Data Mining for the Unique Identification of Patients in the National

Healthcare Systems

D. G. Ramírez-Ríos

, Laura P. Manotas Romero

, Heyder Paez-Logreira

Luis Ramírez

and Yohany Andrés Jimenez Florez

Research Department, Fundación Centro de Investigación en Modelación Empresarial del Caribe,

FCIMEC, Carrera 60 # 64-122, Barranquilla, Colombia

Research Department, Logyca, Av El Dorado # 92-32, Bogotá, Colombia

Keywords: Data Mining, Databases, Secretary of Health, Duplicities, Healthcare System.

Abstract: This paper considers the application of data mining (DM) algorithms as a feasible and necessary strategy for

optimal management of databases (DB) in the national healthcare systems. Specifically it deals with the

management of multiple DB that consider patient’s affiliation information, under the supervision of the

authorities in healthcare, an issue that involves not only the issues of every citizen but also its integral right

to be treated by any institution. We support the idea that the administrative part of the healthcare system

should not obstruct the attention of the patient and a total efficiency must be guaranteed. We believe that

DM algorithms are appropriate for this task and human intervention should be minimized. A case study was

developed in Colombia that considered the multiple affiliations to DB and its integration to a unique DB

managed by the District Health Secretary (DHS, which detected frauds and other type of duplicities. The

mechanism used to approach this, indicates not only a significant reduction of manual intervention of the

DB, but also allows the extraction of data for future analysis, supporting the patient’s need for an efficient

and integral health attention, as well as privacy of personal information registered.

1 INTRODUCTION

The institutions that provide healthcare services and

the administrative entities that affiliate patients in

the public sector are obligated to respond to the

patient’s needs. This implies the activities directly

related to the health services provided for the

patients and the administrative activities that allow

the organization of their affiliations, removals and

other services, which are done to provide an

adequate and integral medical attention. According

to the health policy, the authorities are in charge of

regulating these activities and as part of this, the

management of DB plays an important role in their

daily activities.

The information stored in these DB must be

organized, reliable, free of errors and duplicities,

guaranteing its quality, and thus, assuring a correct

and unique identification of the users and their full

rights to the health services associated to his/her

health plan. “This is why an implementation of

unique health IDs are a requisite for the installation

of politics and applications of the TICs in the sector”

(Oviedo and Fernández, 2010).

The management of inconsistencies that can be

detected inside the health information systems

constitutes in a crucial aspect of the processing of

data and is determinant over the benefits that a

patient will or will not receive in terms of the

services provided by the healthcare system. This is

why a correct identification of users in a DB free of

duplicities or multiple affiliations is strictly

necessary and a responsibility of the entities in

charge of the control and administration of the

health sector.

One of the main problems identified in the

management of DB is the correct identification of

users and the organization of the data in the fields of

the DB. This type of errors may cause problems

during the formal affiliation of the patient (Esp and

Ramírez, 2009). Furthermore and even more critical,

a mistake in the affiliation attempts to the physical

integrity of the patient, possibly aggravating its own

health situation for a withdrawal or delay of the

healthcare service required (McCoy, et al., 2013),

not to mention the elevated costs involved in the

administration of the health entities (McClellan,

2009).

211

G. Ramírez-Ríos D., P. Manotas Romero L., Paez-Logreira H., Ramírez L. and Andrés Jimenez Florez Y..

Data Mining for the Unique Identiﬁcation of Patients in the National Healthcare Systems.

DOI: 10.5220/0005287302110217

In Proceedings of the International Conference on Operations Research and Enterprise Systems (ICORES-2015), pages 211-217

ISBN: 978-989-758-075-8

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

The entities in charge of the administration, attention

and regulation of the subsidized healthcare services

have a, even greater, responsibility of identifying the

users of the health system, without making any

mistakes. The information registered in the DB is

necessary for accessing the government funds that

covers the health plan of a subsidized patient. Thus,

the local government entities of control and

monitoring, periodically make revisions of the data

registered in the information systems and check for

inconsistencies reported with regards to subsidized

users.

