PhysioKit: An Open-Source, Low-Cost Physiological Computing Toolkit for Single- and Multi-User Studies

Figure 1. PhysioKit is a novel physiological computing toolkit which is open-source and cost-effective. Its sensing and signal acquisition layer provides flexibility in using low-cost physiological sensors, and the software application layer enables data acquisition, visualization, and real-time ML-based signal quality assessment, while supporting passive as well as interventional studies for single and multi-user settings.

Figure 2. Flexible PPG setups available in PhysioKit to enhance accessibility.

Figure 3. (A) User-interface for Physiological Computing Toolkit; (B) configuration file for specifying experiment protocol; and (C) configuration file for controlling data acquisition.

Figure 4. SQA-Phys: 1D-CNN-based encoder–decoder architecture for high temporal precision signal quality assessment.

Figure 5. Multi-user setup of PhysioKit for synchronized data acquisition and stimulus presentation.

Figure 6. Data collection protocol for technical evaluation of PhysioKit. Four conditions that comprise the protocol are (A) baseline, (B) easy math task, (C) difficult math task, and (D) guided head movement task.

Figure 7. Bland−Altman scatter plots to compare heart−rate (beats per minute) (A,C), and pulse−rate variability (pNN50) (B,D) with the reference device.

Figure 8. (A): Comparison of pSQI for PPG signals of PhysioKit and the reference device, under different experimental conditions; (B): experimental condition-wise comparison of facial movement.