A Scoping Review of Functional Near-Infrared Spectroscopy (fNIRS) Applications in Game-integrated Learning Systems

TL;DR

The paper addresses how functional Near-Infrared Spectroscopy ($fNIRS$) can illuminate brain processes in game-integrated learning systems (GLS). Using PRISMA-ScR, it identifies 21 empirical GLS-$fNIRS$ studies, classifies them into cognitive/affective and comparative designs, and analyzes platform, device, data types (e.g., $\Delta HbO$, $\Delta HbR$, $OXY$), outcomes, and designs. It finds GLS can match or exceed traditional methods in engagement and learning gains, while $fNIRS$ provides insights into cognitive states but real-time adaptation remains limited due to standardization and interpretability challenges. The review offers design implications, actionable HCI recommendations, and future directions toward brain-informed adaptive learning systems, emphasizing the potential and current constraints of integrating neural data into GLS. Overall, the work lays a foundation for brain-signal-informed adaptive educational environments and highlights pathways to translate neuroimaging findings into practical instructional design.

Abstract

Functional Near-Infrared Spectroscopy (fNIRS) has emerged as a valuable tool to investigate cognitive and emotional processes during learning. We focus specifically on game-integrated learning systems as the context for fNIRS-based brain data analysis. We selected game-integrated learning systems because such systems make learning more engaging, interactive, and immersive, all of which are critical features for adaptive learning design. The goal of this scoping review is to help researchers understand how fNIRS has been used so far to study brain activity in game-integrated learning systems. We also aim to show how brain data captured through fNIRS can support the development of adaptive learning systems by monitoring learners' cognitive states. Using the PRISMA-ScR framework, 1300 papers were screened, and 21 empirical studies were selected for in-depth analysis. Studies were categorized as affective/cognitive response studies or comparative studies, and further analyzed by learning platform, game device, fNIRS configuration, outcome measures, and study design. The findings reveal that game-integrated learning systems can be as effective as traditional methods in improving engagement and involvement. The findings also show that fNIRS offers valuable insights into cognitive states, but it has not yet been widely implemented in real-time adaptive systems. We identify key challenges in standardization and data interpretation and highlight the potential of fNIRS for developing brain-aware, interactive learning environments. This review offers insights to guide future research on using brain data to support adaptive learning and intelligent system design.

Figures (3)

• Figure 1: Scoping review process based on the PRISMA-ScR flowchart prisma_scoping_review.
• Figure 2: Number of publications per year grouped by game-integrated learning system and device type.
• Figure 3: Number of publications per year grouped by fNIRS study type and output measure.