Applying Process Mining and RTLS for Modeling, and Analyzing

Patients’ Pathways

Sina Namaki Araghi

, Franck Fontanili

, Elyes Lamine

, Ludovic Tancerel

and Frederick Benaben

Industrial Engineering Center of Ecole des Mines d’Albi-Carmaux,

University of Toulouse, Allée des science, Albi, France

Maple High Tech, Toulouse, France

Keywords: Business Process Management, Process Mining, Real Time Location Systems.

Abstract: Purpose: This paper aims at introducing a generic approach for visualizing, analyzing and diagnosing

patients’ pathways. This approach could be categorized as a business intelligence approach to extract

knowledge for decision makers in healthcare organizations. The analyses provided by this approach are based

on the location data which are being recorded in the information system (IS) by indoor-Real-Time Location

Systems (RTLS). Findings: Healthcare organizations are getting more eager to learn from the execution of

their processes. They seek different tools and approaches to analyze the processes and visualize the problems.

This paper presents one of the possibilities to provide more understanding of process executions and it is

based on the locations of the patients inside the organization. Approach: Business intelligence approaches

provide new technical and technological solutions for business analysts to improve the quality of products

and services within organizations. The approach of this work helps to visualize patients’ pathways and analyze

them by associating real-time localization and process mining. This approach consists of four phases in which

several functionalities have been defined. These phases are Data, Information, Awareness, and Governance

(DIAG). Additionally, a case study has been designed to illustrate the DIAG approach.

1 INTRODUCTION

The clinical pathways or patients’ pathways consist

of several activities and different steps with the

objective of receiving a service which is the health

treatment. A clinical pathway is a complex

intervention for the mutual decision-making and

organization of care processes for a well-defined

group of patients during a well-defined period

(Vanhaecht et al., 2010). It is desirable to visualize

these pathways as the business processes, which have

the objectives and consist of different tasks and actors

to reach those objectives.

Many initiatives have been introduced within the

last decade to improve the management of clinical

processes. The clinical pathways aim at enhancing the

quality of healthcare services by improving the risk-

adjusted patients’ outcomes, promoting patient

safety, increasing patient satisfaction, and optimizing

the use of resources. These organizational

enhancements can be only justified on the basis of a

relevant diagnosis of the reality behind the processes’

executions and their management on the field. Hence,

equipping decision-makers with tools and methods to

facilitate and support their actions could beneficiate

the organization and its clients.

This paper aims at describing a study case which

illustrates a developing approach named as DIAG.

This method investigates different aspects of the

patients’ pathways and uses multiple sets of

techniques to visualize and analyze the performance

level of each pathway. DIAG has four different

phases; Data, Information, Awareness, and

Governance. This paper focuses on the “information”

phase to highlight the core of this method. It describes

how a proper set of location data from patients’

movements could beneficiate healthcare

organizations by acquiring qualitative and

quantitative analysis.

In the following, section 2 presents the features of

two main tools and techniques in DIAG method. The

third section explains the DIAG through exhibiting a

summary of a study case, which has been performed

at a hospital in France. Finally, the last section shows

the summary of the works and the perspective for the

future works.

540

Araghi, S., Fontanili, F., Lamine, E., Tancerel, L. and Benaben, F.

Applying Process Mining and RTLS for Modeling, and Analyzing Patients’ Pathways .

DOI: 10.5220/0006651605400547

In Proceedings of the 11th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2018) - Volume 5: HEALTHINF, pages 540-547

ISBN: 978-989-758-281-3

2 STATE OF THE ART

This section presents the notion of designing business

process models, the functionality of indoor-RTLS and

process mining.

To analyze the performance level of the

operational processes, organizations are trying to

move from the “data-aware” information systems to

“process-aware” information systems (Aalst, 2004).

This motivates organizations to use business process

management approaches, which consists of modeling

processes and performing quantitative analysis on top

of the process models.

To discover and analyze the business processes

several actions must be taken. To clarify, one should

understand the “life cycle of process modeling”. In

order to design business processes, different activities

should be taken into consideration such as gathering

information, clarification of data and designing the

process flows. After acquiring such models experts

should define criteria in accordance with their

strategies, then a performance measurement should

be defined (Looy and Shafagatova, 2016).

