Ethical Prioritization Framework for the Responsible Integration of AI- and IoMT-Enabled Smart Medical Devices ^†

Abstract

The integration of smart medical devices (SMDs) driven by artificial intelligence (AI) and the internet of medical things (IoMT) is revolutionizing healthcare through improved diagnostics and continuous monitoring. However, their deployment raises significant ethical concerns, including patient safety, data privacy, informed consent, fairness, bias, and regulatory compliance. This paper presents a structured prioritization framework that assesses these ethical considerations according to their severity, contextual impact, and relevance to clinical practice. The usefulness of this prioritization lies in its ability to guide stakeholders to focus on high-impact areas, ensuring that resources and interventions address the most critical ethical risks first. Targeted mitigation strategies support the application of this framework in practice. By aligning innovation with ethical responsibility, this approach promotes safer, fairer, and more reliable healthcare solutions. This ultimately enables the sustainable and socially responsible integration of SMDs into modern medical systems.

1. Introduction

Today, smart medical devices (SMDs) play a crucial role in the progress of digitalization and automation of healthcare ecosystem, further contributing to improved diagnosis and alleviate the burden on medical professionals. More importantly, these smart devices support patient benefit and well-being by providing better access to health information, as well as through continuous monitoring of symptoms and facilitating real-time surveillance [1]. SMDs incorporate up-and-coming technologies, such as the internet of medical things (IoMT) and artificial intelligence (AI) [2]. The World Health Organization (WHO) emphasizes the pivotal role of SMDs in seamlessly integrating long-term healthcare into patients’ everyday routines [3]. Furthermore, the WHO underscores how technology-assisted solutions reduce healthcare costs by preventing hospital readmissions, decreasing emergency room visits, and minimizing unnecessary testing and treatments [4]. In recent years, the field of SMDs has witnessed remarkable progress, largely driven by the growing number of devices that are computerized and connected. Notably, the IoMT market was valued at over 44 USD billion in 2023 and is projected to grow at a compound annual growth rate (CAGR) of 21.2% over the next six years [5].

Various SMDs categorized by their applications, including wearable health monitors, remote patient monitoring, and health diagnosis systems, have been developed to gather data and process information for assessing health and tracking individualized therapy progress [6]. Additionally, social and physical assistance systems are being deployed to aid individuals in receiving timely help and support when needed. In the context of decision support systems, AI-driven tools have shown significant promise in enhancing clinical workflows and improving healthcare delivery [7]. For instance, the work in [8] presents a CNN-LSTM-based healthcare system integrating Apache Kafka, TensorFlow Serving, and federated learning for secure, real-time anomaly detection. It efficiently processes multimodal data, guaranteeing high accuracy, low latency, and strong privacy protection in distributed medical environments. Moreover, the exponential growth of real-time artificial intelligence aids a real-time AI system developed from EUS images effectively differentiates gastrointestinal stromal tumors from leiomyomas, with high accuracy and superior performance to endoscopists. Validated in multiple cohorts, it demonstrates a strong diagnostic performance, particularly for small lesions, improving clinical decision-making in gastrointestinal care [9]. In Ref. [10], this study presents an IoT-based knee-monitoring system using wearable IMUs and edge-cloud computing for telerehabilitation. Validated by hardware and exercise tests, it demonstrates accurate, efficient, and low-power performance, enabling remote and home rehabilitation with real-time feedback.

However, while their integration represents a transformative step forward, it also introduces significant risks, particularly concerning data accuracy and the potential for breaches of sensitive health information. Moreover, in the high-risk area of healthcare, mistakes can have severe consequences for the patient who is the victim of this error. As such, the deployment of SMDs demands thorough ethical considerations. Indeed, integrating AI and IoMT into healthcare in a responsible way depends on proactively addressing complex ethical challenges, including protecting patient privacy, ensuring robust data protection, clarifying liability frameworks, maintaining transparency in algorithmic decision-making, promoting equitable access and outcomes, and maintaining human agency in care decisions. Finding a compromise between innovation and ethical responsibility is essential, as these smart devices only have transformative potential if their deployment prioritizes patient safety, cultivates equity, and fosters trust between individuals, medical practitioners, and communities. Thoughtfully addressing these challenges will not only mitigate risks but also pave the way for sustainable advances that combine technological progress with societal values and human dignity.

