Integrating AI Technologies into Remote Monitoring Patient Systems ^†

Abstract

Improving RPM through AI includes various aspects of healthcare delivery that make systems more efficient, accurate and patient-centric. In this work, the impact and role of AI is explored with a focus on RPM. As a result of the research, it was found that the AI-supported architectures in building RPM have transformed, augmented, and revealed new possibilities of applications and benefits in remote health monitoring. Nine groups of significant AI applications leading to the transformation of remote patient care are identified, analyzed, and discussed. Challenges facing RPM are also discussed. Addressing these challenges requires collaboration among healthcare providers, technology developers, policymakers, and patients to ensure the successful implementation and widespread adoption of remote patient monitoring. The results of this research will allow for an informed decision about the need, benefits, and effectiveness of building a specific AI-based RPM and developing such an architecture with the necessary applications for the specific medical organization.

1. Introduction

Today, discussing and writing about artificial intelligence (AI) is fashionable. Every day, we encounter applications and their integration in various fields of knowledge, which expands their impact on human life. At its core, AI is the simulation of human intelligence by computers or robots programmed to mimic cognitive functions such as learning and solving problems to automate and improve data processing, interpretation, and analysis [1]. These systems and programs, via processing data, can extract knowledge, solve problems, make decisions, and perform various tasks using algorithms that simulate human thinking. AI can be implemented through various technologies and approaches, such as machine learning, genetic algorithms, expert systems, neural networks, and many others. It is used in various fields such as healthcare, finance, automation of production processes, transportation, education, and others.

One of the fastest growing areas in health and healthcare is RPM systems. With the help of various medical sensors and devices, these systems can remotely monitor patients’ health status in real time. Traditionally, RPM systems have been applied to remote monitoring of patients in hospitals during and after surgery, as well as in intensive care units or home care. In these cases, special devices and wireless body sensors are used. They are also particularly useful in pre-hospital care for monitoring the condition of chronically ill patients from a distance, such as patients with difficulty moving. This increases accessibility and convenience for patients, especially in cases of home treatment or when constant monitoring is needed in distant places. The structural diagram of such a system is shown in Figure 1. RPM systems use various technologies to collect data on patient health and transmit them to medical staff for analysis and interpretation, allowing patients’ health to be monitored remotely. This is achieved through integrating various components and technologies, new IoT methodologies in healthcare such as telehealth applications, contact-based wearables and sensors, wireless connectivity and the Internet, software platforms, and visualization interfaces.

In real-time, RPM systems monitor vital signs such as heart rate, blood pressure, breathing, temperature, and blood oxygen levels. Likewise, they offer automatic generation of alarms in case of deviation from normal values of vital signs, which helps in the quick detection of problems and an immediate response by medical personnel. An important feature is their ability to integrate with health information systems stored in the Data Center, allowing automatic data entry and integrated access to patient medical records. Medical professionals can also determine the patient’s health status based on information from digital health records. The data collected by the RPM system is transmitted to specialized software platforms that allow the medical staff to analyze and visualize the collected data. These systems have high data security requirements based on regulatory requirements. They should be easy for both medical staff and patients to use. Easy navigation and an intuitive interface are essential, especially in stressful situations or emergencies. RPM systems play an important role in improving the quality of patient care, reducing hospitalizations and optimizing medical resources.

This study examines the opportunities that emerge from integrating AI-based technology applications and their impact on RPM. The literature review shows an elevated interest in implementing new developments using new engineering and software approaches, both on the part of the global academic community and business.

This research explores AI’s transformative nature in various aspects of the functioning and organization of RPM, medical staff, and patients.

2. Related Works

2.1. Economic Benefit and Significance

The research in [2] focused on the benefits to the patient. A significant reduction was found in the waiting time and the financial cost of a visit to a medical center or the emergency department for patients when implementing RPM. The author in ref. [3] explores the possibilities of predictive analytics and decision support systems in patient care. The research confirms that AI facilitates remote patient monitoring by enabling real-time tracking of patients’ vital and health indicators. The predictive analytics makes it possible to identify potential health problems before they escalate, allowing for proactive interventions. Application of these approaches reduces hospital re-admissions and healthcare costs.

