Artificial intelligence-integrated wearable biomedical devices for cancer management

Abstract

Cancer remains the leading cause of death globally. Early diagnosis and intervention play deterministic roles in improving clinical prognosis. Traditional cancer management heavily depends on central hospital-based imaging and invasive diagnostics, which are intermittent and costly. Moreover, these strategies show limitations to patient compliance and real-time diagnosis. The emergence of wearable biomedical devices (WBDs) has offered a compelling alternative, enabling continuous, non-invasive <i>in situ</i> monitoring of bio-signals and real-time tissue imaging in daily settings. In particular, these devices have recently been integrated with therapeutic modules and artificial intelligence (AI) and have been adapted to closed-loop interventions, allowing for precise, on-demand drug delivery and localized therapy. In this review, we provide an overview of AI-integrated WBDs with their applications in cancer screening, diagnosis, and therapy. To solve the remaining issues of inaccurate screening, delayed intervention and severe side effects, the innovation of WBDs mainly includes conformable wearing structures, adhesive materials and integrated sensors/drug delivery modules. The integration of AI into WBDs has demonstrated high performance in improving signal-to-noise ratio (SNR) and real-time data processing, which significantly enhance the capabilities in long-term monitoring, and patient-specific bio-signal variations. The last session provides future directions for AI-integrated WBDs, focusing on improving SNR, reducing false positives caused by high sensitivity, and addressing patient data privacy concerns during AI training.

Ähnliche Arbeiten