Currently, to gather data related to the process

executions an interview-based process discovery

approach is being used (Marlon Dumas, 2013.). To

succeed in such approach, it necessitates to carry out

a huge load of works to do the on-field observations

and perform interviews with processes’ actors.

Eventually, the quality of the gathered data is in doubt

and it is dependent on the experience of the

interviewer and the process analysts.

The evidence-based process discovery is another

approach which helps to automatically perform

process discovery from the information systems.

However, there are several inconsistencies in the case

of gathered data in the hospital information systems.

For example in the case of patients’ pathways, one

could extract the data from hospital information

system to observe the amount of time that each patient

has spent in the hospital sectors. Since people are in

charge to record the patients’ data, there are lots of

inaccuracies with the registered data.

This research work seeks for an efficient way of

gathering accurate data for visualizing the patients’

pathways. The final results should be a process

diagnosing platform which is able to visualize and

analyze the activities within patients’ pathways. In

addition to the process awareness, the platform could

provide an awareness of the location of each event.

This platform could support the transformation of

current information systems to process and location-

aware information system. The approach in this

research work aims at supporting this transformation

of the information systems by the application of I-

RTLS and process mining. In the following, these two

fields (indoor localization and process mining) will be

explained.

2.1 Indoor Localization Systems

The objective of these systems is to track objects

inside indoor environments. Their functionality is

similar to Global Positioning Systems (GPS) (Drawil

et al., 2013). GPS works by using NAVSTAR

satellites. The GPS functionality does not suit the

indoor localization and the corresponding signals are

not capable of finding an object inside buildings with

accurate coordinates (x, y, z). Therefore, indoor

localization technology is being used for this matter.

These systems consist of two main parts. First, mobile

nodes or tags which are attached to objects that need

to be located. Second, sensors or readers which find

the position of the tags or mobile nodes.

Multiple communication technologies are being

used for indoor positioning of objects in healthcare

and industry. RFID (Radio Frequency

Identification), and Wireless Local Area Networks

(Wi-Fi, ZigBee, Z-Wave, Bluetooth, etc.) are playing

a huge role in these organizations. RFID is an

electronic identification technology used for tracking

goods and people (Curran et al., 2011). This

technology uses radio frequency waves to transfer

data between tags and readers. RFID tags function in

three different states; active, semi-passive, and

passive. Active tags are using an internal battery, or

sometimes they are connected to an external power

source. These tags offer a wide range of hundred feet

or more. Passive tags do not have any energy source,

and their accuracy and readability are limited (Want,

2006). On the other hand, semi-passive tags use

batteries and they need an external power source to

communicate with readers (Lai et al., 2005). The

Readers have the role of transferring radio frequency

signals to the tags and receiving them back in

regulated version (Jin et al., 2008). Thanks to the

semantic event processing, the system could translate

the signals of the tags to the location data (Bok and

Yoo, 2017).

In the context of locating objects, RFID is being

used mainly in manufacturing sectors (Lu et al.,

2006). In the case of healthcare organizations, RFID

could not satisfy all the needs. It could identify the

tagged objects which are passing nearby the different

stations, but the tags should be moving through a very

well structured operation. Also, RFID systems do not

necessarily display the real-time location data. The

real-time information could be very useful in

Applying Process Mining and RTLS for Modeling, and Analyzing Patients’ Pathways

541

manufacturing and hospitals, especially in emergency

situations. Consequently, in some cases, researchers

have integrated RFID with WLAN to acquire real-

time data (Adame et al., 2016).

The other localization systems are classified as

Indoor-Real Time Localization Systems. These

systems have three main parts. Tags, which are

attached to the objects. These tags are active and they

emit signals continuously. Sensors or beacons; these

devices would receive the signals and send them to

the third part of the system which is the location

engine. This part would use different algorithms and

techniques to calculate the exact locations of the tags.

Some of these algorithms are Triangulation,

Trilateration, Angle of Arrival (AOA), Time

Difference of Arrival (TDOA), and Received Signal

Strength Indicator (RSSI) (Luo et al., 2011). Multiple

wireless local area networks have been introduced in

markets for indoor-RTLS, some of them are Wi-Fi,

and ZigBee which has been seen in previous works

that this technology could be applicable in service

sections. It has low energy consumption and multi-

channel control systems, alarm systems, and lighting

control (Cheng, 2009). Ultra-Wide Band (UWB),

these systems have short period pulses as a result

UWB is suitable for indoor positioning technologies.