2.2. Sensors and Distributed Sensor Networks within IoT

The authors in [4] consider AI a new paradigm for distributed sensor networks within the IoT. The authors justify this new theoretical model with the convergence of the architecture, techniques, and platforms for IoT, sensors, devices, and energy approaches for IoT, as well as the convergence of networks and communication for IoT. Convergence creates problems and questions regarding integrating IoT devices and AI, such as data security, privacy, and interoperability. It is appropriate to improve the efficiency of the application of AI in systems built with IoT devices and sensors, as follows:

• Sensory intelligence of sensor devices by incorporating learning algorithms that allow sensors to adapt, learn and make autonomous judgments based on changing data patterns.
• Increasing the energy efficiency of sensors through the use of renewable energy sources, optimization of algorithms for low energy consumption, implementation of low-power communication protocols through optimization of design and resources, reducing the need for frequent battery replacement [5].
• Enhance data security and privacy and promote interoperability by creating strong encryption approaches, authentication protocols, and privacy-preserving algorithms. Sensitive sensor data can be protected from unauthorized access or alteration through robust security measures, training, differentiated privacy, and safe aggregation [6].

The status of intelligent technology-enhanced healthcare sensors under development based on novel materials, device structure, system integration, and AI application scenarios is presented in ref. [7]. The authors analyze advances in wearable sensors for monitoring breathing rate, heart rate, pulse, sweat, and tears, implantable sensors for cardiovascular care, neural signal acquisition, and neurotransmitter monitoring, and soft wearable electronics for precision therapy. Through the integration of IoT, AI, and medical sensors, a closed-loop system can be realized for real-time monitoring, data collection, online analysis, diagnosis, and treatment recommendations. One of the key benefits of AI-enhanced sensors is the realization of personalized healthcare, which facilitates the real-time monitoring of vital signs, enabling rapid detection of abnormalities and timely intervention, thereby saving lives. A significant advantage of modern medical sensors is their seamless integration, AI algorithms, and wireless communication, enabling them to actively participate in managing patients’ health.

2.3. The Underutilization of AI for Medical Needs

Ref. [8] examines how governments and public agencies are addressing the underutilization of AI to improve people’s digital health. The authors share that the primary reasons include lack of trust, skepticism, fear, the administrative burden, the financial disadvantage due to the high cost of investment in hardware, software updates, maintenance, the need for trained personnel to work with AI systems, budgetary requirements, difficulties in financing innovation in the structure itself, and many others related to the clinical staff, doctors, or hospitals. According to the authors, the main drivers for the insufficient use of technology in the health sector are professional reluctance, security of personal data, lack of legal regulations, standards, and infrastructures, and public disbelief in an age of conspiracy theories.

2.4. Guidelines for the Development of AI in Health and Healthcare

As a result of the research in [9], specific considerations for developing AI systems in healthcare are proposed. These suggestions include the following:

• Automation of only those health tasks and roles for which automation is most suitable, such as radiologists, GPs, emergency physicians, cardiologists, oncologists, and imaging specialists.
• Develop personalized AI-based systems for training and upskilling human service providers. The authors believe that the automation of activities through AI potentially removes a fundamental stage of clinical learning, resulting in dehumanized clinicians who may be poorer learners and decision-makers.
• Assessment of the need for health care of people, leading to developing or maintaining the skills of clinical specialties.
• Automating tasks that allow doctors to focus on more satisfying work.
• AI systems will demonstrate improved patient outcomes and healthcare impacts on patient relationships and care.

