INTRODUCTION

Heart disease is the leading cause of death among men and women worldwide, regardless of the racial or ethnic background 12. Although heartdiseaseis a global problem,a study conducted by the National Center for Chronic Disease Prevention and Health Promotion (NCCDPHP) found that,in the United States, one person dies from cardiovascular disease every 36 seconds,and this amountsto 659,000 deaths annually. In addition to being responsible for a quarter of deaths in the U.S., heart diseases cost the public coffers about $363 billion dollars between 2016 and 2017, including the costs of health services, medicines, and lost productivity due to death 2.

Considering the large number of people who suffer from heart disease, coupled with the fact that one person suffers from a heart attack every 40 seconds, it is essentially important to monitor people’s heart and heart attacks, especially among people with cardiovascular problems or who have exhibited heart attack warning signs or symptoms. Heart attack can occur more than once; in many cases, the attack occurs silently, causing severe damage without the person noticing it. Fortunately, there are some warning signs before a heart attack occurs 1.

One wayof monitoring heart attack signs in people is to use constant health measurement sensors. Technological innovations aim to improve people's quality of life in order to contribute to the formation of healthy, intelligent and happy individuals.

Smart health systems involve the use of smart technological devices and systems for treating, monitoring and improving people’s health.With the ever-increasing advances in technology, the use of smart health systems/devices has become increasingly popular and essentially important,since these systems/devices help to expedite and improve access to healthcare and the quality of urban services and infrastructure (social sustainability). In addition, through the internet of things (IoT), internet has become ubiquitous in people's lives. IoT works in different environments in an increasingly sophisticated and intense way, with connections and interactivities involving the following: people with people (conventional internet), equipment with equipment (machinery), people withsmart objects - such as mobile phones and personal assistants, etc.34. By combining the basic needs of individuals with technology, people can have their health monitored in a broad effective way. The monitoring of people’s health can occur in the following different ways: (i) face-to-face monitoring by a health professional in routine consultations; (ii) the use of simplemedical equipment/devices at home; (iii) remote monitoring with specialized support in any environment; and (iv) remote monitoring - specifically indoor, likein vehicles.

Taking the above considerations into account, this paper presents a comprehensive review of the literature on health monitoring in vehicles, and the systems and devices used for this purpose. Most people around the world rely on vehicles as a means of transport to go to work and school, run errands, travel and visit places, and so forth. Nowadays, people spend a considerable amount of their time daily in self-owned/private vehicles and public transport vehicles to go to work/school or perform other daily activities.

Vehicles can thus be used as a possible tool for monitoring people’s health, since people are potentially at risk of being exposed to different types of emotion and state of stress, as they spend a considerable amount of their time in these transport systems. Exposure to different types of emotions and state of stress are among the major factors that aggravate the occurrence of heart attacks, and this can endangerthe life of the driver and the occupants of the vehicle, as well as the lives of third parties. Due to climate change and the need to reduce our over-reliance on petroleum and other fossil fuels globally, electric vehicles will soon be the main means of mobility in urban cities and between cities 5.

As a complex and interdisciplinary subject, a comprehensive review on smart health monitoringrequires the combined use ofstudies and data from different sub-areas, and it is essentially crucial to identify themes with significant scientific production and possible research lines for the development of the study. When faced with a large number of publications to rely on, the main concern is the appropriate selection of works to be used in a specific research project.

In the presentstudy, the articles were initially selected through thorough searches in the databases available, taking the research areas into account. The abstracts, keywords and keyword combinations of the selected articles were thoroughly verified, and the main works chosen for the reviewwereread meticulously. Based on the objectives of the study, we then classified the articles, the type of publication, whether it was complete or not, and determined the works that were most relevant. In the present work, the following factors were considered to elaborate a comprehensive review on the topic: adequate selection of articles; scientific value of articles; and quality of articles 6.

This method of analysis was mainly based on the Proknow-C bibliometric review technique 78. The concept mapping technique was then used as a tool to support the research in the phase of bibliographic study and analysis of the connections between the main works found 910.

The analytical techniqueemployedin this study was aimed atevaluating and developing a comprehensive review of themost relevant publicationswhich could contribute to the search for specialized bibliographic references on the subject.

The bibliographic analysis performed in specialized scientific publications or studies under construction is presented through a time interval from 2011 to 2021. The research methodology adopted in this work involved the following steps:

select a bibliographic portfolio that is scientificallyrecognized and aligned with the theme/subject;

analyze the journals and articles of the bibliographic portfolio surveyed;

present the most relevant works reported on the subject and analyze new works that are yet to be fully developed.

show the connections between these works through the technique of conceptual maps.

With the massive advancesmade in IoT-based technologies and their implementation, there has been a parallel interest in monitoring the optimization of these technologies inthe transport system, as well as in smart vehicles (IoV - internet of vehicles). Thus, in the quest for optimizing these IoT-based technologies in transport systems, health monitoring of intelligent vehicle occupants (smart health in vehicles) is seen tobe essentially important.