UWB performance seems to be better than other

technologies as well as in terms of accuracy (less than

30 cm) and of course the range which is up to 200

meters within an open space. The weakness of this

system is inside buildings with thick walls and full

metal structures, due to the fraction and reflection of

signals in those types of buildings. Bluetooth or

Bluetooth Low Emission (BLE) is the other

technology which is being used in similar approaches

and it has a very low consumption level.

Some aspects of indoor localization have been

discussed in this section. The next part will introduce

process mining and highlights its association with I-

RTLS in the context of healthcare.

2.2 Process Mining

It is obvious that quality of services in hospitals could

directly influence the lives of people. Thus, it is

inevitable that hospitals should plan and act on the

improvement of their processes. The quality of

healthcare processes is highly dependent upon the

way they are being executed. This idea is motivating

researchers to provide healthcare organizations with

new approaches such as process mining to illustrate

the execution of the processes. In addition, execution

of medical and non-medical processes within the

healthcare organizations would result in numerous

recorded events within information systems. These

data would be displayed as a log file. A log file would

consist of a multiple information such as case id,

timestamp start, timestamp end, details of activities,

resources, and other types of information. One of the

challenges for organizations is to extract meaningful

information from these log files (Mans et al., 2013).

Process mining is a set of techniques and methods

which are subjected to discover the business

processes from event logs. Process mining consists of

three main activities: (i) process discovery, (ii)

conformance checking, (iii) enhancement (Aalst,

2016). By using this method information systems are

capable of translating raw log files into usable sets of

information. This information could be the process

models, general statistical analysis of processes,

bottleneck analysis, analysis of a variety of cases in

event logs and etc. There are multiple advantages for

organizations in applying process mining techniques.

In the context of healthcare, study cases show that

process mining could beneficiate directors by

providing several perspectives for decision-making

(Rojas et al., 2016). A control-flow perspective could

show the execution of the processes. A performance

perspective highlights the problems and bottlenecks

while executing the processes. Currently, these

analyses are basically related to analyzing the time-

oriented data (Fernandez-Llatas et al., 2015).

Additionally, conformance checking would help

organizations to see where the deviations are in a

process. The resource usage analysis is the other

perspective offered by process mining (Caron et al.,

2014).

Process mining techniques and algorithms are

being developed in many tools. ProM (http://www.

processmining.org/), Disco (https://fluxicon.com),

QPR (https://www.qpr.com), Celonis (https://www.

qpr.com) are front-runners in the market. Most of the

healthcare cases are being developed by ProM (Rojas

et al., 2016), which is extremely powerful in applying

multiple techniques on event logs (Dongen et al.,

2005). Disco is a user-friendly and powerful tool,

however, it does not contain all the analyzing

techniques like ProM. On the other hand, Celonis and

QPR are two tools which are known and being used

mainly in manufacturing industries.

Applying process mining and indoor localization

systems did not receive sufficient amount of

attention, however, it shows promising results in

healthcare. For instance, Fernandez-Llatas et al used

this idea in order to monitor the behavior of the

patients in a nursing home by analyzing the pathways

of the patients (Fernández-Llatas et al., 2013). In

another case, they presented an approach for applying

HEALTHINF 2018 - 11th International Conference on Health Informatics

542

RTLS and process mining in a hospital in Valencia

which also shows intriguing outcomes in case of

observing the movement of patients and to analyze

the time aspect of the process execution (Fernandez-

Llatas et al., 2015). In both cases, they have used

PALIA ILS suits which is a web-based process

mining tool to discover the process maps. PALIA

stands for Parallel Activity-based Log Inference

Algorithm and is able to infer workflows from

activity log samples. And ILS is for Indoor

Localization Systems. Also, they have introduced a

methodology which supports their application. This

methodology consists of several steps as ILS

installation, ILS data gathering, a semantic

aggrupation of areas, process filtering, process

discovery, process conformance, process

enhancement, and process improvement. In their

methodology the difference between process

enhancement and process improvement is vague.

In this paper, the process improvement actions are

outside the application of process mining and indoor

localization systems because the resolution to

improve and change a process is only feasible by

decisions coming from human knowledge.