2.5. Legislation on AI in Healthcare

The need to create a legal framework for accessing and sharing information for AI applications, developing legal standards to protect each patient’s personal information, and using these data only collectively for public purposes is discussed in ref. [1]. The authors note that social consensus needs to be reached on critical aspects of AI, including data sharing, privacy, and accountability. Another important conclusion related to the implementation of AI technologies is the reduction in jobs that will be automated as a result of the application of AI. According to the authors, these personnel should be redirected to other hospital and pre-hospital care positions. It is also important to address the education of doctors, nurses, healthcare professionals, and patients as early as university by introducing immediate training on the application of AI in healthcare.

The authors [10] discuss the opportunities and challenges for the development and safe implementation of AI in healthcare, reaching conclusions in support of increased security, such as the development and implementation of legislation for the protection of health and medical data, introduction of models for personalized risk assessments, calibrated and efficient application update protocols, determination of risk and uncertainty for all possible applications, automated systems and algorithms to be able to adjust and respond to uncertainty and unpredictability, and the introduction of learning not only for AI-based systems, but also for clinicians and medical and health personnel to use AI technologies.

As a result of the many proposals and fears about the deployment of AI for medical and healthcare needs, an EU law on AI was adopted in March 2024, the first-ever comprehensive legal framework for AI worldwide. The new rules aim to promote the application of trustworthy AI in Europe and beyond, ensuring that AI systems respect fundamental rights, safety, and ethical principles and address the risks of very powerful and impactful AI models [11].

3. Applications on AI for RPM

3.1. Remote Monitoring of the Patient, Based on AI

RPM using AI requires integrating various elements and developing software and hardware solutions to ensure monitoring systems’ reliability, efficiency, and security. Traditional deep learning and machine learning are the main methods used in RPM AI applications for detecting and predicting vital signs and classifying patients’ physical activities. These systems can be contact or non-contact, monitoring vital signs such as heart rate, pulse, respiratory rate, blood pressure, and blood oxygen volume, as the deterioration in these vital signs affects the human health system. There is also a trend for AI-based systems to detect various patient activities, such as falls and mobility-related illnesses [12]. The authors also note that AI has enormous potential for providing various health services, assessing disease risk, predicting early deterioration of health, providing ongoing patient care, and reducing complications during disease progression. AI-powered wearables and sensors collect continuous data from vital signs, health metrics, and other activity levels, providing a comprehensive real-time view of a patient’s health. Machine learning algorithms analyze these data to detect anomalies and trends, alerting healthcare providers to deviations from normal parameters. AI algorithms can be applied to large data sets to identify and analyze patterns and trends in patient data, facilitating personalized medicine by adapting treatment plans to patients based on their characteristics and medical history [13].

The critical components for building an RPM system are shown in Figure 2 and include the following:

• Sensors and monitoring devices. These can be wearable sensors for heart rate, blood pressure, blood oxygen levels, temperature, and other biometric data, as well as video surveillance cameras.
• Internet of Things (IoT) and connectivity. Sensors and devices connect to a central platform via the Internet, allowing continuous patient data tracking.
• Cloud infrastructure. The data collected by the sensors are sent and stored in a cloud infrastructure where they can be processed and analyzed.
• AI and data analytics. AI algorithms can continuously analyze data from wearable sensors and other remote monitoring devices to track vital signs, activity levels, sleep patterns, and other health indicators in real time. This allows medical staff to intervene immediately if abnormalities or potential problems are detected in the patient’s health. Only machine learning algorithms can be included to predict the development of diseases or warn of impending dangerous conditions.
• Interface for healthcare professionals and patients. RPMs are often integrated with health information systems, enabling automatic data entry and integrated access to patient medical records. The information analyzed by the system should be accessible to health professionals and patients themselves, especially for home treatment or the need for constant monitoring in distant locations. This can be done through web interfaces, mobile applications, or other types of software applications.
• Security and data protection. The collection and processing of medical data require high security and protection of patients’ data. Such a system must comply with legal data protection requirements, such as the (GDPR) General Data Protection Regulation in the European Union.