INTERNET OF THINGS AND HEALTH MONITORING

According toRay 11, although IoT is interconnected with nano electronic devices, connections, sensors, smartphones, embedded systems, cloud computing and software, it is still searching for its own form of connection, and thishas already resulted in remarkable advances such as the use of universal solution media for the connected scenario. In the case of health monitoring, IoT reflects a set of elements with their respective sensors connected to the internet 12. In fact, the impacts of the application ofIoT-based technologies can be clearly observed in all areas of society. IoT-based technologieshave been employedfor the monitoring of a wide range of things; these technologies have been used for the monitoring of the human body andthe surveillance of an entire country. The use of IoT-based technologies for monitoring people and things of different scope and in different environments can lead tothe development of a chain of sustainable and intelligent systems. This is important because it is the connections between these systems that produce true intelligent systems which are intended to be usedfor the improvement of people's lives. Oneof the major advantages of usingIoT-based technologies for the monitoring of human body (smart health monitoring of human body) is thatthe vital signs of the individual can be effectively monitored from afar, and this can be useful for patients’ treatment and disease prevention, as well as for the integration of information on patients and health professionals13. In the transport system,healthmonitoringhas not yet beenproperly conceived or regulated;as such, it is a subject that needs to be thoroughlyinvestigated and explored.

SMART TRANSPORT AND SMART VEHICLES

Intelligent transport systems, also referred to as smart transport, smart transportation and mobility, intelligent transportation systems or smart vehicle systems, can be smart roads or smart highways with warning messages and signs for diversion according to weather conditions and unexpected events such as accidents or traffic jams 5. Intelligent transport systems will become part and parcel of smart citiesin the near future; this may be associatedwith the nature of the contents involved in the implementation and development of loT-based applications 14.

The concept of smart vehicle emerged as a viable solution to urgent road-related problems including traffic management, driver comfort and health, road safety and on-demand provisioning services. With the availability of on-board vehicle services, smart vehicles will play a key role infacilitating the construction of smart cities. Smart vehicles can share and store digital content, detect and monitor their surroundings, and mobilize services on demand. But providing these services is challenging due to different needs, costs and quantity of services demanded. The concept of Smart Vehicle as a Service (using the acronym SVaaS) is related to the idea of providing continuous vehicular services to those who need them14. In Brazil, the rate of urbanization stood at approximately 84.4% in 2018, which is the highest rate recorded since the 1960s. This dramatic rise in urbanization has led to the construction of new houses and buildings, with sharp increase in the fleet of vehicles, new industries, and so forth,whichhas caused drastic changes in the environment 15. Certainly, this is one of the major reasonswhy we needsmart vehicles in the cities; vehicles that can help improve our health and the environment,while promotingthe sustainability of our planet.

When a function/feature is added to a vehicle, the vehicle needs to possess suitable properties in order to meet the new demand or capacity. Bearing that in mind, the automotive industry is developing systems that will enable connected vehicles to monitor driver behavior (driving monitoring system - DMS). The application of thesesystems helps to monitor the behavior of drivers and prevent the occurrence of accidents on the road.Researchhas shown that,in most cases,the drivers’ emotional distraction andloss of focus while driving are caused by fatigue, drowsiness and alcohol, or even changes in the environment - these factors can cause road accident16. In the case of smart vehicles, apart from drivers, passengers can also be monitored while traveling. However, due to data privacy laws, passengers may have to grant their permissionfor the monitoring of their health during the ride, and accept to pay for additional health care services. This decision can be made based on the same principle adopted for recommended services in e-health and e-tourism scenarios, which have become increasingly important 17.

Consumers of low-value vehicles sometimes look for "premium-like" versions to satisfy their vanity and/or personal needs. To remove the stigma attached to cheaper and simpler cars, carmanufacturers are working on the aesthetic appeal of these vehicles in order to raise the self-esteem of the owners 18. In future studies, due to falling prices of sensors and controllers, it may be possible to think of popular, intelligent vehicles (including health monitoring services) that add value to owners at the time of purchase and offer health benefits to occupants.

HEALTH MONITORING OF SMART VEHICLE OCCUPANTS

There are devices or equipment that can detect and monitor the health status of the driver before and during thejourney. However, cables should not be connected to the driver as part of the health monitoring system,as this may cause distraction, leading to loss of focus. One of the smart health monitoring tools that can help tackle this problem is stretchable (elastic) conductive ink (SCI), which has been used on fabric and through electronic devices for monitoring purposes in diverse areas including health, automotive, and communications16. SCI operateslike an adhesive tapeand its flexibility allows the transmission of signals to a circuit, even on an irregular or constantly changing surface. Over the past few years, advances in the design of health systems in vehicles have becomea subject of discussion in the literature. Current problems faced by health care providers, as well as potential challenges in the field of health monitoring are identified and compared to monitoring systems in other areas. In view of this, the aim of this study is to investigate the application of advanced technology to monitor the health of smart vehicle occupants.The study also aimed to explore the systemic interaction of advanced technologies with cities in a sustainable and smart way.The contribution of the present study to scientific knowledge lies mainly in providing helpful insights into the expansion of the benefits of advanced technologies in our modern societies.

MATERIALS AND METHODS

The presentstudy is characterized as exploratory and descriptive in nature. As an exploratory work, this study does not require the elaboration of hypotheses; it is confined to the definition of objectives and the search for more information on a given subject of study which is conducted through the gathering of a bibliographic portfolio of articles on the research topic.The study is descriptive in nature, as it describes the characteristics of thearticles and their references through the discovery of the frequency with which each article is cited, and their relationship and connection with other articles19. As for the research approach, it is characterized as qualitative. To conduct the present review, we conducted qualitative research in order to identify the articles that constituted the bibliographic portfolio and their references, and quantitative research which involved counting the variables investigated in order to determine the volume of articles in each sub-area; this was done in line with the research techniques described by Richardson 20.