In modern businesses, due to the challenges of

designing and monitoring business processes, there is

a necessity of using suitable analysis techniques

(Aalst, 1998). Also, it has been inferred from other

works that analyzing business processes with

exclusive criteria is necessary in order to classify the

business processes’ attributes, identify the

bottlenecks, and compare process variants (Vergidis

et al., 2008). Additionally, Vergidis et al have

classified process analysis into three groups of

diagrammatic models, business process languages,

and mathematical models. Needless to say, their

approach is based on the primary analysis of

processes which is the method one could use to

visualize the processes.

It has been seen in the work of Zakarian in

(Zakarian, 2001) that using diagrammatic process

modeling techniques provide a qualitative notation on

the process executions. However, these techniques

suffer from the lack of tools which show the

quantitative analysis and performance level of

processes. Furthermore, generating business process

models is impractical without diagnosis based on

relevant key performance indicators. This analysis

should be in line with the main objective, which is

improving the performance of physical processes.

Improving business processes is only feasible by

understanding the features of each activity and events

in the process in accordance with customer and

process owner’s perspectives. For example, a task

such as recording the history of patient’s treatment

could be a non-value added activity from patient’s

perspective. But from the hospital’s point of view,

this task should be identified as a value-added

activity. As Dumas et al define in (Marlon Dumas,

2013) we could have three types of activities, value-

added, non-value added, and business value-added

activities. This research work aims at analyzing

business processes based on this classification of

activities. However, the question is how to discover

the activities in the business processes and how to

distinguish them.

3 DETAILED PROPOSAL

It has been shown that business intelligence (BI)

approaches in healthcare present new solutions for

business analysts within the organizations especially

the healthcare sector for improving the quality of

medical care and patient’s quality of life (Machado

and Abelha, 2001). It could be inferred that BI

approaches encompass the strategies and

technologies used by enterprises to help business

analysts to understand and improve the quality level

of business processes (Dedić and Stanier, 2016). It’s

been proven that BI tools are working efficiently with

healthcare data and they are able to generate real-time

information and knowledge relevant to the success of

healthcare organizations. Moreover, BI approaches

profit healthcare professionals in making vital

decisions inside hospitals, clinics, paramedics’

circulation and management of the administrative

works.

This paper aims at suggesting DIAG approach

which is embracing the BI notion and tries to generate

knowledge from the location data for decision makers

in hospitals. DIAG stands for four levels which could

transform the raw data into knowledge and awareness

for healthcare experts. These levels are Data,

Information, Awareness, and Governance.

The IDEF0 model presented in figure1 gives an

overview of the DIAG approach. In the first phase,

one function has been defined as gathering. This

function is mainly concerned with monitoring the

movements of the objects and receiving the primary

data come from localization systems. As it has been

explained before, these systems use the positioning

algorithms to locate the tags. The log refining

function executes different data mining techniques in

the context of RIO-DIAG platform. This function

cleans and transforms the collected data in the first

function into the proper event logs which suit the

process mining techniques in the following function

Applying Process Mining and RTLS for Modeling, and Analyzing Patients’ Pathways

543

Figure 1: The DIAG methodology.

modeling. The modeling function uses the process

discovery algorithms and illustrates the execution of

a processes based on the qualitative process analysis.

Fourth, in analyzing function, different techniques are

being used in order to provide a quantitative analysis

for users. The performance level of processes will be

evaluated in this function, and one could see and

comprehend the way patients are circulating in the

organization to receive their treatment.

In the diagnosing function, multiple sets of

techniques will be used to highlight the real cause

behind the problems and weaknesses in the

performance of processes. Last but not least, in the

prognosing by using queuing theory and other

simulation techniques different scenarios and

improvement actions will be suggested. RIO-DIAG is

the platform which is working based on these

functions, and by fulfilling four different phases of

DIAG approach provides opportunities to apply BI on

monitoring patients’ pathways. Describing all the

functionalities is out of the scope of this paper,

however, modeling and analyzing, two of the main

functions of this approach will be illustrated.

To clarify this method a study case has been

developed to show how one could visualize and

analyze the patients’ pathways. In this experiment, an

indoor-RTLS had been used to collect the location

data. After installing the sensors and calibrating the

location systems, patients received the tags. These

tags communicate with sensors by UWB technology.

