Rationale and Design of the AutoHealth Study: Prospective clinical validation of continuous multimodal in-vehicle health monitoring in real-world mobility

Figures

• 

  Figure 1. Conceptual framework of Automotive Health and multimodal in-vehicle sensing. This figure illustrates the Automotive Health ecosystem, integrating optical, electrical, acoustic, and radar-based sensing modalities with vehicle telemetry. Biosignals are preprocessed into physiological features and behavioural indicators, which feed real-time analytics to derive actionable cardiovascular risk insights. These outputs can be interfaced with electronic health records (EHR), remote monitoring systems, and preventive care pathways enabling continuous physiological assessment and timely interventions during routine mobility. rPPG, remote photoplethysmography; ECG, electrocardiogram; PCG, phonocardiography; HR, heart rate; HRV, heart rate variability; SpO2, peripheral oxygen saturation.

• 

  Figure 2. Study design, cohort allocation, and data collection workflow of the AutoHealth trial. Participants are screened and assigned to one of five equally sized cohorts (healthy volunteers, increased cardiometabolic risk, heart failure with preserved ejection fraction [HFpEF], heart failure with reduced ejection fraction [HFrEF], or persistent atrial fibrilla1on). The study includes an extensive baseline clinical phenotyping visit, followed by a second visit with structured multimodal in-vehicle data collection. A follow-up telephone assessment is conducted four weeks later. PROMs, patient reported outcome measures; BIA, bioelectrical impedance analysis; BP, blood pressure; HR, heart rate; HRV, heart rate variability; RR, respiratory rate; SpO2, peripheral oxygen saturation; ECG, electrocardiogram; CPET, cardiopulmonary exercise testing.

• 

  Figure 3. Standardized in-vehicle protocol with cognitive, acoustic, and physical stressors.(A) The closed test course consists of a predefined sequence of straight segments, turns, slalom manoeuvres and a controlled braking zone. This layout enables reproducible driving dynamics, controlled exposure to acceleration and deceleration forces, and standardized elicitation of autonomic and haemodynamic responses. Telemetry-based segmentation of the course allows phase-specific analysis of sensor robustness, motion effects, and physiological reactivity. (B) The in-vehicle stress assessment consists of alternating rest phases and structured stress interventions. Rest phases are recorded with the vehicle stationary and the participant’s eyes closed. Participants then drive a standardized course during which cognitive load is applied using N-back and mental arithmetic tasks supplemented by acoustic and optical stressors. In the final round, a physical stressor is introduced through target braking at 30 km/h and g-force exposure. The protocol concludes with a final resting period to capture recovery dynamics. VAS, visual analogue scale.