Digital health innovation and artificial intelligence (AI)-enabled wearable medical devices are at a pivotal moment.
In early 2026, the US Food and Drug Administration (FDA) issued updated guidance documents aimed at easing regulation around two digital health product classes: wearable devices (such as smartwatches/rings and body-worn sensors) and clinical decision support software, an area of growing interest and application of AI models.1–4
The integration of AI and wearable bioelectronics is transforming digital health care by supporting proactive care, individualized treatment, and data-driven clinical decision-making.5
Shalini Prasad, PhD, Professor of Bioengineering and Biomedical Engineering at the University of Texas at Dallas, framed the expanding role of and opportunities for AI-assisted noninvasive technologies. “In health care broadly and in pulmonary medicine, prescreening, diagnostics, disease characterization, and treatment response assessments are all performed using separate sets of tools, associated with varying accuracies and sensitivities,” she said. Clinicians then interpret these diverse longitudinal datasets to inform care decisions.
Dr. Prasad said that coupling noninvasive smart wearable devices for collecting multimodal data with AI-enabled data analytics may help address two key challenges with this paradigm: health care access and improvement of patient-centric care and care efficiency.
AI techniques, particularly machine learning, play a key role in combining data from multiple sources to support accurate disease diagnosis.6 However, the path from noninvasive measurement of clinical parameters (eg, a blood-based COPD-associated inflammation biomarker or cough frequency/intensity) to a readily usable, clinically validated AI-assisted device traverses a complex health care terrain and diverse scientific disciplines.
Laying the groundwork for noninvasive technologies
Dr. Prasad, who cofounded a company that develops advanced biosensors for health care applications, said the key question with these noninvasive technologies is how to get multimodal data from different sensors and integrate them to generate reliable, accurate, clinically meaningful, and contextual findings that are actionable.
Dr. Prasad and colleagues are working on three classes of noninvasive sensors:
- Wearable sweat-based biomarker tracking platforms, enabling direct mapping of body biochemistry7–10
- Electrochemical sensors for monitoring markers in bodily fluids (eg, to monitor plasma levels of von Willebrand factor A2 domain as an early indicator of COPD flare)11
- Exhaled breath analyses platforms for profiling chemical compounds (eg, nitric oxide as a marker of respiratory distress)12–13
Wearables with photoplethysmography, ECG, and other biosensors with machine-learning software are improving the ability to detect cardiovascular problems, while also encouraging patients to stay engaged and adopt healthier habits.14
From innovation to improved patient care
More than 200 sensor-based digital health technology devices have been FDA-authorized for marketing in the United States since 2014.15 Diagnostic-support wearables have evolved from simple tracking tools into multisensor, connected systems with emerging technological and clinical trends shaping their next phase of adoption.16 However, no drugs based on sensor-based digital health technology end points have been approved yet.17
Dr. Prasad said the process for developing AI-enabled sensor platforms includes many steps. Meta-analyses can help identify quantifiable clinically relevant markers that can be translated to digital biomarkers. Then, sensor- and device-related technical challenges must be addressed.
Sensor-generated data can also be paired with conventional data, like biopsy or spirometry findings, and weighted appropriately for AI algorithm-based analysis.
Importantly, sensors and AI algorithms must be clinically tested and validated. Ambrose Chiang, MD, Senior Attending Physician at Cleveland VA Medical Center and UH Cleveland Medical Center, and Clinical Professor at Case Western Reserve University, cautioned that AI-enabled devices need to be assessed in well-designed clinical studies—rather than validation studies—that are adequately powered for robust statistical analyses, have sufficient sample sizes, and include diverse populations, preferably in multicenter trials.
In a recent review, Dr. Chiang and coauthors discuss common research pitfalls around novel OSA-detecting wearables for home sleep apnea testing. Dr. Chiang said independent, cross-device, head-to-head comparative evaluations to better understand the performance differences among AI-enabled OSA-detecting wearables are needed.18
Giovanni Ferrara, MD, PhD, Professor of Medicine and Director of the Division of Pulmonary Medicine at the University of Alberta, Canada, also underscored the complexity around AI-assisted wearables, adding, “We are still at the dawn of this new way to collect data directly from patients in a continuous manner. It is extremely exciting.”
“In pulmonary disease management, for instance, AI-enabled methods can help detect and analyze pulmonary sounds, like cough and wheezing,” Dr. Ferrara said, discussing a pilot trial he led on the feasibility and performance of an objective cough monitoring device. In the trial, algorithm-enabled cough detection using data collected via an on-body sensor worn by participants at home was referenced against validated questionnaires. Cough intensity, but not cough count, showed trends reflecting questionnaire scores.19
Dr. Ferrara said feasibility studies are important to show that such devices can be used by patients at home and over prolonged periods.
In another analysis using the same wearable, fine particulate matter smaller than 2.5 micrometers in diameter and ozone levels were associated with cough episode increases in patients with progressive pulmonary fibrosis.20
Embedding digital health in care, research
Despite the complexity and hurdles, sensor-based digital health technology is a rapidly growing field.
