi-layer networks. The analysis compared the Initial Phase of Intuition Training (ITIP) with the Advanced Phase (ITAP) using multi-layer metrics such as modularity and participation coefficient, as well as single-layer graph theory metrics like node strength and local efficiency. Statistical significance was assessed using ANOVA with False Discovery Rate correction. Results indicated that brain network activity became significantly more stable and robust in the advanced phase of training compared to the initial phase. Specifically, the JTF-MDBN demonstrated stronger connection strength during alert state tasks. Multi-layer analysis revealed that modularity was significantly higher in the non-alert state than in the alert state within the alpha and beta bands, suggesting distinct organizational patterns for risk anticipation. Crucially, in the W4 time window (one second before collision), specific network features differentiated imminent collision scenarios from non-collision events. Single-layer analysis further identified statistical differences in node strength and local efficiency in the alpha and beta bands during the W4 and W5 windows. These identified biomarkers were used to train a linear kernel Support Vector Machine classifier, which achieved a collision risk prediction accuracy of 87.5%. The findings demonstrate that dynamic brain network biomarkers can effectively distinguish between collision and non-collision scenarios, validating the feasibility of using EEG-based network analysis for driving safety applications. The study highlights that intuition training enhances the stability and connectivity of brain networks, particularly in the alpha and beta bands. These results provide a scientific basis for developing brain-computer interface-based intelligent driving assistance systems, offering a pathway to improve hazard perception and reduce traffic accidents through real-time neural monitoring.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• anticipation

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data