One of the methodological instruments used in this study involved the construction of knowledge based on the interests and delimitations of the researcher, which is in line with the constructivist analytical approach Proknow-C (Knowledge Development Process-Constructivist) proposed by Ensslin et al.21. Based on the work of Ensslin et al.22, and with the aim of developing the bibliographic portfolio of the subject of this review, the research process was divided into three main phases. The first two phases (A and B) were based on the Proknow-C constructivist analysis,while the last phase(C) was proposed by the authors of the present review9, 10:

Selection of the gross articles bank (gross portfolio or GP);

Filtering of the database of articles; and

Elaboration of the conceptual map and analysis.

The first phase, i.e., selection (A), was subdivided into the following four stages:

definition of databases and/or search systems;

construction of search protocol with the definitionof keywords and the search string;

adherence test;

search for articles in the databases with the keywords found in the search string.

The second phase, related to filtering (B), was subdivided into the following six stages:

elimination of relevant, repeated articles, with inconsistent data or dubious sources, and with publication dates outside the delimited horizon;

alignment by reading and analyzing titles (preliminary selection);

analysis of titles and their connection with the main topic(thematic selection);

analysis of abstracts and their connection with the main topic (selection of abstracts);

analysis via synthetic reading,taking the main topic of the review into consideration (selection of synthetic reading);

analysis via detailed reading and classification, taking the main topic of the review into consideration (full reading selection).

In the third phase (C), we created conceptual maps in order to obtain a complementary graphic analysis of the articles and a schematic representation of the authors' ideas, providing contextual and conceptual visualization, with a view to facilitatingthe understanding or visualization of the most important points of the content and the possible gaps in research. All thephasesmentioned above are illustrated in Figure 1 in the form of a funnel. The conceptual maps were elaborated using the CmapTools tool¹. These maps enabled us to visually identify connections between the differentarticlesfoundand to effectively analyzethe research gaps. As already seen, the use of conceptual maps, in support of the bibliographic survey stage, has advantages for the organization of the information of the main works found, as exposed by Moreira 10, that is: a) it emphasizes the structure of the works presented; b) emphasizes the importance of the concepts presented in hierarchical order; c) promotes an integrated view of the theme and also d) facilitates the understanding of the theme addressed. In addition, we developed an initial research protocol in order to direct the searches and provide support for the methodology adopted. With regard to the treatment of the articles, the following software tools were employed: i) Mendeley was used for the organization, collection and exclusion of articles in the GP; ii) JabRef was used for the transfer of articles to spreadsheets; and Microsoft Office Excel 2016 was used for the preparation of spreadsheets and graphs^2,3.

Figure 1:
Funnel illustrating the search and selection process.

FINDINGS

Thesearch using the search engines of the Capes Journal Portal⁴(Brazilian internet portal which contains articles and scientific publications from national and international academic journals administered by Capes - Brazilian public research funding agency) and Google Scholaryielded 1337 works. The steps were divided according to the topics belowand the methodology employed in the study. In phase 1, databases and/or search engines were selected. We opted for direct search in Google Scholar and in the Capes Journal Portal, in line with the search protocol. In the Capes Journal Portal, the databases were filtered via the search in the following areas:i) Applied Social Sciences with Information Sciences as sub-area; ii) Exact and Land Sciences with Computer Sciences as sub-area; iii) Multidisciplinary Fieldwith Interdisciplinary Subjects as sub-area;iv) Health Sciences; and v) Engineering-Technology-Management. Finally, the common databases related to the sub-areas were used for the composition of the gross articles bank (GP- Gross Portfolio), and the 12 major databases (based on the view of the authors of the present study) were selected; the databases included the following: i) Academic Search Premier - ASP (EBSCO);ii) Britannica Academic

Edition; iii) EMBASE (Elsevier); iv) Directory of Open Access Journals - DOAJ; v) OECD - iLibrary; vi) Bioline International; vii) SciELO.ORG; viii) ScienceDirect (Elsevier); ix) SCOPUS (Elsevier); x) Technology Collection (ProQuest); xi) Web of Science - Main Collection (Clarivate Analytics); and xii) Wiley Online Library.

As mentionedpreviously, for the filtering of the selected articles, we conducted a search for articles in the Capes Journal Portal and Google Scholar search engines, with the general search through the string search generated from the basic protocol of the review (Figure 1, item 2 - Representativeness test of the Bibliographic Portfolio). In the third stage of the search and selection process (Figure 1, item 3), the search string test was performed directly in order to obtain the GP. To perform this analysis, three articles were randomly chosen in order to identify the need to include new keywords. Based on the procedures executed, we concluded that it would not be necessary to include new keywords, because the keywords used were found in 2 of the 3 articles chosen randomly, and in the 100 most relevant articles. This outcome pointed to the alignment of the articles with the research topic. In the fourth stage, for the selection of the GP and using the search protocol, the search axis was defined with the following terms: "smart health"AND "driver" OR "occupant"; "monitoring" OR "data gathering";"IoT" OR "internet of things"; "smart" AND “vehicle” or “car”. Based on that, weobtained five combinations of words, with the cross-checking of the axis used as a guide in the search for scientific articles. Figure 2 shows the combinations of keywords with the words of the search axis.

Figure 2:
Search axis layout and search keywords.

The search process began after the selection of the databases. The search was conducted using the combinations defined by the keywords with the research axis; thecombinations were made in Englishand covered the periodbetween 2011 and 2021. In this initial selection, 107 articles were found in the Capes Journal Portal and 1230 articles were found in Google Scholar; this amounted to a total of 1337 articles - which was used to compose the GP.