Given the actual situation, the information

system plays an important role in guaranteeing the

stability of the healthcare system. DBs are a key

element in the administration of the information of

users in the subsidized healthcare system. In order to

guarantee that the system counts with complete and

clean data (information free of errors), several DBs

of users must be integrated, such as, the DB that

contains the deceased, the new affiliations, the

withdrawals, the transferred, among others. This

integration must be done at a timely basis given that

periodically there must be a report and the efficiency

of the system must be maintained. DM is considered

for this matter, given that it involves techniques and

algorithms that allow correct and optimal

management of DBs, as well as the use of

information to gain knowledge over the population

involved.

This research takes into account a point of view

of the problem described above with respect to the

correct identification and administration of

inconsistencies in the registered data in the

healthcare system. Particularly, this research

identifies DM as an appropriate tool used for the

timely detection of inconsistencies by many

information systems. While some believe that DM is

a robust and complex tool to be used for the

detection of duplicities in the DB registrations in any

information system, we believe it’s completely

necessary in order to get clean data and at the same

time, obtain new knowledge from the data and a

profound analysis of its behavior with respect to the

abnormalities presented, that can become

compelling to the overall quality of the system.

The paper is organized as follows: On section 2,

a the state of art in DM applied to the Health Sector

is given, supported by some applications with

respect to duplicity detection on DBs; section 3,

presents the case study developed, Unique

Identification System for Users (SIUU) for the

Health Sector, and guidelines of the solution are

proposed; then, on section 4, Advantages and

Disadvantages of DM in a SIUU, shows the

importance of the DM for the detection of

duplicities, patients’ needs and the complexity of the

solution; the last section presents concluding

remarks and considerations to take into account

when implementing the project.

2 DATA MINING APPLIED TO

THE HEALTH SECTOR

A DB is a set of data that belong to the same context

and are stored in a structural way for its further use

(Date and Date, 1990). A DB provides institutions

the access to information, in a way that it can be

visualized, managed and updated, according to the

access rights given (Batra, Parashar, Sachdeva, and

Mehndiratta, 2013). With respect to the case study

developed under this research, the DB identified as

FOSYGA (MinSalud, 2014) is in charge of storing

the Colombian healthcare information system with

respect to the affiliation information. This DM

provides access to sensitive information of the users

registered in the system, which represents close to

91,69% of the entire Colombian population (DANE,

2013).

One of the most wearying activities to be done in

terms of the administration of information is to keep

the DB updated. In the Colombian healthcare

subsidized system (RSS), local authorities must

guarantee that the data updated is free of errors,

since the payment given for the healthcare attention

of a user that no longer belongs to the system is

absorbed by the entities that offer the service and are

not benefitting any other users. The identification of

multiple registrations in this type of DB allows for a

correct use of the government funds for healthcare

services.

This same issue has been identified and

approached in other countries, such as New Zealand,

England, Spain, among others. In these countries,

they have created a unique identification system for

patients and have established some technological

and legal frameworks in order to support and

regulate the processes of affiliation and registration

of patients in the system (Oviedo and Fernández,

2010).Yet, the problem is still present with or

without the implementation of a unique identifica-

tion system, given that the DB must be integrated

and the data must be clean in order to use this

information in the decision making process. DM has

been approached to solve this issue, given that it

gives the controlling entities the capacity to

automatically classify and correct errors in the data.

ICORES2015-InternationalConferenceonOperationsResearchandEnterpriseSystems

212

DM has been important, not only in the organization

and consolidation of DBs, it has also been of

assistance with regards to statistics, mainly in the

identification of vulnerable populations, types of

treatments, geographic relationships and even the

knowledge of the evolution of epidemics (Kaur and

Wasan, 2006). Nevertheless, when DM is used for

the regulation of DBs in the management of users,

its impact is not too significant (Holzinger and

Jurisica, 2014).