The transmitted signals would be calculated in the

location engine by using TDOA and AOA

measurement techniques. In this experiment, fourteen

patients have been monitored. Each patient has a case

ID based on the tag identification. In this case, there

was no return of the same patient to the hospital.

Twelve different activities have been identified based

on the name of the zones in the hospitals. Each time a

patient went inside or outside a section an event had

been registered in the primary log file. The entities in

the log file are related to the type of indoor-RTLS.

In this experiment, the log file consists of patients’

ID, activities’ details, timestamp start, and timestamp

end. Figure 2 shows a part of the event log related to

the patient 171. The primary event log consists of one

hundred events, and it is transformed and prepared for

the modeling function. As it has been shown, the first

and last events related to the case 171 have the

duration of “00:00”. This indicates that these events

are only instants while the process is being executed

and they have been seen as a “passage” in the facility.

Figure 3 shows that the longest event for the

patient 171 was staying in the “Consultation Box 1”

to meet with the doctor. In addition to this data, the

distance of the whole pathway was two hundred and

sixty-two meters. In addition, it has been shown that

the patient has spent approximately 54 minutes in the

facility. Ability to reach such analysis is the result of

HEALTHINF 2018 - 11th International Conference on Health Informatics

544

applying RTLS and acquiring the accurate data

related to the processes’ executions. Now, the

question is the value classification of these activities

and which share of the process duration was value-

added for the patients. This knowledge of the value

class of activities is being provided by the healthcare

experts. Furthermore, there is a need to select the data

about one specific patient’s pathway and to analyze it

separately.

Figure 2: The event log for the case 171.

Figure 3: The basic analysis for patient 171.

To analyze the patient’s pathways as the business

processes, it would be formidable to acquire a

qualitative analysis by displaying the process map

and a quantitative analysis based on different aspects

of the performance of the process. Thanks to process

mining, figure 4 shows the process map related to the

patient 171. Such process map provides a care-flow

perspective on the patient’s pathways. Explanations

regarding the discovery algorithms are broader than

the scope of this paper since the core and possible

outcomes of DIAG approach is being illustrated in

this article.

To acquire a performance perspective on process’

execution, a Critical To Quality (CTQ)

(Montgomery, 2007) has been defined for this

experiment as “Process Efficiency” in the hospital. A

quantitative indicator has been proposed as “Length

of Stay” or duration to analyze the performance level

of the processes based on the CTQ. The analysis

related to the indicator has been shown in table 1.

These analyses are based on the value-class of the

activities and the data coming from localization

systems. By looking at the process efficiency, one

could observe that the 85% of the process duration

was related to executing the value-added activities

and close to 15% of process duration was associated

with the waiting time and staying in the queue. The

decision-making actions are now dependent on the

knowledge of the healthcare experts to determine how

they can improve the non-value added parts of the

process. Although, the simulation step could be a

compatible solution to see several options for

enhancing the situation.

Indeed, one could infer that a logical decision

could be made only by receiving the proper

knowledge which is the aim of this approach. As it

has been presented, this approach could support the

quality of the decisions by providing the care-flow

and performance perspectives on the processes.

Figure 4: The process map for patient 171.

Applying Process Mining and RTLS for Modeling, and Analyzing Patients’ Pathways

545

Table 1: Process efficiency analysis based on Length of

Stay.

Activities (patient

171)

Duration

VA/ NVA

Process

Efficiency

Consultation

Department

PASSAGE

85.2 %

Waiting Room

4:39

NVA

Registration Desk

9:02

Waiting Room 3

1:11

NVA

Pre-Consultation Box

6:01

Waiting Room 3

PASSAGE

Consultation Box1

29:17

Registration Desk

2:37

Gate

2:19

NVA

Exit area

PASSAGE

4 CONCLUSION

This paper aimed at presenting the possibilities and

advantages of applying process mining on location

data to provide more awareness regarding the

patients’ pathways. This paper identified the DIAG

approach which uses the indoor-RTLS, process

mining, and business process management techniques

to visualize, analyze and diagnose the patients’

pathways. This approach embraces the hierarchy of

data, information, awareness, and knowledge which

extract the location data from the movements of the

patients and transform these data to the knowledge.

