Encyclopedia
Please note this is an old version of this entry, which may differ significantly from the current revision.
With technological advancements, smart health monitoring systems are gaining growing importance and popularity. Today, business trends are changing from physical infrastructure to online services. With the restrictions imposed during COVID-19, medical services have been changed. The concepts of smart homes, smart appliances, and smart medical systems have gained popularity. The Internet of Things (IoT) has revolutionized communication and data collection by incorporating smart sensors for data collection from diverse sources.
- IoT
- smart healthcare
- patient mortality prediction
1. Introduction
We are living in an age in which technology has revolutionized the world. The technological revolution has happened in communication, business, education, medical care, and many other areas. Digital technology removes barriers of distance, enlarges thinking, and benefits businesses. In particular, since the spread of COVID-19, the use of modern smart applications has increased massively. Such applications are used for a variety of applications, including medical care. Furthermore, the birth of the Internet is considered a major contribution to technological advancement [1]. The Internet removes barriers to digital technology and provides access everywhere. Social media is another source of information and has a large impact on the young generation. Handy technology provides instant data on social media.
The Internet of Things (IoT) is an emerging field of computer science that has impacted the world in a short period of time. The concepts of smart homes [2], smart-resource-based cities [3], smart driving [4], and smart farming [5] have changed the living and working styles of many people around the world. Smart devices are embedded in smart homes and cities, and things are controlled by smart devices. In the near future, real-time human activities will be monitored by smart devices, and real-time data will be collected by tagging sensors within the human body [6]. Human health can be monitored in real-time by sending the information obtained using the sensors to medical consultants. The Internet of medical things (IoMT) is the field in which health equipment is connected with IoT [7]. Authors have discussed IoMt using wearable devices and AI [8]. This concept gives new directions to the health field and opens up new doors of development. As in the medical field, accurate and timely diagnosis and intervention are the main factors in health-related issues, so substantial work is required to fulfill this gap and develop intelligent systems for health monitoring.
With the development of information technology (IT), medical fields are being revolutionized in developed countries. In a traditional medical system, huge crowds, power consumption, and routine work all burden the system and lead to delays in facilitation. In IoMT, wearable sensor devices are connected to the care provider’s smart devices, and they can monitor the real-time patient health record and treat the patient accordingly [9]. IoMT provides a low-cost and quick solution by remotely monitoring the patient’s health. With the spread and rise of chronic diseases, the medical systems of underdeveloped countries face significant issues in managing large numbers of patients due to a shortage of staffing and resources. In the medical field, quick responses significantly impact saving lives. Figure 1 shows the IoMT mechanism in the physical environment.
Artificial intelligence (AI) is a domain of computer science that induces intelligence in machines and makes smart devices capable of working without human intervention. Smart devices make smooth connectivity using AI and work innovatively. AI processes help to find hidden patterns in the large volume of data received from smart devices [10]. Moreover, the AI process also makes recommendations to improve the performance of the systems. The domains of AI and machine learning (ML) greatly help in solving today’s complex problems. In every field of life, computational systems are designed using AI and ML to solve significant dynamic problems. Combining IoT, AI, and ML can change how people live and interact in their daily lives. In the medical field, large datasets are gathered using smart devices and sensors. AI and ML algorithms are applied to find underlying hidden patterns to diagnose different diseases.
Using AI-based solutions, it has become easy for medical staff to work on large datasets and provide future recommendations to prevent diseases. The artificial intelligence of medical things (AIoMT) combines AI approaches with health diagnosis approaches to help in the medical field. The idea behind AIoMT is to prevent unnecessary stays in the hospital and avoid the associated health charges [11]. With the growth in the population, traditional methods of medical diagnosis need to fulfill the demand of the growing community. Because of the limited resources considering the increasing population growth, finding solutions for the efficient management of these resources is a high priority.
