A Review of Sensor Technologies and IoT Innovations in Telemedicine & E-Health (IoMT): Advancements in Healthcare Instrumentation, Telehealth and E-health Sensor Devices

In contemporary society, terms like IoT, big data, and smart sensors permeate nearly every aspect of human activity, whether generating information, processing it, or enabling smart computing actions on sensory devices. While attention is often focused on smart computers, mini-computers, or wearables, one underexplored area is the sensors that enable interconnected systems to provide essential data points and process them in real-time. The term IoT refers to components and devices that connect various systems, distributing their input and output across different nodes on a shared network. Big data, on the other hand, addresses the “3Vs”-volume, velocity, and variety-of the data generated by these interconnected systems. Data points are constantly produced by network nodes, varying in quantity, quality, and sampling rate. The computing power needed to generate, aggregate, and analyze this data depends heavily on the capabilities of the devices involved. “Smart sensors” are devices equipped with feedback or advanced functionality, such as processing raw data or automating outputs to achieve the desired result. These sensors are crucial for efficient data management in IoT systems. These systems have many use cases in our everyday life as well as our health. As a result, IoT can be used as the backbone of fusing data for e-health applications.

Recently, these technologies have increasingly formed the backbone of AI processes. While AI-particularly with the rise of systems like ChatGPT and Generative Adversarial Networks (GANs)- has gained significant attention, understanding the infrastructure that supports these operations is now more important than ever [1]. Telemedicine stands for improving the methods of diagnosing, treating and monitor patients. In order to achieve this, in most telemedicine applications, AI and other resource-intensive solutions are used that require significant processing power and energy. This article focuses on defining sensors, examining their generations and advancements, and identifying low-cost, energyefficient mini devices that can be used to build an sensor network infrastructure. These solutions are invaluable for healthcare enterprises, researchers, and educators as scalable, reliable, and affordable tools, to assist scientists into understanding where these AI operations are executed and what are the parts of IoT in healthcare [2-4]. This becomes even more critical as the concept of IoT evolves into the Internet of People (IoP), where human activities function as sensor nodes, providing the necessary input for sensory devices based on our daily needs thus the broad interaction of each patient in terms of monitoring its telecare [5].

Sensors are devices designed to detect physical quantities and produce measurable outputs. The concept of sensors has evolved from the human senses-what we hear and see-to sophisticated instruments capable of precise measurement, addressing a wide range of everyday needs. Since their inception, sensors have played a pivotal role in our lives, transitioning from rudimentary tools to technological marvels. The term “smart sensor” emerged to describe devices with built-in information processing capabilities, distinguishing them from earlier models. Selecting the right sensor for a specific application involves carefully considering key characteristics such as accuracy, tolerance, linearity, distinctness, repeatability, and sensitivity. A highly accurate sensor especially in telemedicine and e-health application must provide minimal errors, while tolerance measurements must assess its ability to withstand variations. Linearity ensures a consistent relationship between input and output, while distinctness, repeatability, and sensitivity contribute to a sensor’s ability to detect and measure changes precisely [6-7]. Based on their advancements and new features, sensors can be classified into several generations:
a. First Generation: These smart sensors primarily relied on external electronic circuitry for signal processing and amplification. They were often physically connected to their installation sites, which limited their flexibility.
b. Second Generation: This generation introduced a degree of autonomy, with sensors placed remotely and connected to analog circuitry for on-site adjustment and modulation.
c. Third Generation: Marking a significant leap forward, this generation integrated signal determination modules onto the same physical platform as the sensor. However, analog-todigital conversion and microprocessor functions remained external.
d. Fourth Generation: This generation further refined integration by combining regulation circuits with monolithic or hybrid integrated circuits. Although the transducer and digital processing circuits remained separate, the output could be bidirectionally interfaced with a microprocessor for automated control.
e. Fifth Generation (Current): The latest generation represents the pinnacle of integration, with the analog-to-digital converter now located within the same monolithic or hybrid integrated circuit as the signal conditioner. These sensors produce digital signals that can be directly communicated with microcontrollers and modern computer systems. Through communication buses or wired networks, multiple sensors can be integrated into a host system, providing extensive data

The key advantages of fifth-generation smart sensors include:
a. Multiple Signals: The ability to gather data from various sensors simultaneously.
b. Self-Correction: Automatic detection and correction of errors caused by environmental factors or component aging.
c. Seamless Integration: Compatibility with modern computer systems and communication networks.
d. Advanced Functionality: The ability to perform complex tasks and analyses.

As technology advances, we can expect even more sophisticated smart sensors to emerge, revolutionizing industries such as healthcare, manufacturing, environmental monitoring, and autonomous systems. The sensor generations and the keywords presented above for the latest sensor generation are presented in Figure 1.

Figure 1:Generations of sensors and key characteristics of the latest sensor generation.