This leaves us the following questions: What if

there existed such an integrated and complete DB of

the users of the whole country, which contains from

the registers of ID number to the number of

deceased? How could it be exploited by using DM?

What is the real impact in applying the complex

algorithms found in DM into the management of

user-based DB? The response to these questions

leads us to consider several points of view. Some

opinions may enhance the benefits, others the

disadvantages of DM in several situations (Marcano

Aular and Talavera, 2007) (Harrison, 2013)

(Yucatan, 2014) (Harrison, 2013) (DeBariloche,

2014) (Pagliery, 2014).

There are multiple areas in which DM has been

successfully implemented for the optimal

management of DB (Hsu-Hao, 2012) in the health

sector or user DB. Dávila Hernández and Sánchez

Corales consider the concept “Clinical Decision

Making Support Systems” (CDSS), which have

proven to be fundamental in reducing the medical

errors and improving the healthcare processes. A

CDSS employs DM as a study method and is used

for the classification of data, generating new

knowledge from stored data. This research explains

the contributions to the diagnosis of diseases using

DM through the combination of two mathematical

models. These models were applied to a case study

on arterial hypertension and, as a result, behavior

patterns were discovered with relation to factors that

can raise the risk of having the disease (Davila

Hernandez and Sanchez Corrales, 2012).

On the other hand, (Viveros, Nearhos, and

Rothman , 1996) discuss the effectiveness of two

techniques in DM that can be used to analyze and

predict behavior patterns unknown to DBs that are

registered by health insurance companies. The DB

used were associated to pathology and general

practitioners services. The techniques used in DM

were association rules for pathology services and

neural segmentation for the consolidation and

evaluation of both DB integrated. The study

demonstrates that DM algorithms can be used

satisfactorily in huge data sets at a reasonable time

and without employing too much computational

effort. These results can be transformed into

quantitative benefits and support decision making.

Among the results shown, the study found an

overpayment of more than $550.000 US per year, a

figure that was not found in the conventional

monitoring techniques.

It is possible to observe that DM is applied with

much greater frequency in the follow up of

treatments, diseases, patients and medicine in the

health sector. With respect to the issues encountered

in the administration of registered patient

information of nationwide DBs that guarantees a

correct and error free identification of users, DM is

not too popular and it has been observed that for

these cases, conventional algorithms are used for the

detection of duplicities or other abnormalities.

Such a case is explained in (McCoy, et al.,

2013), which evaluates the percentage of duplicities

encountered in the Electronic Health Records (EHR)

of five entities. This research establishes what is

known as coincidence indicators, applied for the

correct detection of duplicities, which can be easily

adjusted to any entity. The algorithms employed

have shown to be effective, given the increase in the

amount of duplicities encountered.

On the other hand, DM and other advanced

techniques used for the treatment of information

stored in DBs have been applied for the detection of

duplicities. In (Elmagarmid, Ipeirotis, and Verykios,

2007) a literature review is given and several

methods used for solving the detection of duplicities

are analyzed. For example, probabilistic approaches,

automatic learning techniques and other variations,

presenting the metrics and support tools in the

application of systems for the detection of errors in

DM.

3 UNIQUE IDENTIFICATION

SYSTEM FOR USERS (SIUU)

FOR THE HEALTH SECTOR, A

CASE STUDY

A case study was implemented in the District Health

Secretary (DHS), the dependency of the District

Authorities, in charge of directing, coordinating and

supervising the District’s Health System. It is in

charge of providing healthcare services for the

benefit of the community, such as, promotion,

prevention of diseases, protection of the

environment and health restoration. In order to fulfill

all of its functions satisfactorily, the DHS requires

DataMiningfortheUniqueIdentificationofPatientsintheNationalHealthcareSystems

213

integrated information of their users and statistics of

the behavior of the health sector population in their

municipality, in general.