After COVID-19, a remote healthcare system is needed for the real-time diagnosis of various diseases. To improve healthcare facilities, it is the responsibility of academia and industry to make combined efforts to overcome these issues. Some efforts have been made in the recent past to design a smart bed concept and smart point-of-care (PoC) devices, but due to population growth, further efforts are needed. According to the world health organization (WHO), 17.9 million deaths are recorded yearly because of cardiac diseases [12]. Heart disease is a major challenge in the medical field. Heart disease arises due to the narrowing of arteries. Heart failure is the primary cause of death in most underdeveloped countries due to inadequate instant response. Old-aged people are the predominant group of heart patients; however, young patients might also be victims of this disease. In the US, one of every nine deaths are caused by heart disease [13]. Chest pain and fatigue are the primary symptoms of this disease, but it happens even without such symptoms. There is a need to develop a remote health monitoring system for heart patient that continuously monitors the patient’s activity. In the case of an emergency, remote health-monitoring systems can require an instant response. IoMT provides a lot of success in medicine, especially in rural areas where providing instant help is challenging. Early disease detection and medication can save a lot of lives. Remote health monitoring also reduces the cost of routine diagnosis due to eliminating the traveling cost and other medical overhead.
Figure 1. IoMT mechanism.
The IoMT is a multifaceted domain of IoT and medical technique and faces many challenges. Medicine is a sensitive field where the survival of human beings is essential. Therefore, a remote monitoring system in IoMT with AI is needed. The software systems in medical areas must be appropriately and extensively verified to meet medical standards. The IoMT-induced AI system must satisfy the following criteria:
-
Security The AI-inspired remote monitoring system of cardiac disease must be secure. It must not be vulnerable to adversarial attacks by other devices. The exact measurements it provides are the main advantage of these systems.
-
Reliablity The reliability of medical software systems must be achieved. A non-reliable system cannot be used in any field, particularly when the privacy and confidentiality of patients are concerned. Internal implementation issues or external problems should be ensured regarding accurate measurements. An auto-correction algorithm must be implemented in every system to avoid damage.
-
Safety The remote monitoring software must be economical and not affect the environment. The system must be human-friendly so that there is no negative impact on human lives. The medical operator, user, and designer should be able to interact with the system without harm.
2. Internet-of-Things-Based Smart Monitoring System
In this modern age, multiple technologies have been developed to monitor medical data. With the invention of sensors, the development has been taking place at a rapid pace. Sensors are used for real-time monitoring and data collection. Smart devices are full of sensors and are used for data collection. In [15], a technique based on a ring sensor is used to monitor the patient. A wearable ring containing the sensor is used for real-time monitoring of cardiac patients. In [16], a technique based on an ear sensor is used for continuous monitoring of heart patients. These sensors are small in size and wearable among persons of different ages alike which makes the monitoring process easy.
Heart patients need an instant response in case of an emergency. Most fatalities happen due to delays in early diagnosis and inefficient methods. In [17], the authors designed a technique to take advantage of IoT technology using cloud sources for early diagnosis. One of the biggest challenges in medical fields is maintaining data. Clouds have huge volumes of storage capacity, so combining IoT technology with the cloud can have a huge impact. The study [18] designed an energy-efficient protocol for healthcare applications using dynamic channel coding by combining physical and multiple access layers. The aim is to optimize energy usage and maintain a lifetime of nodes in a network.
During the COVID-19 pandemic, the health of front-end medical staff became vulnerable to this disease, and extra care was required to deal with patients. Consequently, remote health monitoring applications helped medical staff to diagnose patients effectively. Such systems are realized using IoT devices [19,20]. In addition, image processing applications are installed at different public points for real-time surveillance of the public. The data from such installments can be obtained via IoMT by medical staff for further analysis [21].
Due to the advantages offered by IoMT technology, several approaches and systems have been presented recently using IoT connectors (Table 1). Jain et al. [22] proposed a health-monitoring system based on near-infrared spectroscopy and ML to analyze the glucose level. Today, IoMT edge devices are frequently used to monitor human health. It is a positive trend to help humans achieve fast-tracking and accurate results. The error ratio of the proposed glucose level detector is minimized as compared to methods available in the literature. Shui-Hua Wang et al. [23] proposed a method to classify diseases such as COVID-19, pulmonary diseases, tuberculosis, and pneumonia. The suggested method helps medical staff to diagnose diseases more accurately. This method shows better results in detecting various diseases. The authors designed an approach for remote-controlled ambulance service to improve healthcare in [24]. AI approaches are used to obtain real-time results. AI approaches are helpful in real-time applications, especially in medical applications where a fast and accurate response is needed. Similarly, Harshal Arbat et al. [25] used an approach to monitor heart patients by managing a smart health band. The band is used to measure the heart rate data, which are used in analyzing the patient’s health.