In the next step (Figure 1, item 5),based on the total number of articles collected, the first filtering procedure was performed in the GP.In this filtering phase, repeated or redundant articles were removed from the GP, prioritizing the articles by relevance up to the top 100 of the results obtained from each search engine (Capes Journal Portal and Google Scholar). For an article to be selected, thetitle of the articlewas required to be related to smart health or smart car/vehicle/transport or internet of things. Also, with the aid of the Mendeley managing tool for updating data/information (this tool was used in order to search for relevant information about the sources of the articles on the internet), articles with inconsistent data were removed from the GP. Thus, only articles published between 2011 and 2021 following this filtering procedure, the number of articles was reduced to 43%, leaving only 574 articles (BASIC SELECTION).

Subsequently, new filtrations were performed. In the sixth stage, the second filter was applied; this filtering procedure involved reading all the titles of the articles generated in the first filtering procedure andremoving those that were not aligned with any of the topics proposed. After reading through the 574 titles, only 346 articles remained (26% of the GP - PRELIMINARY SELECTION or ANALYSIS OF TITLES).

In the seventh stage (Figure 1, item 7), the third filter was applied; this filtering procedureinvolved reading all the titles of the 346 articles andremoving those articles that were not aligned with the topic proposed (smart health, smart car/vehicle/transport), directly or indirectly. After this filtering procedure,only 285 articles remained (21% of the GP - TOPIC SELECTION).

In the next stage (Figure 1, item 8), the fourth filter was applied; this filtering procedureinvolved reading all the abstracts of the 285 articles andremoving those articles that were not aligned with the topic proposed. After this filtering procedure, only 94 articles remained (7% of the GP - ANALYSIS OF ABSTRACTS ON THE TOPIC).

In the ninth stage, the fifth filter was applied; this filtering procedureinvolved contextual/synthetic reading of the 94 articles and removing those articles that were not aligned with the topicproposed. After thisfiltering procedure, only 23 articles remained (5% of the GP - SYNTHETIC READING).

In the tenth stage, the sixth filter was applied; this filtering procedureinvolved a thorough reading of all the 23 articles that remained in the previous stage andremoving those articles that were not aligned with the topic proposed. The articles were analyzed meticulously in order to obtain relevant information on the following elements: i) Introduction and justification of the study; ii) Keywords; iii)The study proposal and the research questions it seeks toanswer (Research problem); iv)Theresearch methodology employed; v)The results and conclusions drawn from the study,as well as suggestions for future work (future perspectives); vi) Thekind of article it is/which category it falls into; and vii) Personal observations, with the exploration of the content of the articles and the analysis of the most relevant articles that deserve to be included in the literature review. The filtering procedure performed in this stageresulted in the exclusion of 7 articles and only 16 articlesremained (approximately 1% of the GP - DETAILED READING), as can be seen in link 1⁵ (Table of 16 articles).

In the eleventh stage (Figure 1, item 11),aftera detailed analysis of the 16 articles that remained in the last stage, the conceptual maps of each article were created and interconnected by the topics, with the intersection of the studies. In essence, we were able to graphically analyze the contextual level of the interconnections between the articles, research methodologies and future perspectives, as can be seen in link 1 too. We sought to findresearch gaps and bibliographic references that were more directly connected to the topic under investigation and the focus of the research (review).

Based on the analysis of the results obtained,we noted the following: i) the Proknow-C methodology7, with the appropriate complementations, was found to be systematic and aided the structuring of the literature review, and the complementations were constructivist in nature; ii) the research tools (Mendeley, Jabref, Cmaptools and Excel) employedin the study were consistentwith the proposed methodology - the tools applied enabled us to develop an analytical framework (work model) which can be reused by other researchers; iii) the conceptual map proposed helpedschematize the ideas of the main authors of the articles that were obtained afterthe filtering procedures, providing a contextualization of the articles, their interconnections and complementary works, and allowing us to perform an effective analysis of research gaps and future perspectives of the studies.

DISCUSSION

The assessment of an individual's health can be done through continuous (uninterrupted) or non-continuous monitoring, invasively or noninvasively, by obstructive methods with minimal or little obstruction, or via unobstructive or discreet methods23-26. Vehicles are transport instruments used for various services, but mainly for the mobility/transport of people and goods. Discreet health monitoring in vehicles is an innovative monitoring technique34 thatseeks to use sensors to monitor one’s health without presenting any inconveniences to the normal activity of the vehicle occupants. Thus, the space in which monitoring can be performed is the interior of the vehicle, and it is possible to monitor the health of all vehicle occupants 4. Several studies have reported the use of discreet health monitoring in vehicles and the benefits that arise from that342324252627. Based on the studies reported in the literature, the literature reviews show the monitoring devices that are currently in the market and what innovations and advances can be made in the area of smart health monitoring so that vehicle occupants can receive rapid treatmentwhen they need it, apart from having their health monitored during the journey 3. When it comes to health monitoring/control in vehicles, the vehicle seats can be equipped withsmart devices that can perform electrocardiography, balistocardiography and real-time cardiography with vital sensors 28. Certainly, a database system will be needed in order to store the simultaneous results of the tests 29.

