Autism in adolescence and adulthood.

The increasing global burden of chronic diseases, such as diabetes, cardiovascular disease, and cancer, together with the rapid growth of the aging population, has become a major public health challenge worldwide [–]. At the same time, substantial disparities in access to diagnostic, therapeutic, and post-treatment management persist across regions [, ]. Despite these growing demands, current healthcare systems remain largely based on centralized, one-size-fits-all models that are increasingly inadequate for delivering comprehensive, affordable, and consistently high-quality care. To address these limitations, the concept of precision diagnosis has emerged as a transformative framework in modern healthcare [–]. Precision diagnosis aims to identify disease onset, progression, and heterogeneity at the individual level by leveraging continuous, high-resolution, and multi-parameter data. However, existing clinical practices are predominantly episodic and rely on isolated measurements obtained in controlled settings, limiting their ability to capture dynamic physiological changes, interindividual variability, and early-stage disease signal [, –]. As a result, these approaches fall short of fully realizing the potential of precision diagnosis. In response, healthcare is undergoing a profound shift toward wearable health devices and digital diagnostics [–]. These technologies enable continuous and longitudinal monitoring of individual health status, facilitate early detection of clinically significant physiological alterations, and expand access to advanced diagnostic capabilities, particularly in resource-limited settings [, ]. Wearable diagnostic systems are integrated analytical platforms that consolidate biosensor functionalities with miniaturized electronic components, mobile computing interfaces, and machine learning (ML) algorithms for automatic data interpretation and outcome prediction []. By analyzing a wide range of biofluids, these systems enable molecular-level health assessment in formats that can be worn directly on the body or incorporated into garments and body-worn accessories, allowing continuous tracking of subtle temporal fluctuations in physiological status and supporting precision diagnosis [–]. In this context, precision diagnosis is not limited to the detection of individual biomarkers but refers to a clinically actionable framework enabled by continuous, highly sensitive, and selective monitoring of molecular targets in complex biological matrices [, , ]. Such capabilities provide deeper insight into underlying biochemical processes associated with health and disease and support personalized assessment of disease onset, progression, and prognosis. In addition, the integration of molecular and electrophysiological signals enables multi-parameter data fusion, thereby overcoming the limitations of single-biomarker thresholds and improving differential diagnostic accuracy [, , ]. More importantly, these systems establish personalized reference frameworks that distinguish normal interindividual variability from true pathological changes within the same individual. Together, these capabilities define a precision wearable diagnostic system as a platform integrating high-fidelity sensing, longitudinal monitoring, multidimensional data interpretation, and individualized clinical assessment. Wearable medical systems have emerged for several decades, as early as the 1960s, the Holter monitor pioneered the ambulatory recording of cardiac electrical signals []. In recent era, the convergence of Internet infrastructure, semiconductor technology, and big data has driven a fundamental transformation in wearable system architectures, catalyzing the development of fully integrated biochemical sensing platforms capable of extracting rich molecular information (Fig. ) []. Operating under diverse transduction mechanisms, these systems employ the following principal sensing modalities: direct electrical sensors, including potentiometric, resistive, capacitive, and inductive sensors; electrochemical sensors based on biochemical interactions or catalytic redox reactions; and optical sensor Each modality offers distinct advantages: potentiometric and resistive sensors are amenable to array-based signal acquisition [, –]; capacitive sensors offer rapid response kinetics [, ]; field-effect transistors (FETs) and their derivatives provide intrinsic signal amplification [, , ]; inductive sensors support non-contact measurement [, ]; electrochemical sensors combine analyte selectivity with straightforward miniaturization []; and optical sensors demonstrate superior electromagnetic interference immunity with trace-level analyte detection [–].