Feasibility of fNIRS for assessing n-back task-related prefrontal cortex oxygenation after continuous cycling at varying intensities: an evaluation of protocol feasibility

Introduction While previous research has shown that single bouts of acute exercise can enhance cognitive performance, the physiological mechanisms behind this remain unclear. Functional near-infrared spectroscopy (fNIRS) constitutes a promising tool to capture brain activity in this research field, but the feasibility of implementing such setups in combination with stationary cycling has not yet been evaluated. Thus, in this exploratory feasibility and methodological validation study, we investigated (1) procedural feasibility of the experimental protocol and (2) scientific feasibility, including dual-metric data quality assessment via scalp coupling index (SCI) and quality testing of near-infrared scans (QT-NIRS), as well as behavioral and prefrontal hemodynamic indicators of n-back task engagement. Methods Of 14 eligible, nine young and healthy participants (2 female, 7 male) completed 20 min of low-, moderate- and high-intensity continuous-cycling in a randomized cross-over design. Prefrontal cortex (PFC) activation was captured during numerical n-back tasks, administered before and with a 10-min delay after exercise. Procedural feasibility was descriptively assessed via participant attrition, protocol timing adherence, fNIRS setup reliability, target exercise load estimation, and monitoring of affect and fatigue. Scientific feasibility was evaluated through average SCI values, agreement between oxygenated and deoxygenated hemoglobin (oxyHb, deoxyHb) β -coefficients derived from SCI- and QT-NIRS-based pruning (Lin’s concordance correlation coefficient, CCC), and n-back task sensitivity via accuracy, reaction time, and load-dependent prefrontal hemodynamic responses. Results Technical difficulties ( n = 1), physical discomfort due to cap wear ( n = 2), and undisclosed reasons ( n = 2) were reasons for study discontinuation. Protocol timing adherence and fNIRS setup reliability were satisfactory overall, and target exercise load estimation was successful after manual adaptation. Affect and fatigue were stable across timepoints. The channel-averaged SCI value of 0.964 ( SE = 0.003) and near-optimal β -agreement between techniques (oxyHb CCC = 0.958; deoxyHb CCC = 0.983) suggested that good quality data were retrieved from the setup. N-back task behavioral and hemodynamic outcomes were load-dependent. Discussion These findings provide preliminary support for the scientific feasibility of assessing n-back task-induced changes in PFC activation using fNIRS. Future studies should consider minor adjustments to pre-experimental design and fitness assessment to potentially reduce attrition and improve target exercise load estimation.