In order to yield a smart system that protects and monitors the behavior of drivers 30, innovative advances need to be made in intelligent systems and computing 31; these include the development and application of multimodal biometric systems for secure user identification based on deep learning and the enhancement of the machine learning-assisted advanced driver assistance system (ADAS). Multimodal biometric systems merge two or more biometric features,such as face, fingerprint, palm and iris. Thus, the input of a single sensor can be performed through the application of different types of algorithms, or more than one sensor could be used to capture single or multiple biometric strokes. Thus, the use of highly advanced multimodal systems can significantly improve image recognition rates 31.

Smart tools and equipment can be used for this type of monitoring. Smart health monitoring systems (SHMS) - derived from health monitoring systems (HMS), are generic systems used for health monitoring. The design and modeling of these devices have been an object of research in the area for the past few years32. Another interesting monitoring tool is the seven-tier model architecture for vehicle internet;as pointed out in the literature, this tool can only be operated whenthe devices (sensors, personal devices, actuators, among others) are able to communicate with other equipment and machinesthrough different technologies 33. Other forms of communications which can be integrated in vehicles are V2X (vehicle-to-everything), I2X (infrastructure-to-everything) and P2X (pedestrian-to-everything) (which allowcommunication between vehicles, infrastructure and pedestrians) and their applications 34. Some of the systems employedin driver monitoring include the following: driver monitoring systems (DMS) - also referred to as driving monitoring systems 162526272928, ADAS, and advanced driver monitoring for assistance systems (ADMAS)35. Driver monitoring is done in different ways and using different techniques; the techniques employed fall into the following categories:vehicle data monitoring (VDM); physiological signal monitoring (PSM); and monitoring of actions in driving (action monitoring - AM). The PSM systems are divided into intrusive (contact method) and discreet (contactless method). In intrusive PSM systems, vital characteristics such as breathing, electroencephalogram (EEG), electrocardiogram (ECG), digital signal processing (DSP), are recorded. Driver monitoring focuses on driver surveillance monitoring, attention and stress in general36.

In-vehicle health monitoring includes environmental, physiological and behavioral monitoring. Environmental monitoring involves monitoringthe temperature in the vehicle, air quality, humidity, weather conditions, and speed; these data are captured by default in order to ensurethe welfare of the vehicle occupant and provide the driver with the needed assistance through driver assistance systems. Physiological monitoring typically involves monitoringthe vital signs of the vehicle occupant, including heart rate (HR), breathing rate (BR), body surface temperature, and skin impedance. More advanced parameters can be measured with special sensing devices. Behavioral monitoring involves quantifying physical activities that occur during the trip and which reflect the driver's attention, tiredness and well-being. Wang et al. 4conducted an interesting study on health-supervised vehicle technology, where they investigatedthe use of discreet/non-obstructive monitoring in vehicles. As noted by the authors, in-vehicle health monitoring allows the occupant of the vehicle to perform regular medical evaluations/check-ups during the driving time; this means that a number of specific sensors need to be installed in the vehicle for the monitoring to occur effectively. Haux et al. 37 show that in sensor-enhanced private spaces (for example, inside vehicles), health-related information is continuously obtained and critical changes or events can be captured automatically. The information gathered in these sensor-enhanced spaces reflects natural reality and, from this, a wide range of health services can be provided, such as emergency detection, disease management and real-time health feedback; preventive medical advice or diagnostics can also be provided.

Among the sensors that can be used for health monitoring in vehicles include the following: mechanical (e.g. ballistocardiography sensors, microphones, accelerometers), optical (e.g. video and infrared cameras, pulse oximeter), electromagnetic (e.g. thermometers, radar sensor, contact electrodes), and others (such as GPS, gas sensors) 4 . Discreet monitoring of vital signs in automotive environments can be performed via the application of the following: ECG sensors, capacitive ECG (cECG), radar, ballistocardiography (BCG), seismocardiography (SCG); video imaging; photoplethysmography (PPG); photoplethysmography imaging (PPGI); magnetic induction (MI); body surface thermography, displacements and temperature measurement; perfusion and intrathoracic impedance; etc.27.

The use of camera sensors allows both the direct monitoring of the activities of the driver 38 and the measurement of his/her vital signs. For illustration purposes, oxygen saturation can bemeasured by attaching LED and photodiode to the steering wheel 39and through the analysis of images captured by a camera attached to the windscreen or control panel 40. With regard to oxygen saturation monitoring, this monitoring mechanism is a way of detecting whether the vehicle occupantneeds medical care, like in cases where the condition of a COVID-19-infected person has worsened. Heart rate variation (HRV) can be measured by deploying a piezoelectric sensor or accelerometer on the 41, or through a radar signal 42.Many of these sensors allow the monitoring of all the vehicle occupants in real time. With regard to physiological monitoring, Naziyok, Zeleke and Röhrig43 studied contactless monitoring, where they highlighted the use of ballistocardiography (BCG), radar sensing and thermography as ways to obtain heartbeat (HR), HRV, and cardiopulmonary signals, respectively.