The real-time location data provide more accuracy in

tracking patients’ activities in the hospitals and

process mining permits to model those activities as

business process models. The advantage in this

approach is in providing the quantitative and

qualitative analysis based on the value class of the

activities in the organizations.

To illustrate the possible outcomes of this

approach an experiment has been performed at a

hospital in France. Additionally, an overview has

been provided for the major techniques and

technologies in this approach.

To evolve this research work, the upcoming works

are oriented towards refining this approach and

providing the details about the rules and regulations

for transforming the location data into prepared event

logs for process mining tools.

REFERENCES

Aalst, W. M. P. V. D., 1998. The Application of Petri Nets

to Workflow Management.

Aalst, W. M. P. van der, 2016. Process Mining: Data

Science in Action. Springer.

Aalst, W. M. P. van der, 2004. Business Process Management

Demystified: A Tutorial on Models, Systems and

Standards for Workflow Management, in: Desel, J.,

Reisig, W., Rozenberg, G. (Eds.), Lectures on

Concurrency and Petri Nets, Lecture Notes in Computer

Science. Springer Berlin Heidelberg, pp. 1–65.

Adame, T., Bel, A., Carreras, A., Melià-Seguí, J., Oliver,

M., Pous, R., n.d. CUIDATS: An RFID–WSN hybrid

monitoring system for smart health care environments.

Future Gener. Comput. Syst. doi:10.1016/j.future.

2016.12.023

Bok, K., Yoo, J., 2017. RFID Based Indoor Positioning

System Using Event Filtering. J. Electr. Eng. Technol.

12, 335–345. doi:10.5370/JEET.2017.12.1.335

Caron, F., Vanthienen, J., Vanhaecht, K., Limbergen, E.V.,

De Weerdt, J., Baesens, B., 2014. Monitoring care

processes in the gynecologic oncology department.

Comput. Biol. Med. 44, 88–96. doi:10.1016/

j.compbiomed.2013.10.015

Cheng, Y. M., 2009. Using ZigBee and Room-Based

Location Technology to Constructing an Indoor

Location-Based Service Platform, in: 2009 Fifth

International Conference on Intelligent Information

Hiding and Multimedia Signal Processing. Presented at

the 2009 Fifth International Conference on Intelligent

Information Hiding and Multimedia Signal Processing,

pp. 803–806. doi:10.1109/IIH-MSP.2009.106

Curran, K., Furey, E., Lunney, T., Santos, J., Woods, D.,

McCaughey, A., 2011. An Evaluation of Indoor

Location Determination Technologies. J Locat Based

Serv 5, 61–78. doi:10.1080/17489725.2011.562927

Dedić, N., Stanier, C., 2016. Measuring the Success of

Changes to Existing Business Intelligence Solutions to

Improve Business Intelligence Reporting, in: Research

and Practical Issues of Enterprise Information Systems,

Lecture Notes in Business Information Processing.

Presented at the International Conference on Research

and Practical Issues of Enterprise Information Systems,

Springer, Cham, pp. 225–236. doi:10.1007/978-3-319-

49944-4_17

Dongen, B. F. van, Medeiros, A. K. A. de, Verbeek, H. M.

W., Weijters, A. J. M. M., Aalst, W. M. P. van der,

2005. The ProM Framework: A New Era in Process

Mining Tool Support, in: Applications and Theory of

HEALTHINF 2018 - 11th International Conference on Health Informatics

546

Petri Nets 2005. Presented at the International

Conference on Application and Theory of Petri Nets,

Springer, Berlin, Heidelberg, pp. 444–454.

doi:10.1007/11494744_25

Drawil, N. M., Amar, H. M., Basir, O. A., 2013. GPS

Localization Accuracy Classification: A Context-Based

Approach. IEEE Trans. Intell. Transp. Syst. 14, 262–

273. doi:10.1109/TITS.2012.2213815

Fernández-Llatas, C., Benedi, J.-M., García-Gómez, J. M.,

Traver, V., 2013. Process Mining for Individualized

Behavior Modeling Using Wireless Tracking in

Nursing Homes. Sensors 13, 15434–15451.

doi:10.3390/s131115434

Fernandez-Llatas, C., Lizondo, A., Monton, E., Benedi,

J.-M., Traver, V., 2015. Process Mining Methodology

for Health Process Tracking Using Real-Time Indoor

Location Systems. Sensors 15, 29821–29840.

doi:10.3390/s151229769

Fundamentals of Business Process Management, Marlon

Dumas, Springer, n.d.