Table 1. IOT connectors and their applications.
| Ref. | Standards | Purpose |
|---|---|---|
| [26] | Bluetooth | Monitoring and detection of heart sounds |
| [27] | Bluetooth | Fitness health monitoring when engaging in physical activity |
| [28] | Wifi | Remote monitoring of health condition |
| [29] | 3G | Provide fast data communication during emergencies |
| [30] | Satellites | Track patients and approach them in an emergency |
| [31] | NFC | Home surveillance for less-skilled individuals |
| [32] | 2G, 3G | Skincare in real-time |
| [33] | RFID | Locating and tracking medical equipment quickly |
Smart devices are used today to gather health-related data to analyze human health. The study [34] designed an approach to monitor human health using a mobile application. Mobile applications can be merged with IoT and cloud computing to boost the limits of the health monitoring system. Cloud technology has enough storage to store large volumes of IoT data and provide processing services. This approach also covers the brain signal to measure the stress of the human body. Michael Fischer et al. [35] proposed a technique for non-professionals to know about various diseases. Instructions are given to the bot, which help to diagnose the patient. The integration of the bot with smart devices helps in providing better services. The accuracy of the technique is low compared to other techniques, yet it is a remarkable step toward automated disease diagnosis. The complete summary of IoT-based works is shown in Table 2.
Table 2. Summary of IoT-based works.
| Ref. | Year | IoMT |
|---|---|---|
| [34] | 2016 | Cloud-based remote ECG system |
| [36] | 2016 | Secure healthcare system for patient |
| [37] | 2016 | Intelligent healthcare system for patients in medicine |
| [38] | 2016 | IoT-based kidney anomaly recognition system |
| [39] | 2016 | IoT enabled healthcare |
| [40] | 2016 | IoT-based mobile medical health care system |
Along the same lines, Reference [41] developed a system consisting of cloud, IoT sensors, and IoMT devices to deal with cardiac patients. The system measures the patient’s eye movement, body temperature, and oxygen level for heart disease. Similarly, authors designed a model to monitor heart disease in [42]. The proposed approach utilizes sensor data and an ensemble model in a fog environment. The model performs early diagnosis of heart patients. In [43], an Adaptive Neuro-Fuzzy inference system with multiple kernel learning is used to identify cardiac disease. The system provides better results, although the computational complexity is high. An automated approach is designed in [44] to distinguish between people at high risk of heart failure and those at low risk. The authors used the classification and regression tree (CART) and achieved specificity and sensitivity values of 63% and 93%, respectively. The existing state-of-the-art studies on cardiac disease is shown in Table 3.
Table 3. Existing studies on cardiac disease.
| Ref. | Methods | Data | Limitations |
|---|---|---|---|
| [43] | Neuro-Fuzzy Inference and Multi-Kernel Learning | KEGG Metabolic Dataset | Evaluating parameters are not comparable. |
| [45] | CNN, BiGRU, BiLSTM Ensemble | Cardiac Disease Dataset | The dataset used in this approach is not a standard dataset |
| [46] | Gradient Boosting, XG Boost, Random Forest Ensemble | Hungarian, Cleveland, Z-Alizadeh Sani Dataset | The complexity and price of the model are increased by stacking three models. |
| [47] | Generalized Discriminant Analysis and Fisher Method | SR-CAD and NSR-CAD | There is lack of training on large datasets. Training on heart rate variability needs to improve. |
| [48] | Random Forest, KNN, Decision Tree, Naïve Bayes | Hungary, Cleveland, VA Long Beach, Switzerland Datasets | There should be more experimentation with model combinations and feature choices. |
This entry is adapted from the peer-reviewed paper 10.3390/s23104580
This entry is offline, you can click here to edit this entry!