3.2. Integration of AI in RPM

The integration of AI has led to transformative improvements such as image analysis systems, machine learning algorithms to predict health risks, and AI-driven chatbots to provide consultations [12]. Predictive analytics and decision support systems enable early diagnosis and disease prevention by identifying risk factors and patterns, contributing to improved patient outcomes and cost-effective healthcare. AI-facilitated remote patient monitoring enables proactive health interventions by tracking and identifying potential health problems and vital signs in real-time [3,12]. Integrating AI in RPM can bring benefits, positives, and challenges for patients, medical staff, and medical service providers. These results can be to systematised systematized, according to systematized systematized, as shown in Figure 3.

3.2.1. Real-Time Health Monitoring

Integrating AI with intelligent sensors and wearables enables continuous patient monitoring. AI algorithms can continuously analyze data from wearable sensors, other medical devices, or remote sensing images to track vital signs, activity levels, sleep patterns, and other health indicators in real time. This allows healthcare professionals to receive an immediate alert when irregularities occur in the patient’s vital signs, such as blood pressure, heart rate, and respiratory activity, and to intervene immediately if abnormalities are detected [3].

3.2.2. Early Detection of Health Deterioration Is Vital for Chronically and Acutely Ill Persons [14]

AI models for early detection of health problems are based on algorithms that process and analyze the collected data, considering factors such as current health status, age, gender, and history of diseases. These data include vital signs such as blood pressure, heart rate, and breathing rate. After establishing baselines, the AI system continuously monitors incoming data for deviations from the norm. Thus, AI can identify signs of health deterioration, detect abnormalities, and alert healthcare professionals, enabling immediate, appropriate action. AI can also predict potential health problems based on historical data tendencies. For example, if a patient’s heart rate variability gradually decreases, AI can alert medical professionals to the increased risk of a cardiac event. Early detection of diseases through the integration of AI in RPM leads to the following:

• Timely reaction and intervention when detecting minimal changes in patient’s health status, which can prevent the progression of health problems and reduce the likelihood of complications;
• Reducing the need for patients to be hospitalized, as they can receive medical and health care in their homes, which saves costs for patients and health systems;
• Improving the quality of life for patients by avoiding possible health crises;
• Generating predictability of the patient’s condition, helping medical professionals plan treatment more precisely, and make effective medication adjustments and lifestyle recommendations to patients.

At the same time, it should be noted that the early detection of health problems faces challenges related to the accuracy of AI algorithms in minimizing false positives and negatives, data security, and patients’ active participation in remote monitoring of their health by complying with wearable devices.

3.2.3. Predictive Analytics for Risk Stratification

AI models can use the data to produce predictive analytics that group patients into low-, moderate-, and high-risk groups based on future health events. Such future health events may include re-admission to hospital, exacerbation of chronic diseases, or complications after treatment. By stratifying patients based on their levels of risk, interventions can be prioritized, focusing on high-risk patients who need immediate intervention and allocating resources more efficiently. Predictive analytics is a powerful instrument that combines AI and RPM to identify patients at high risk of adverse health events. By analyzing patient data and patterns, AI and RPM-driven systems can predict potential health complications and enable healthcare providers to intervene proactively. In predictive analytics, AI algorithms analyze documented patient data, real-time health metrics, and relevant information to predict future health outcomes.

Vital components of AI-enabled predictive analytics are as follows:

• Collection of data on vital signs, laboratory results, medications taken, lifestyle habits, and data from health records.
• AI machine learning algorithms to identify patterns, trends, and correlations within the collected data. These algorithms always learn from new data, improving their predictive accuracy over time.
• Risk stratification of patients into three risk groups based on the likelihood of adverse events. Stratification helps efficiently allocate resources and prioritize interventions for each patient. High-risk patients receive personalized care and interventions.
• Generate alerts and notifications when patterns indicating potential health deterioration are detected for immediate interventions to prevent or mitigate adverse outcomes before symptoms become severe.