The first step to obtaining the information of

the health systems corresponds to the affiliation and

registration of the users in the system. Counting with

unique and reliable information of each patient is

very important for guaranteeing a timely, efficient

and economical attention in their healthcare plan.

The process of affiliation and registration of the

patients is in charge of the Public Health Entities

(EPS) that belong to the health system. These

entities have the obligation of reporting the

affiliations made to the National Health System and

to the corresponding DHS. On the other hand, the

DHS must supervise and control that the new

affiliations are correct and unique for each patient.

In their process of supervision, the DHS

compares the DB of the reported affiliations

nationwide to other DBs, such as the National

Registration, in order to verify data and

identification of the users. These DBs are not

integrated nor standardized, which implies that

reported inconsistencies come in different formats,

hindering the automatic integration of the DB.

This process is done at a monthly basis, where

the EPS reports and sends documents weekly but the

validation and verification takes longer, many times

creating unsatisfied users and difficulties in the

optimal response to situations presented.

The process of verification and validation of

registered users face two delicate decisions: (1)

removal of users from the RSS successfully

registered in the system, because they are considered

duplicates or invalid, creating a "false positive"; or

(2) not detecting invalid users or belonging to

another health system (including RSS, generating a

duplication), and thus creating a "false negative".

Both scenarios pose severe consequences, which

implies having a user without health services or

having a user affiliated twice, having a user who

does not have the right to be benefitted or having

one that does not even exist.

In order to effectively operate, a unique

identification DB (SIIU), integrated and normalized,

is proposed. For the structural model of the DB, data

modelling and UML Class diagrams have been

employed (Teorey, Lightstone, Nadeau, and

Jagadish, 2011).With respect to the new integrated

DB, it is possible to apply the techniques commonly

used for the detection of duplicities, particularly

through automatic learning. With the clean data and

the duplicities identified, it is possible to apply DM

for generating new knowledge from registered data,

specifically the information concerning new

affiliations, with regards to duplicities identified and

common mistakes presented in the system. Other

information concerning scope of healthcare attention

and services provided, are also considered.

Figure 1 presents the actual and proposed

mechanism for the affiliation of users and the

detection of duplicities. The actual process begins

with the EPS that sends the information both

physically and digitally to the DHS, yet, the formats

may vary among entities and the electronic media

used for digital documents. The proposed

mechanism is based on a server platform for the

DHS in order to receive the requests digitally in a

standardized document. The affiliation processes are

to be supported by ITS and technologies for

scanning documents automatically.

Figure 1: Affiliation process. Current and Proposed.

With the proposed mechanism, a few

administrative steps are eliminated, yet the times of

response and supervision processes are reduced

significantly, making the process more efficient.

With respect to the DBs, integrated and normalized,

some DM techniques are applied in order to detect

duplicities and other abnormalities (Elmagarmid,

Ipeirotis, and Verykios, 2007).

The proposed mechanism considered in this case

study is known to improve the times it takes in the

processes of affiliation, validation and correction of

ICORES2015-InternationalConferenceonOperationsResearchandEnterpriseSystems

214

registered data. Additional to this, valuable

information that can be extracted from this DB

through DM, such as, vulnerable population, most

common emergency cases, types of attention

required in the different age ranges, epidemic alerts,

and entities with the most claims registered, among

others.

4 ADVANTAGES AND

DISADVANTAGES OF SIUU

Data mining is composed by a set of techniques

widely recognized and applied in numerous fields.

However, their use in certain applications must be

evaluated from the perspective of the requirements,

the ability of technological development and

acquiring real customer needs, aside from the

fashions and preferences based on developer

experience.

From an initial review "DM is a technique that

optimizes and improves the effectiveness in

detecting duplicate records in databases." The

detection of duplicate records is one of the simplest

processes in Data Cleansing. Cleaning records in a

database is studied in conjunction with data mining

and other areas.

