Devender Kumar

Rania Islambouli, Laura Geiger, Daniela Wurhofer et al.

Monitoring exercise intensity is critical for safe and effective physical activity, particularly for individuals with cardiovascular disease, where overexertion can pose serious risks. Although physiological measures such as heart rate are widely used for avoiding overexertion, they can be unreliable in certain cases, such as when affected by medication or when wearables are worn too loosely. We introduce AktivTalk, a mobile prototype that digitizes the clinically validated Talk Test to support voice-based, in-the-moment self-assessment of exertion. In a within-subject study with 20 participants, we collected exertion-labeled voice samples and found that AktivTalk was rated as highly usable and preferred over conductor-guided assessment. We further explored automated exertion classification from Talk Test speech. Using MFCC-based features with class balancing and cross-validation, a lightweight neural classifier achieved up to 90% accuracy for detecting high vs.non-high exertion from Talk Test recordings. This work highlights the potential of structured voice interactions for accessible exertion assessment and motivates future passive exertion monitoring from speech.

David Haag, Devender Kumar, Sebastian Gruber et al.

We evaluated the viability of using Large Language Models (LLMs) to trigger and personalize content in Just-in-Time Adaptive Interventions (JITAIs) in digital health. As an interaction pattern representative of context-aware computing, JITAIs are being explored for their potential to support sustainable behavior change, adapting interventions to an individual's current context and needs. Challenging traditional JITAI implementation models, which face severe scalability and flexibility limitations, we tested GPT-4 for suggesting JITAIs in the use case of heart-healthy activity in cardiac rehabilitation. Using three personas representing patients affected by CVD with varying severeness and five context sets per persona, we generated 450 JITAI decisions and messages. These were systematically evaluated against those created by 10 laypersons (LayPs) and 10 healthcare professionals (HCPs). GPT-4-generated JITAIs surpassed human-generated intervention suggestions, outperforming both LayPs and HCPs across all metrics (i.e., appropriateness, engagement, effectiveness, and professionalism). These results highlight the potential of LLMs to enhance JITAI implementations in personalized health interventions, demonstrating how generative AI could revolutionize context-aware computing.