In previous sections, we explored the types of sensor devices used to develop a telemedicine application. When discussing an application, it’s essential to consider both the software componentssuch as a GUI or an API for patient measurements-and the hardware used in these devices. Although we have not presented casespecific machinery used in telemedicine (also referred to in recent literature as telehealth or telecare), it’s important to understand the components that make up these systems. The technology and infrastructure supporting secure, reliable telemedicine platforms can be divided into two main parts: the front end and the back end. The front end refers to the application interface, while the back end comprises various components explained in this paper, including IoT sensors and devices. The IoT primarily constitutes the back end, where sensors are used to collect data. Earlier generations of sensors provide less accurate and “smart” data but can capture measurements quickly and at a lower cost. The latest sensor generation, however, offers more advanced functionality, delivering feedback, or even processing data directly on the node, rather than merely providing raw data samples [8]. This is particularly important for telemedicine, as these systems can provide critical, case-specific data essential for patient care. Telemedicine for intensive care, for example, is not simply an IoT platform (a network of interconnected devices collecting data points) but rather a complex system architecture that also incorporates AI and Big data. This allows smart sensors to analyze data at the source before sending it to the front-end application [9]. Healthcare applications for these technologies are numerous, ranging from heart monitors with real-time reporting and medical alert systems to ingestible sensors, medication dispensers, and a variety of wireless sensor nodes. The goal is to leverage this technology to enable real-time tracking of patients’ vital signs and to provide alerts, potentially preventing harmful events [10-11].

Beyond the types of sensors themselves, it’s important to consider the devices that can quickly, reliably, and cost-effectively build an infrastructure for testing, monitoring, and controlling sensor networks. In recent years, mini computing devices like the Raspberry Pi have become the go-to solution for these purposes. Due to its balanced performance, extensive community support, and cost-effectiveness, Raspberry Pi has stood out among minicomputers [12]. While other options, such as the Onion Omega2+ [13] and ASUS Tinker Board [14] offer unique advantages, Raspberry Pi’s combination of storage, speed, processing power, and ease of use makes it especially suitable for educational applications. Devices like the Omega2+ offer compelling features, but their limitations in processing power and community support have restricted their suitability for certain projects. For future reference, when selecting devices for sensor development, particularly in telemedicine and eHealth applications, it is crucial to consider the specific needs of each project. Future research should explore a comparison of low-cost, low-power devices like Raspberry Pi with even more affordable options, examining their potential applications across various domains [15-16]. These advancements will help build the sensor networks that will serve as the IT, cloud, and data middleware infrastructure for the future. Although the end goal is to enable real-time tracking of patients and transition IoT into the Internet of Medical Things (IoMT), the objective is also to enable medical devices to communicate autonomously over a network, using patient data and activities with minimal to no direction from the patient. As such, telehealth should not only be about real-time interactive communication between patients and healthcare professionals but also about tools that provide accurate readings and precise actions for performing actual procedures or, alternatively, building a robust electronic medical record.

1. Hemalatha K, Amanullah M (2024) Effective hybrid deep learning model of GAN and LSTM for clustering and data aggregation in wireless sensor networks. Int J Sensors Wireless Commun Control 14(2): 122-133.
2. Portilla L, Loganathan K, Faber H, Eid A, Hester JG, et al. (2023) Wirelessly powered large-area electronics for the internet of things. Nature Electronics 6(1): 10-17.
3. Roy SS, Puthal D, Sharma S, Mohanty SP, Zomaya AY (2018) Building a sustainable internet of things: Energy-efficient routing using low-power sensors will meet the need. IEEE Consum Electron Mag 7(2): 42-49.
4. Wan T, Shao B, Ma S, Zhou Y, Li Q, et al. (2023) In‐sensor computing: Materials, devices, and integration technologies. Adv Mater 35(37): e2203830.
5. Akan OB, Dinc E, Kuscu M, Cetinkaya O, Bilgin BA (2023) Internet of Everything (IoE)-from molecules to the universe. IEEE Commun Mag 61(10): 122-128.
6. Formica D, Schena E (2021) Smart sensors for healthcare and medical applications. Sensors 21(2): 543.
7. Azizan A, Ahmed W, Razak AHA (2024) Sensing health: A bibliometric analysis of wearable sensors in healthcare. Health and Technology 14(1): 15-34.
8. Ghemari Z (2024) Applications of vibration sensors in medicine: Enhancing healthcare through innovative monitoring. Prog in Med Devices 2(2): 83-88.
9. Guo X, Sun Y, Sun X, Li J, Wu J, et al. (2024) Doping engineering of conductive polymers and their application in physical sensors for healthcare monitoring. Macromolecular Rapid Communications 45(1): 2300246.
10. Karthikeyan S, Raj RA, Cruz MV, Chen L, Vishal JA, et al. (2023) A systematic analysis on raspberry pi prototyping: Uses, challenges, benefits, and drawbacks. IEEE Internet Things J 10(16): 14397-14417.
11. Sengan S, Khalaf OI, Priyadarsini S, Sharma DK, Amarendra K, et al. (2022) Smart healthcare security device on medical IoT using raspberry pi. International Journal of Reliable and Quality E-Healthcare (IJRQEH) 11(3): 1-11.
12. Choubey CK, Thakur P, Hussain I, Deshalahre D, Chavda L, et al. (2024) Performance analysis of IoT-based healthcare monitoring system using raspberry Pi. IEEE, pp. 1-8.
13. Kodali RK, Boppana L, Bandaru H, Avuthu SD (2021) Low cost IoT application using onion omega2+. IEEE, pp. 1-4.
14. Jahan N, Rupa FY, Sarkar S, Hossain S, Kabir SS (2023) Performance analysis of ASUS tinker and mobile net V2 in face mask detection on different datasets. Machine Intelligence and Emerging Technologies, pp. 283-293.
15. Álvarez JL, Mozo JD, Durán E (2021) Analysis of single-board architectures integrating sensor technologies. Sensors (Basel) 21(18): 6303.
16. Mbunge E, Muchemwa B, Batani J, Jiyane S (2021) Sensors and healthcare 5.0: Transformative shift in virtual care through emerging digital health technologies. Global Health Journal 5(4): 169-177.