Predictive analytics helps allocate resources more efficiently, leading to lower healthcare costs, informed decision-making, and reduced re-admissions based on AI-generated predictions. When using predictive analytics, it is important to ensure patient privacy [14,15].

3.2.4. Personalized Treatment and Care Plans

AI algorithms’ processing of large patient databases, including medical history, investigations, genetics, lifestyle factors and treatment responses, medications taken, and results of previous treatments, allows for the generation of personalized treatment and care plans adjusted to the needs and preferences of individual patients. This approach allows for more targeted interventions and improves treatment outcomes. Personalized treatment plans are part of effective health care, especially in managing chronic diseases and complex medical cases. Important to the development of personalized treatment and care plans with the help of AI are the following:

• The integration of data from different sources.
• Predictive analytics: AI algorithms can predict potential outcomes of different treatment plans based on historical medical data. This analysis makes it possible to evaluate the effectiveness of different treatment plans and interventions before their implementation.
• The risk assessment that AI can offer, assessing complications or adverse events depending on individual medical history, to identify high-risk patients who need targeted interventions.
• Treatment recommendations: AI-generated data analysis can produce personalized treatment recommendations. These recommendations may include medication adjustments, lifestyle changes, dietary changes, and physical activity guidelines.

Personalized treatment and care plans through RPM with AI optimize the probability of treatment success and patient satisfaction. They also lead to the optimization of interventions, adjustment of treatment plans based on changing health data, and more efficient allocation of resources because interventions are targeted and tailored, reducing unnecessary procedures and medications. A significant challenge in personalized treatment and care is accurate and comprehensive data and compatibility across systems [14,16].

3.2.5. Medication Adherence Monitoring

Treatment nonadherence is a significant challenge, resulting in compromised treatment outcomes and increased healthcare costs. AI technology is vital in promoting medication adherence and prescribed treatment through personalized patient reminders. Through RPM, AI monitors patient behavior and responses to treatment, identifies adherence patterns, and predicts potential problems with medication irregularity. The following are important for the construction of AI monitoring when taking medicines [14]:

• Analyzing the behavior of engagement patterns by AI algorithms to predict when a patient may forget or miss doses of medication;
• Personalized reminders and notifications sent by AI to patients to take their medications as prescribed [17];
• The integration of data from different sources;
• Predictive analytics, allowing timely intervention before medication irregularity becomes a severe health problem;
• AI patient engagement by providing motivational, educational content for patients that clarifies the significance and importance of regular medication administration [18].

AI applications lead to patient adherence, better treatment outcomes, reduced health risk, and chronic disease management. Improved adherence can lead to fewer hospitalizations, emergency department visits, and disease-related complications, resulting in cost savings for patients and healthcare systems. AI empowers patients to actively supervise their health by providing them with tools and resources to stay on top of their medications. There are different approaches to building nonadherence models, for example, through a chatbot.

3.2.6. Remote Diagnostics and Triage

Remote AI patient diagnosis uses AI technology to remotely assess patients’ medical conditions or health problems without needing in-person consultations. This diagnostic overcomes geographical barriers and improves the efficiency of diagnostic processes. This approach uses AI algorithms to analyze different types of patient data collected from remote monitoring devices, digital health applications, or telemedicine platforms, as shown in Figure 2. Additionally, AI-driven diagnostics have the potential to increase the expertise of healthcare specialists, leading to more accurate and timely diagnoses, ultimately improving patient outcomes and satisfaction. AI diagnostic tools can analyze remote symptoms, images, and test results, predict diagnoses, conduct remote consultations, and guide patients to proper levels of care. AI is integrated into the remote diagnosis process as follows:

• Continuous monitoring, collection, and analysis of patient data and integration with electronic health records;
• The predictive analytics offered by the AI algorithms enable proactive interventions to prevent adverse outcomes;
• Decision support, such as recommendations for diagnostic tests, treatment options, or adjustments to treatment regimens based on the patient’s current condition and historical data;
• Automated patient triage and service and care for patients with the greatest need [19];
• Personalized care by adapting care plans to the individual needs of patients;
• Remote consultations: AI virtual assistants can facilitate remote consultations between patients and medical staff by collecting relevant information, answering common questions and facilitating communication [20].