With the use of the conceptual map, we were able to identify the interconnections between the articles and their limitations. We foundthe existence of some gaps in research that need to be explored. Wang et al. 4 conducted a literature review on the application of sensing technology in health monitoring in vehicles, where they provided an interesting guide forthe use of sensors in monitoring the health and well-being of vehicle occupants; the guide was developed based on the following questions:

(1) Which sensors are suitable for data collection in the vehicle? (Sensor)

(2) Where should the sensors be placed? (Location)

(3) What biosignals or vital signs can be monitored in the vehicle? (Biosignal)

(4) For what purposes can health data be used? (Purpose)

Based on the interconnections observed between the columns sensor, location, biosignal and purpose and the references found in the research, Wang et al. 4 concluded that there is higher or lower density of studies in each of the 4 dimensions and their sub-categorizations, which are represented by the lines between them and the filling of the box 4. Looking at the dotted lines between the columns"biosignal" and "purpose" (Figure 3), one will see the interconnection of "heartbeat (HR)" and "electrocardiography (ECG)" with "heart diseases" as being a natural promising possibility of future work; this is because both "heartbeat" and "ECG" have several interconnections when it comes to the application of sensors in vehicles. Thus, Wang et al. 4 showed that, despite the plethora of research on the development and application of ECG detection and HR measurement sensors, no studieshave been reported in the literature regarding theuse of the data collected for solving clinical problems (diagnosis).

The present literature review systematically summarized the state of the art of the most current research on sensor technology applied for discreet health monitoring in vehicles. Apart from answering the research questions proposed, the review pointed out future perspectives and suggestions for future research based on the gaps in research that need to be explored in the area. Most of the studies related to the object of this review found in our bibliographic research were focused on the development of sensor technology and its application targeted at ensuring safety driving. There is a scarcity of health-oriented research related to the use of physiological data obtained from sensors applied in vehicles for diagnosis and other medical purposes 4.

Figure 3:
In-vehicle health monitoring: Interconnections between Location, Sensors, Biosignal and Purpose columns, based on the references found.
Source:WANG et al., 2020.

Wang et al. identified only 11 articles investigating the application of sensors for health monitoring in vehicles;remarkably, 9 of these articles were focused on safety driving and only 2 articles focused their attention on health monitoring. Surprisingly, none of the articles focused their attention on the use of physiological data obtained from sensors installed in vehicles for the diagnosis of diseases; however, warning that cardiovascular diseases are the ones that have the most devices and studies in their measurement4.

The results showed that most of the articles were published in the last 4 years; this shows that the research topic is in its incipient stage and more studies are gradually being conducted on the topic. Previous studies (published more than 4 years ago) reported in the literature have shed some light on the subject and have proposed the development of specific sensors and their potential application for health monitoring in vehicles.

In total, 16 papers were reviewed in detail in this article, considering the aspects relevant to the questions proposed for the construction of the concept map, including the following:i) introduction and justification for the study; ii)the study proposal and the research questions it seeks to answer (Research problem); iii) the research methodology employed; iv)the results and conclusions drawn from the study; and v) perspectives and suggestions for future work. The conceptual map was a practical and visual tool that enabled us tohave a clearer understanding of the context of the studies and their interconnections.

CONCLUSION

In this present work, we conducted a comprehensive review of studies reported in the literature - between 2011 and 2021, related to smart health monitoring in smart vehicles. The articlesanalyzed in this review were published in international scientific journals of high representativeness. A total of 1337 articles were analyzed. As seen in the results of this work, most of the articles were published in the last 4 years; demonstrating that more studies are needed on the subject and with new research approaches.The objectives that were initially proposed in this study were achieved. Applying the Proknow-C methodology 7in the search for literature on the research topic enabled us to obtain a portfolio of 16 important articles of scientific relevance aligned with the topic investigated.

A conceptual map was created using the 16 relevant articles, where the articles were interconnected; this map also helped identify gaps in research on the subject matter that needs to be deeply explored in future research.

Considering that the research topic is recent and it is most widely studied in the fields of engineering and computing, a good number of articles found in our literature search were confined to the area of technology. About the application of the proposed literature review methodology for the topicinvestigated, based on the results obtained and the level of relevance of the selected articles, it can be concluded that although the proposed methodology is of great academic and professional relevance,itshould be complemented with conceptual mapswhen a contextual analysis of interrelated articles on a particular topic is needed.

The findings of the review pointed to a predominance of studies related to safety drivingover health monitoring, though we were able to find a few studies that reported the adoption of technological innovations for health monitoring, which served as the theoretical basis for the present work.

As for suggestions for future research, we recommendimproving the stages of the filtering process by enhancingthe accuracy of the search criteria and comparing the literature review methodology with other methodologies through the analysis of the final list of articles obtained at the end of the filtering stages. The inclusion of other methods for the systematic analysis of articlesmay also help expand the researcher’s knowledge on the subject.

Received: 08/10/2022

Accepted: 29/11/2023

REFERENCES

1 C. F. Ski, K. M. King-Shier, and D. R. Thompson, “Gender, socioeconomic and ethnic/racial disparities in cardiovascular disease: A time for change,” Int. J. Cardiol., vol. 170, no. 3, pp. 255-257, Jan. 2014, doi: 10.1016/J.IJCARD.2013.10.082.

2 CDC, “Heart Disease and Stroke Statistics-2021 Update A Report from the American Heart Association,” Lippincott Williams and Wilkins, 2021. doi: 10.1161/CIR.0000000000000950.

3 S. Heuer, B. Chamadiya, A. Gharbi, C. Kunze, and M. Wagner, “Unobtrusive in-vehicle biosignal instrumentation for advanced driver assistance and active safety,” in Proceedings of 2010 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2010, 2010, pp. 252-256, doi: 10.1109/IECBES.2010.5742238.

4 J. Wang, J. M. Warnecke, M. Haghi, and T. M. Deserno, “Unobtrusive health monitoring in private spaces: The smart vehicle,” Sensors (Switzerland), vol. 20, no. 9, p. 2442, 2020, doi: 10.3390/s20092442.