“Many institutions have established digital health offices, and pharmaceutical companies are also exploring these technologies in the context of chronic lung disease management,” Dr. Ferrara said.
“At the University of Alberta, we established a digital health unit within our clinical trials office to integrate these new technologies and data collection capabilities within randomized controlled and observational clinical trials.”
Wearable health technologies are advancing quickly, but their impact depends on solving persistent issues in data quality, interoperability, and equitable performance across diverse patient populations.21 Regulatory acceptance of clinical trial end points derived from digital health technologies requires early, structured engagement with regulators and a clear evidence package demonstrating the end point’s analytical validity, clinical relevance, reliability, and fit-for-purpose performance.22
Sleep medicine leading the way
Dr. Ferrara said that within pulmonology, sleep medicine is perhaps at the leading edge of sensor-based digital health technology, as many wearable home sleep apnea testing devices are available.
Dr. Chiang said such devices are very user-friendly, with automated scoring, minimal setup requirements, and a more streamlined user experience. Additionally, some can be used without pairing with a smartphone, and most have no wires, catheters, or chest belts. He said that AI-enabled wearable devices will be increasingly important in three main areas of sleep medicine:
- To diagnose OSA in uncomplicated cases with a high probability of moderate to severe OSA
- For follow-up of OSA treatment modalities, especially for patients using oral/dental appliances or nasal expiratory PAP devices (they can be easily worn for multiple nights to mitigate the effect of night-to-night variability)
- For OSA risk stratification, given FDA clearance of AI-enabled commercial sleep trackers for this purpose
Though several wearable home sleep apnea testing devices or software-as-a-medical-device products have been FDA-cleared for OSA diagnosis, Dr. Chiang said clinicians need to understand the strengths and limitations of these devices as well as the nuances of patient selection criteria when considering them.18 Clinicians also need to understand the physiology and mechanisms behind these devices, he said.
Beyond clinical validation and regulatory clearance
Many challenges can hamper the uptake of sensor-based digital health technology and software-as-a-medical-device technology in routine clinical practice.
Currently, Dr. Chiang said, clinical adoption of wearables among sleep specialists remains limited, in part due to the rapid pace of expansion of FDA-cleared devices and emerging reimbursement changes. Reimbursement for home sleep apnea testing is evolving, with changes anticipated in early 2027.23
“Another issue is the absence of consensus-backed clinical practice guidelines,” Dr. Chiang said.
The American Academy of Sleep Medicine (AASM) recently updated its position statements on AI, and an AASM Task Force is drafting guidelines on OSA diagnostic testing clinical practices, to be published this year, which will address the use of novel wearables for OSA testing at home.
Cross-disciplinary collaboration
Smart health care systems that combine “Internet of Medical Things” data with secure infrastructure, such as blockchain and deep learning, highlight why clinicians, engineers, data scientists, and cybersecurity experts must work closely together to translate digital health tools into reliable clinical care.24
Dr. Chiang said he urges clinicians to collaborate with AI experts and engineers, many of whom lead start-ups, for advancing clinical research. Drs. Ferrara and Prasad agreed.
“Digital health innovations require multidisciplinary collaboration and partnership,” Dr. Ferrara said. “At present, the hard lift is to validate these technologies and show that they are helpful for clinical patient management. Medical, operational, clinical trial, engineering, and computing science experts must be brought together, and we need big data experts.”
Dr. Prasad added that innovation happens in a collaborative ecosystem. When physicians who are technology-focused are willing to work with clinically oriented technologists/engineers, she said, the “bench to bedside” paradigm can become a reality.