3.2.7. Behavioral Health Monitoring

AI algorithms can analyze patients’ digital interactions, such as social media activity, online searches, or smartphone usage patterns, to assess their mental health status and detect signs of depression, anxiety, or other behavioral health conditions. This non-invasive approach allows continuous monitoring of patients outside the clinical setting. Behavioral health monitoring is the systematic observation, assessment, and tracking of an individual’s mental health and behavior over time [21]. AI drives web-based and smartphone apps, mostly used for self-help and guided cognitive behavioral therapy for anxiety and depression [22]. This involves using various instruments and techniques to gather information about a person’s emotional state and psychological and behavioral patterns. Key aspects of AI monitoring of behavioral health include the following:

• Initial assessment of the individual’s mental health. Assessment tools may include questionnaires, interviews, standardized tests, and observations by health professionals. The assessment aims to identify any existing mental health conditions, symptoms, risk factors, or areas of concern.
• Monitoring and tracking behavior and symptoms over time. Monitoring can be done through regular check-ins with healthcare providers, self-monitoring tools such as mood diaries, or mobile apps or wearables that track physiological indicators such as heart rate or sleep patterns.
• Intervention aimed at managing mental health problems or preventing their escalation, based on information from behavioral health monitoring. Interventions may include psychotherapy, drug treatment, lifestyle changes, stress management techniques, or referral to other mental health professionals or support services.
• Prevention to prevent the onset or recurrence of mental health problems. By identifying risk factors or early warning signs, implemented strategies can mitigate these risks and promote mental health [23].
• Analyzing data to identify trends, patterns, or correlations to assist in program development or public health initiatives.
• Feedback on and correction of patients’ progress and adjustment of treatment plans or strategies as needed.

Behavioral health monitoring is a proactive approach to promoting mental health and well-being by carefully observing and responding to changes in behavior, emotions, and psychological functioning. It plays a crucial role in early intervention, ongoing support, and treatment planning for individuals with mental health problems.

3.2.8. Automation of Administrative Tasks

AI can automate several administrative tasks in healthcare facilities, streamlining operations, improving efficiency, and reducing the administrative burden for healthcare professionals, allowing medical staff to focus on core medical tasks [24]. Here are a few ways AI is being applied in this context:

• Optimization of medical processes, and scheduling of examinations, appointments, and medical operations. In the context of these tasks, AI in remote monitoring systems can analyze historical data, patient preferences, and medics’ time to optimize the planning of examinations and operations. This helps reduce wait times, prevent overbooking, and improve overall patient satisfaction.
• Medical transcription AI-driven speech recognition technology can automatically transcribe medical dictations, converting spoken words to text. This reduces the time and effort required for manual transcription.
• Billing and coding of medical services provided through AI algorithms that can review medical records and automatically assign appropriate billing codes to medical activities performed. This helps ensure billing accuracy while reducing errors and minimizing the risk of lost revenue due to coding discrepancies.
• Electronic health record (EHR) management. AI can enhance EHR systems by automatically organizing and summarizing patient data, extracting relevant information from unstructured clinical notes, and facilitating data entry through voice recognition or natural language processing technologies.
• Patient triage and routing. AI-driven chatbots or virtual assistants can interact with patients, sort through their symptoms, provide basic advice, and direct them to the appropriate medical healthcare resources or staff.
• Resource allocation: AI algorithms can analyze patient flow, resource utilization, and staffing patterns to optimize resource allocation in healthcare facilities. This includes predicting patient volume, identifying potential bottlenecks, and reallocating resources to improve operational efficiency.
• Fraud detection. AI analytics algorithms can analyze healthcare claims data to detect patterns or practices that indicate fraudulent activity, such as billing for unnecessary procedures or services. By flagging suspicious claims for further investigation, AI helps mitigate financial losses and maintain the integrity of healthcare systems.