5 N. Gershenfeld, R. Krikorian, and D. Cohen, “The Internet of Things,” Sci. Am., vol. 291, no. 4, pp. 76-81, 2004, doi: 10.1038/scientificamerican1004-76.

6 A. C. Braga et al., “Estudo bibliométrico para construção científica de modelo de inovação tecnológica em redes de empresas,” Cienc. da Inf., vol. 45, no. 1, pp. 74-87, Dec. 2016, doi: 10.18225/ci.inf..v45i1.

7 L. de Moraes, R. Garcia, L. Ensslin, M. J. da Conceição, and S. M. de Carvalho, “The multicriteria analysis for construction of benchmarkers to support the Clinical Engineering in the Healthcare Technology Management,” Eur. J. Oper. Res., vol. 200, no. 2, pp. 607-615, 2010, doi: 10.1016/j.ejor.2009.01.018.

8 Afonso M. H. F., J. V. de Souza, S. R. Ensslin, and L. Ensslin, “Como construir conhecimento sobre o tema de pesquisa? Aplicação do processo proknow-c na busca de literatura sobre avaliação do desenvolvimento sustentável,” Rev. Gestão Soc. e Ambient., vol. 5, no. 2, pp. 47-62, 2012, doi: 10.24857/rgsa.v5i2.424.

9 M. S. T. Araújo and R. Formenton, “Utilização De Mapa Conceitual Como Ferramenta De Análise De Trabalhos Científicos,” Holos, vol. 1, pp. 171-181, 2015, doi: 10.15628/holos.2015.2130.

10 M. A. Moreira, A teoria da aprendizagem significativa e sua emplementação em sala de aula, vol. 1. Editora Universidade de Brasilia, 2006.

11 P. P. Ray, “A survey on Internet of Things architectures,” J. King Saud Univ. - Comput. Inf. Sci., vol. 30, no. 3, pp. 291-319, Oct. 2018, doi: 10.1016/j.jksuci.2016.10.003.

12 E. Oriwoh and M. Conrad, “‘Things’ in the Internet of Things: Towards a Definition,” Int. J. Internet Things, vol. 4, no. 1, pp. 1-5, 2015.

13 F. Lacerda, “Arquitetura da Informação Pervasiva: projetos de ecossistemas de informação na Internet das Coisas,” 2016, Accessed: Jun. 16, 2017. Online. Available: http://repositorio.unb.br/handle/10482/19646.

14 M. Keertikumar, M. Shubham, and R. M. Banakar, “Evolution of IoT in smart vehicles: An overview,” in International Conference on Green Computing and Internet of Things (ICGCIoT), Oct. 2015, pp. 804-809, doi: 10.1109/ICGCIoT.2015.7380573.

15 E. M. et al. Nieri, “Vista de Condiciones de los árboles urbanos: un estudio de revisión,” Rev. Cienc. y Tecnol., vol. Ano 20, no. 30, pp. 56-61, 2018, Accessed: Mar. 07, 2022. Online. Available: https://www.fceqyn.unam.edu.ar/recyt/index.php/recyt/article/view/252/179.

16 A. K. Ab Wahid et al., “Driving monitoring system application with stretchable conductive inks: A review,” Malaca - Malasia, 2020.

17 Y. Pérez Almaguer, N. Martín Dueñas, E. Carballo Cruz, and R. Yera, “Una revisión de los sistemas recomendadores grupales como herramienta innovadora en el área del turismo,” Rev. Cienc. y Tecnol., no. 35, pp. 44-53, 2021, doi: 10.36995/j.recyt.2021.35.006.

18 A. Demaison, J. G. Botura, and L. C. Paschoarelli, “A função percebida no desenho de carros ‘populares’ e sua importância para o mercado,” Rev. Cienc. y Tecnol., vol. 35, no. 1, pp. 35-43, Aug. 2021, doi: 10.36995/J.RECYT.2021.35.005.

19 A. C. GIL, Métodos e Técnicas de Pesquisa Social, 7th ed. São Paulo: Grupo GEN, 2019.

20 R. J. Richardson, Pesquisa Social: Métodos e Técnicas, 4th ed. São Paulo: ATLAS, 2017.

21 L. Ensslin and S. R. Ensslin, “Processo de Investigação e Análise Bibliométrica: Avaliação da Qualidade dos Serviços Bancários Research Process and Bibliometric Analysis: Bank Service Quality Assessment,” vol. 17, no. 4, pp. 325-349, 2013.

22 L. Ensslin, G. M. Neto, and S. M. D. Noronha, Apoio à decisão: metodologias para estruturação de problemas e avaliação multicritério de alternativas. Florianópolis, SC.: Insular, 2001.

23 K. C. N. Silva, “Monitoramento da saúde humana através de sensores: análise de incertezas contextuais através da teoria da evidência de dempster-shafer,” Escola Politécnica da Universidade de São Paulo, 2013.

24 T. M. G. de A. Barbosa, “Uma arquitetura de redes de sensores do corpo humano,” UNIVERSIDADE DE BRASÍLIA FACULDADE DE TECNOLOGIA, 2008.

25 Z. Hou, J. Xiang, Y. Dong, X. Xue, H. Xiong, and B. Yang, “Capturing electrocardiogram signals from chairs by multiple capacitively coupled unipolar electrodes,” Sensors (Switzerland), vol. 18, no. 9, p. 2835, 2018, doi: 10.3390/s18092835.