References
1. Fiore K. FDA moves to ‘get out of the way’ of wearables, AI in the clinic. MedPage Today. January 9, 2026. www.medpagetoday.com/publichealthpolicy/healthpolicy/119352
2. Ardic N, Dinc R. Emerging trends in multi-modal artificial intelligence for clinical decision support: a narrative review. Health Informatics J. 2025;31(3):14604582251366141. doi:10.1177/14604582251366141
3. US Food and Drug Administration. General wellness: policy for low risk devices — guidance for industry and Food and Drug Administration staff. FDA. January 6, 2026. www.fda.gov/media/90652/download
4. US Food and Drug Administration. Clinical decision support software: guidance for industry and Food and Drug Administration staff. FDA. January 6, 2026. www.fda.gov/media/109618/download
5. Huang G, Chen X, Liao C. AI-driven wearable bioelectronics in digital healthcare. Biosensors. 2025;15(7):410. doi:10.3390/bios15070410
6. Kumar Y, Koul A, Singla R, Ijaz M. Artificial intelligence in disease diagnosis: a systematic literature review, synthesizing framework and future research agenda. J Ambient Intell Humaniz Comput. 2023;14(7):8459-8486. doi:10.1007/s12652-021-03612-z
7. Sardesai AU, Greyling CF, Lin KC, Kumar RM, Muthukumar S, Prasad S. A new paradigm in tracking the dynamics of glucose and cortisol: an observational study from human sweat enabled by a skin sensor. Biosens Bioelectron X. 2023;14:100377. doi:10.1016/j.biosx.2023.100377
8. Ramasubramanya A, Singh P, Lin KC, Prasad S, Muthukumar S. CIRCA: circadian inference of rhythmicity using comparative analysis from non-invasive continuous measurements of cortisol and melatonin in passive perspiration. Biosens Bioelectron X. 2025;26:100656. doi:10.1016/j.biosx.2025.100656
9. Singh P, Kumar A, Karmakar S, Lin KC, Muthukumar S, Prasad S. Monitoring TNF-⍺ monoclonal antibody via sweat cytokine profiling using the IBD AWARE®: a non-invasive approach to assess therapeutic response. Sens Biosensing Res. 2025;50:100914. doi:10.1016/j.sbsr.2025.100914
10. Shahub S, Kumar RM, Lin KC, et al. Continuous monitoring of CRP, IL-6, and calprotectin in inflammatory bowel disease using a perspiration-based wearable device. Inflamm Bowel Dis. 2025;31(3):647-654. doi:10.1093/ibd/izae054
11. David BE, Madhurantakam S, Karnam JB, Muthukumar S, Prasad S. Electrochemical profiling of vWFA2 for systemic inflammatory state detection. ACS Meas Sci Au. 2024;4(6):721-728. doi:10.1021/acsmeasuresciau.4c00060
12. Banga I, Paul A, France K, et al. E.Co.Tech-electrochemical handheld breathalyzer COVID sensing technology. Sci Rep. 2022;12(1):4370. doi:10.1038/s41598-022-08321-x
13. Subawickrama Mallika Widanaarachchige NR, Paul A, Banga IK, Bhide A, Muthukumar S, Prasad S. Advancements in breathomics: special focus on electrochemical sensing and AI for chronic disease diagnosis and monitoring. ACS Omega. 2025;10(5):4187-4196. doi:10.1021/acsomega.4c10008
14. Nazir A, Nazir A, Shah Wali Jamal M, et al. Wearable technology and its potential role in cardiovascular health monitoring and disease management. Health Sci Rep. 2025;8(11):e71486. doi:10.1002/hsr2.71486
15. US Food and Drug Administration. Medical devices that incorporate sensor-based digital health technology. FDA. December 5, 2025. https://www.fda.gov/medical-devices/digital-health-center-excellence/medical-devices-incorporate-sensor-based-digital-health-technology
16. Escobar-Linero E, Muñoz-Saavedra L, Luna-Perejón F, Sevillano JL, Domínguez-Morales M. Wearable health devices for diagnosis support: evolution and future tendencies. Sensors (Basel). 2023;23(3):1678. doi:10.3390/s23031678
17. Bakker JP, Izmailova ES, Clement A, et al. Regulatory pathways for qualification and acceptance of digital health technology‐derived clinical trial endpoints: considerations for sponsors. Clin Pharmacol Ther. 2024;117(1):56-72. doi:10.1002/cpt.3398
18. Chiang AA, Jerkins E, Holfinger S, et al. OSA diagnosis goes wearable: are the latest devices ready to shine? J Clin Sleep Med. 2024;20(11):1823-1838. doi:10.5664/jcsm.11290
19. Feist MD, Huang Y, Kalluri M, et al. Decoding objective cough features in progressive pulmonary fibrosis: a 6-month feasibility study. Am J Med. 2026;139(1):99-107.e9. doi:10.1016/j.amjmed.2025.07.027
20. Vander Linden H, Zanini U, Kalluri M, et al. The effect of pollutant exposure on cough in progressive pulmonary fibrosis. IJTLD Open. 2025;2(10):622-624. doi:10.5588/ijtldopen.25.0427
21. Canali S, Schiaffonati V, Aliverti A. Challenges and recommendations for wearable devices in digital health: data quality, interoperability, health equity, fairness. PLOS Digit Health. 2022;1(10):e0000104. doi:10.1371/journal.pdig.0000104
22. Bakker JP, Izmailova ES, Clement A, et al. Regulatory pathways for qualification and acceptance of digital health technology-derived clinical trial endpoints: considerations for sponsors. Clin Pharmacol Ther. 2025;117(1):56-72. doi:10.1002/cpt.3398
23. Roy S. Proposed home sleep testing CPT codes clear high hurdle. Sleep Review. April 30, 2025. https://sleepreviewmag.com/sleep-diagnostics/home-testing/home-apnea-testing/proposed-home-sleep-testing-cpt-codes/
24. Albakri A, Alqahtani YM. Internet of Medical Things with a blockchain-assisted smart healthcare system using metaheuristics with a deep learning model. Applied Sciences. 2023; 13(10):6108. doi:10.3390/app13106108