3.2.9. Training and Development of Medical Staff

Integrating AI into the training and development of medical staff can give them the knowledge, skills, and tools needed to deliver high-quality care in a digital health environment and improve their ability to effectively use and interpret data collected from such systems [25]. AI can be incorporated into training and development programs for medical staff in the context of remote patient monitoring through the following:

• AI-based training modules that provide nursing staff with comprehensive instructions for effectively using RPM. These modules may include interactive simulations, real-world case studies, and quizzes to reinforce learning.
• Data interpretation skills using AI-based analysis tools. This includes identifying patterns, trends, and anomalies in patient data to make and take informed decisions about patient care.
• Customized training modules and programs for individual nursing staff members based on their learning preferences, knowledge gaps, and clinical responsibilities. This ensures that training is tailored to their specific needs and goals.
• Distance learning through AI virtual reality (VR) or augmented reality (AR) simulations to provide remote training sessions to medical staff, especially those working in geographically dispersed locations or at times when in-person training is not feasible.
• Quality assurance and compliance monitoring of care provided by medical staff. This includes ensuring compliance with established protocols, identifying areas for improvement, and mitigating risks related to patient safety and data privacy.
• Faster research and development collaboration between medical staff and AI developers and co-creation of innovative remote patient monitoring solutions. This collaborative approach allows medical staff to contribute their clinical expertise while leveraging AI technology to improve patient care.

Observations show that AI has yet to be widely applied in practical training, so it is necessary to strengthen the theoretical guidance of medical education and synchronize it with the rapid development of AI. Medical educators should provide advanced training to medical students on new technologies applicable to medicine. Research teams should be from multidisciplinary fields to ensure the applicability of AI in medical education.

4. Challenges of Remote Patient Monitoring

• Barriers connected with technology: Not all patients have access to or are comfortable using the necessary technology for remote monitoring, such as smartphones or wearable devices. This can create disparities in healthcare access.
• Privacy concerns and data security: The collection and transmission of sensitive health data raise concerns about privacy and security. Ensuring a response from regulations like HIPAA is crucial to protect patient information.
• Integration with existing healthcare systems: Integrating RPM data with other systems connected to healthcare and electronic health records (EHRs) can be challenging. Seamless integration is necessary for healthcare providers to view a patient’s health comprehensively.
• Reimbursement and payment models: The lack of standardized reimbursement models for remote patient monitoring services can hinder widespread adoption. Establishing fair and consistent reimbursement is essential for healthcare providers to invest in RPM.
• Provider workload and alert fatigue: Continuous monitoring generates significant data, leading to potential alert fatigue for healthcare providers. Effective systems for prioritizing and responding to alerts are needed to avoid overwhelming healthcare professionals.
• Patient engagement and compliance: Keeping patients engaged and motivated to participate actively in remote monitoring is challenging. Some patients may become complacent or disinterested over time, impacting the effectiveness of the monitoring program.

5. Conclusions

AI plays a key role in RPM, revolutionizing how the healthcare sector can monitor and manage patient health remotely. AI transforms patient care and contributes to clinical impact within specific healthcare contexts and nursing staff.

This work explores the impact and role of AI with a focus on RPM. The research found that AI-supported architectures in building RPM have transformed and augmented RPM and revealed new possibilities of applications and benefits in remote health monitoring.

Nine significant AI applications transforming remote patient care are identified, analyzed, and discussed.

Remote patient monitoring systems face challenges, which are also discussed. Addressing these challenges requires collaboration among healthcare providers, technology developers, policymakers, and patients to ensure the successful implementation and widespread adoption of remote patient monitoring.

The results of this research will allow an informed decision about the need, benefits, and effectiveness of building an AI-based RPM and developing such architecture with the necessary applications for the particular medical organization.