This study examines the complex ethical challenges associated with the integration of SMDs into healthcare through a discussion of their dual roles as transformative tools and sources of risk. It highlights the critical need for dynamic ethical guidelines to govern their deployment, balancing innovation and accountability. By proposing practical strategies at the institutional level, the research attempts to guide the ethical adoption of SMDs in an era of rapid technological progress. The following sections examine the ethical implications of AI and IoMT, assessing their respective relative relevance in healthcare systems. The main contribution of this study lies in the systematic prioritization of ethical concerns, such as privacy, fairness, responsibility, and human agency, according to their societal, clinical, and systemic implications. This approach aims to enable stakeholders to address high-impact issues first, fostering trust and ensuring that technological advances are consistent with the core values of safe, equitable, and patient-centered care.

2. Ethical Considerations in AI and IoMT

As the healthcare sector harnesses the revolutionary potential of AI and IoMT, it must navigate an intricate ethical landscape shaped by rapid technological advancement. These innovations promise to redefine patient care through enhanced diagnostics, personalized treatments, and real-time health monitoring, while optimizing operational efficiency across hospitals and clinics. Yet, their integration into healthcare systems introduces profound ethical dilemmas that demand rigorous scrutiny, proactive governance, and multidisciplinary collaboration. Hereby, we explore the ethical dimensions to ensuring that AI and IoMT serve as forces for good, balancing their transformative potential with societal values, patient rights, and long-term sustainability.

Figure 1 encapsulates the multifaceted ethical landscape surrounding the integration of AI and the IoMT in healthcare. These technologies promise transformative advancements, yet their deployment demands rigorous ethical scrutiny to ensure they align with societal values, patient rights, and global sustainability goals. Below is a comprehensive interpretation of the interconnected ethical domains and their implication. The radar chart visualizes ethical considerations for AI-based and IoMT-based systems in healthcare, comparing how each aligns with key ethical dimensions.

AI and IoMT technologies present distinct ethical considerations across key domains. AI scores higher in areas such as patient safety and efficacy due to its central role in diagnosis, decision-making, and predictive modeling, which have direct impacts on patient outcomes. It also raises heightened concerns around informed consent and bias, as patients must understand complex algorithmic decisions, and underrepresented groups risk unfair treatment if training data lacks diversity. In contrast, IoMT excels in regulatory compliance and sustainability, as it involves numerous hardware components subject to strict approval processes and environmental concerns. Both technologies share high stakes in data privacy and security, particularly due to the sensitive nature of health data and widespread connectivity. Equity and accessibility remain pressing for both, though AI may pose greater challenges if algorithmic deployment fails to account for marginalized groups.

Cultural sensitivity, end-of-life management, and innovation risk are also crucial, with AI typically facing greater scrutiny due to its autonomy, while IoMT raises concerns about device access and lifecycle ethics.

3. Prioritizing Ethical Considerations in the Integration of AI and IoMT-Based SMDs

The utility of this prioritization framework lies in its ability to guide healthcare stakeholders, such as developers, clinicians, policymakers, and regulatory bodies, in making informed, ethical decisions throughout the lifecycle of SMDs. Categorizing ethical concerns based on severity, real-world risks, and potential mitigation strategies helps identify which issues demand immediate attention and resource allocation. For example, prioritizing patient safety, data privacy, and informed consent ensures that foundational principles of medical ethics are upheld, thereby reducing harm and fostering trust. Additionally, it enables proactive planning for moderate or context-dependent issues like cultural sensitivity or sustainability by integrating ethical foresight into design and deployment. This structured approach enhances transparency, accountability, and inclusivity in digital health innovation, ultimately supporting safer, fairer, and more socially responsible technology adoption in healthcare systems.

The

Table 1. Structured application of ethical considerations to SMDs mapped to their severity.