Jin, X., Lee, X., Kong, N., Yan, B., 2008. Efficient

Complex Event Processing over RFID Data Stream, in:

Seventh IEEE/ACIS International Conference on

Computer and Information Science (Icis 2008).

Presented at the Seventh IEEE/ACIS International

Conference on Computer and Information Science (icis

2008), pp. 75–81. doi:10.1109/ICIS.2008.60

Lai, E., Redfern, A., Wright, P., 2005. Vibration

Powered Battery-Assisted Passive RFID Tag, in:

Embedded and Ubiquitous Computing – EUC 2005

Workshops. Presented at the International Conference

on Embedded and Ubiquitous Computing, Springer,

Berlin, Heidelberg, pp. 1058–1068. doi:10.1007/

11596042_108

Looy, A. V., Shafagatova, A., 2016. Business process

performance measurement: a structured literature

review of indicators, measures and metrics.

SpringerPlus 5, 1797. doi:10.1186/s40064-016-3498-1

Lu, B. H., Bateman, R. J., Cheng, K., 2006. RFID enabled

manufacturing: fundamentals, methodology and

applications. Int. J. Agile Syst. Manag. 1, 73–92.

doi:10.1504/IJASM.2006.008860

Luo, X., O’Brien, W. J., Julien, C. L., 2011. Comparative

evaluation of Received Signal-Strength Index (RSSI)

based indoor localization techniques for construction

jobsites. Adv. Eng. Inform., Information mining and

retrieval in design 25, 355–363. doi:10.1016/j.aei.

2010.09.003

Machado, J., Abelha, A., 2001. Applying Business

Intelligence to Clinical and Healthcare Organizations.

IGI Global.

Mans, R. S., Aalst, W. M. P. van der, Vanwersch, R. J. B.,

Moleman, A. J., 2013. Process Mining in Healthcare:

Data Challenges When Answering Frequently Posed

Questions, in: Lenz, R., Miksch, S., Peleg, M., Reichert,

M., Riaño, D., Teije, A. ten (Eds.), Process Support and

Knowledge Representation in Health Care, Lecture

Notes in Computer Science. Springer Berlin

Heidelberg, pp. 140–153.

Miclo, R., Fontanili, F., Marquès, G., Bomert, P., Lauras,

M., 2015. RTLS-based Process Mining: Towards an

automatic process diagnosis in healthcare, in: 2015

IEEE International Conference on Automation Science

and Engineering (CASE). Presented at the 2015 IEEE

International Conference on Automation Science and

Engineering (CASE), pp. 1397–1402. doi:10.1109/

CoASE.2015.7294294

Montgomery, D. C., 2007. Introduction to statistical quality

control. John Wiley & Sons.

Rojas, E., Munoz-Gama, J., Sepúlveda, M., Capurro, D.,

2016. Process mining in healthcare: A literature review.

J. Biomed. Inform. 61, 224–236. doi:10.1016/

j.jbi.2016.04.007

Vanhaecht, K., Sermeus, W., Peers, J., Lodewijckx, C.,

Deneckere, S., Leigheb, F., Decramer, M., Panella, M.,

2010. The impact of care pathways for exacerbation of

Chronic Obstructive Pulmonary Disease: rationale and

design of a cluster randomized controlled trial. Trials

11, 111. doi:10.1186/1745-6215-11-111

Vergidis, K., Tiwari, A., Majeed, B., 2008. Business

Process Analysis and Optimization: Beyond

Reengineering. IEEE Trans. Syst. Man Cybern. Part C

Appl. Rev. 38, 69–82. doi:10.1109/TSMCC.2007.

905812

Want, R., 2006. An introduction to RFID technology.

IEEE Pervasive Comput. 5, 25–33. doi:10.1109/

MPRV.2006.2

Zakarian, A., 2001. Analysis of Process Models: A Fuzzy

Logic Approach. Int. J. Adv. Manuf. Technol. 17, 444–

452. doi:10.1007/s001700170162.

Applying Process Mining and RTLS for Modeling, and Analyzing Patients’ Pathways

547