26 R. C. Bessa Junior and B. C. C. de Leão, “Monitorização do débito cardíaco: vantagens e desvantagens dos métodos disponíveis,” RMMG - Revista Médica de Minas Gerais, 2010.

27 S. Leonhardt, L. Leicht, and D. Teichmann, “Unobtrusive vital sign monitoring in automotive environments-A review,” Sensors (Switzerland), vol. 18, no. 9, Sep. 2018, doi: 10.3390/s18093080.

28 M. Sidikova et al., “Vital sign monitoring in car seats based on electrocardiography, ballistocardiography and seismocardiography: A review,” Sensors (Switzerland), vol. 20, no. 19, pp. 1-29, 2020, doi: 10.3390/s20195699.

29 N. El Haouij, J. M. Poggi, S. Sevestre-Ghalila, R. Ghozi, and M. Jadane, “AffectiveROAD system and database to assess driver’s attention,” in Proceedings of the ACM Symposium on Applied Computing, Apr. 2018, pp. 800-803, doi: 10.1145/3167132.3167395.

30 Z. Wu et al., “Access to Smartness: An Intelligent Behaviour Monitoring and Safeguard System for Drivers,” in Advances in Intelligent Systems and Computing, 2021, vol. 1183, pp. 536-544, doi: 10.1007/978-981-15-5856-6_53.

31 X. Yang, R. S. Sherratt, D. Nilanjan, and J. Amit, Advances in Intelligent Systems and Computing 1183 Proceedings of Fifth International Congress on Information and Communication Technology, vol. 2. 2020.

32 M. M. Baig, H. Gholamhosseini, M. Mansoor Baig, H. Gholamhosseini, M. M. Baig, and H. Gholamhosseini, “Smart health monitoring systems: An overview of design and modeling,” J. Med. Syst., vol. 37, no. 2, Apr. 2013, doi: 10.1007/s10916-012-9898-z.

33 J. Contreras-Castillo, S. Zeadally, J. A. Guerrero Ibáñez, and J. A. G. Ibáñez, “A seven-layered model architecture for Internet of Vehicles,” J. Inf. Telecommun., vol. 1, no. 1, pp. 4-22, 2017, doi: 10.1080/24751839.2017.1295601.

34 J. M. L. Domingues, T. J. M. Sanguino, J. M. L. Domínguez, and T. J. Mateo Sanguino, “Review on V2X, I2X, and P2X Communications and Their Applications: A Comprehensive Analysis over Time,” Sensors (Switzerland), vol. 19, no. 12, pp. 1-29, 2019, doi: 10.3390/s19122756.

35 A. Koesdwiady, R. Soua, F. Karray, and M. S. Kamel, “Recent Trends in Driver Safety Monitoring Systems: State of the Art and Challenges,” IEEE Trans. Veh. Technol., vol. 66, no. 6, pp. 4550-4563, Jun. 2017, doi: 10.1109/TVT.2016.2631604.

36 E. Husni and G. Boy, “Car driver attitude monitoring system using fuzzy logic with the internet of things,” ICIC Express Lett., vol. 12, no. 11, pp. 1115-1122, Nov. 2018, doi: 10.24507/icicel.12.11.1115.

37 R. Haux, S. Koch, N. H. Lovell, M. Marschollek, N. Nakashima, and K. H. Wolf, “Health-Enabling and Ambient Assistive Technologies: Past, Present, Future,” Yearb. Med. Inform., pp. S76-S91, 2016, doi: 10.15265/IYS-2016-s008.

38 R. A. Naqvi, M. Arsalan, G. Batchuluun, H. S. Yoon, and K. R. Park, “Deep Learning-Based Gaze Detection System for Automobile Drivers Using a NIR Camera Sensor,” Sensors 2018, Vol. 18, Page 456, vol. 18, no. 2, p. 456, Feb. 2018, doi: 10.3390/S18020456.

39 H. J. Baek, H. B. Lee, J. S. Kim, J. M. Choi, K. K. Kim, and K. S. Park, “Nonintrusive Biological Signal Monitoring in a Car to Evaluate a Driver’s Stress and Health State,” https://home.liebertpub.com/tmj, vol. 15, no. 2, pp. 182-189, Mar. 2009, doi: 10.1089/TMJ.2008.0090.

40 A. R. Guazzi et al., “Non-contact measurement of oxygen saturation with an RGB camera,” Biomed. Opt. Express, Vol. 6, Issue 9, pp. 3320-3338, vol. 6, no. 9, pp. 3320-3338, Sep. 2015, doi: 10.1364/BOE.6.003320.

41 J. Wang, J. M. Warnecke, and T. M. Deserno, “In-vehicle respiratory rate estimation using accelerometers,” Stud. Health Technol. Inform., vol. 261, pp. 97-102, 2019, doi: 10.3233/978-1-61499-975-1-97.

42 S. Kyu Leem, F. Khan, S. Ho Cho, J. Ureña, Á. Hernández Alonso, and J. Jesús García Domínguez, “Vital Sign Monitoring and Mobile Phone Usage Detection Using IR-UWB Radar for Intended Use in Car Crash Prevention,” Sensors 2017, Vol. 17, Page 1240, vol. 17, no. 6, p. 1240, May 2017, doi: 10.3390/S17061240.

43 T. P. Naziyok, A. A. Zeleke, and R. Röhrig, “Contactless patient monitoring for general wards: A systematic technology review,” Stud. Health Technol. Inform., vol. 228, pp. 707-711, 2017, doi: 10.3233/978-1-61499-678-1-707.

Notes