Ethical Consideration	Application to SMDs	Severity	Examples and Risks	Mitigation Strategies
Patient Safety Efficacy	Risks from software malfunctions, hacking, or algorithmic errors in devices (e.g., pacemakers, insulin pumps.)	High	A hacked insulin pump delivering lethal doses.	Rigorous pre-market testing (e.g., FDA’s Safer Technologies Program).
			AI misdiagnosing cancer due to flawed training data.	Cybersecurity hardening (e.g., encryption, regular updates).
				Clear, jargon-free consent forms.
Informed Consent	Users must understand how data is collected, shared, and used (e.g., continuous monitoring by wearables).	High	Patients unaware their smartwatch data is sold to third parties.	Opt-in/opt-out controls for data sharing.
				End-to-end encryption.
Data Privacy	IoT devices transmit sensitive health data (e.g., glucose levels, heart rhythms) vulnerable to breaches.	High	Hackers accessing a cardiac implant’s Bluetooth interface.	Anonymization of datasets.
				Subsidized pricing models.
Equity	High costs and digital literacy gaps exclude marginalized groups from AI-driven diagnostics or remote monitoring.	Moderate-High	AI diabetic retinopathy tools unavailable in rural areas with poor internet.	Offline functionality.
				Community training programs.
				Diverse training datasets.
Bias	Biased algorithms in skin cancer apps or diagnostic tools due to underrepresentation of diverse populations in training data.	Moderate-High	Pulse oximeters overestimating oxygen levels in darker-skinned patients.	Open-source algorithm auditing.
				Explainable AI interfaces.
Regulatory	Balancing innovation speed with compliance to evolving standards (e.g., EU MDR for software-as-a-medical-device).	Moderate	Unapproved AI radiology tools causing misdiagnoses.	Proactive engagement with regulators (e.g., FDA’s Digital Health Pre-Cert Program).
				Modular design for upgrades.
Sustainability	E-waste from obsolete wearables or non-recyclable components in implants.	Context-Dependent	Discarded smart glucose monitors leaching toxins into landfills.	Take-back programs.
				Biodegradable materials.
				Co-design with local communities.
Cultural Sensitivity	Devices conflicting with cultural norms (e.g., fertility trackers opposed in conservative communities).	Context-Dependent	Smart contraceptive apps rejected in regions with stigmatized family planning.	Customizable privacy settings.
				Clear legal frameworks for data ownership.
End-of-Life	Ethical deactivation of IoT-enabled life-support devices or deletion of cloud-stored patient data after death.	Context-Dependent	Family disputes over turning off a smart ventilator.	Advance directives for device deactivation.
				Staged clinical trials.
Risk	Pressure to adopt unproven AI updates in devices (e.g., neural implants) without long-term safety data.	Moderate	Experimental brain-computer interfaces causing unforeseen side effects.	Ethical review boards for iterative updates.
				Post-market surveillance.

provides a comprehensive overview of the ethical challenges associated with SMDs powered by AI and IoMT, systematically evaluating each consideration based on its application, severity, risks, and mitigation strategies. High-priority concerns such as patient safety, informed consent, and data privacy highlight the urgent need for secure design, transparent communication, and robust Cybersecurity, as failures in these areas can result in life-threatening outcomes. Moderate to high risks include bias, equity, and regulatory compliance, where algorithmic discrimination and unequal access can undermine fairness and effectiveness, necessitating inclusive datasets, explainable AI, and regulatory alignment. Context-dependent issues like sustainability, cultural sensitivity, and end-of-life management reflect the importance of adaptability and local engagement to ensure ethical alignment with societal values and environmental responsibility. The table ultimately serves as a strategic guide for healthcare stakeholders, enabling them to prioritize high-impact ethical domains while fostering responsible and equitable innovation.

4. Conclusions

In conclusion, AI and IoMT in healthcare have immense potential to revolutionize patient care and healthcare delivery. However, their responsible and ethical use depends on considering the multifaceted considerations described in this manuscript, particularly through the conscious prioritization of competing values and goals. By focusing on patient well-being, transparency, fairness, and equitable resource allocation, stakeholders can find complex trade-offs, such as balancing innovation and safety, efficiency and privacy, or scalability and personalization of care. Healthcare systems, technologists, policymakers and professionals must work together to establish clear ethical hierarchies, ensuring that key priorities such as patient safety and efficacy, societal benefit, and harm mitigation guide decision-making.

As we move forward, the ethical pathway for AI and IoMT must remain dynamic. Continuous reflection, adaptation, and strategic prioritization will be essential to harness the full potential of these technologies while preserving well-being and trust. In the ever-changing healthcare landscape, ethics and intentional prioritization must remain central, so that AI and IoMT serve as tools for empowerment and healing. Only by rigorously prioritizing ethical imperatives can we ensure that these advances equitably benefit individuals and society now and in the future.