Comparing the records, one by one, is the most

reliable duplication detection process in a database.

However, this technique is time consuming,

demands lots of processing resources and is

conditioned on the number of records to evaluate

(Wai Lup, Mong Li, and Tok Wang, 2001).

Other techniques have been proposed, using

algorithms such as “nearest neighbor”, to reduce the

consumption of time and resources in duplication

detection algorithms. Another technique is to have a

stack of records previously detected and prioritize

these to be the first to be evaluated.

Lower complexity algorithms of Data Mining

and Data Cleansing are applied in duplicate

detection such as Soundex, assessing a 95.99%

effective records (Elmagarmid, Ipeirotis, and

Verykios, 2007). These algorithms or methods are

suitable for detecting duplicities on individual fields.

Nevertheless, the detection record consisting of

multiple fields is a more complex problem, which

requires the application of probabilistic approaches

and supervised machine learning techniques, used in

DM.

A second opinion suggests that "DM is a very

complex and robust technique to be applied to a

single user registration process." DM is not an easy

task and consumes a lot of human and equipment

resources. DM implementation involves the

acquisition of query and analysis tools and training

the users (Xintong, Hongzhi, Song, and Hong,

2014).

The challenge of DM, but also one of its

advantages, is to be a framework that integrates

multiple approaches from different disciplines and

knowledge areas (Bellazzi and Zupan, 2008). The

application of DM in a specific domain problem

requires that developers are not only experts in DM

but also acquire a considerable level of knowledge

about the problem itself. Application of Data Mining

in a specific field require a proper analysis of the

problem domain and modeling solution (Shu-Hsien,

Pei-Hui, and Pei-Yuan, 2012) to establish an

appropriate methodology for the problem, for

example, detection of records and duplication in the

Health Sector.

When analyzing the information related to

people and the method of implementation of DM for

Data Cleansing, some negative aspects may appear,

which are classified by several authors in four

general key factors claves (Han and Gao,

2008):Security: Although entities can manage large

amounts of data that contain personal and

confidential information of the users, in occasions,

there are no mechanisms that prevent the loss and/or

stealing of data, generating a risk to the security of

the users registered.

1. Privacy: DM requires data to be exposed to

the processes applied, so it is necessary to

accompany this information with the

appropriate security techniques and

encryption protocols.

2. Accuracy: Sometimes an error in the data

processing could generate a huge problem if

the information is interpreted or processed

incorrectly.

3. Complexity: Huge investment for processing

information.

From these two points of view, the application of

DM in the District’s DB for a system of duplication

detection must be evaluated. Assessing customer

needs and the relevance of the proposed solution.

For this case study, sessions were made for the

capture of requirements and modeling the problem

domain. The group was formed by researchers and

stakeholders involved from the perspective of the

business. This process is recorded in class diagrams

and requirement diagrams for the domain model by

using UML.

From these diagrams, the identification of

processes that can be automated or implemented in

DataMiningfortheUniqueIdentificationofPatientsintheNationalHealthcareSystems

215

software was performed, the selection of features or

cases of use that guide the development process,

defining an architecture for the project and therefore

the implementation of the solution.

5 CONCLUSIONS AND FURTHER

RESEARCH DIRECTIONS

In this paper, the specific problem of detecting

duplicate records and Data Cleansing on the User

Record System of the DHS is presented. In addition,

positions are presented against the application of

DM to solve this problem, emphasizing the

importance of using software development

methodologies, modeling languages like UML and

analysis requirements for making the right decision

of software architecture and techniques to be applied

in the solution.

It is evident that the detection of duplicate

records is a problem of special attention for the

DHS, affecting the available economic, quality

assurance and patient care time.

A web architecture is proposed to streamline the

registration process for new members of the

healthcare system and the digitalization of the

documents that must be delivered, reducing the use

of paper and human intervention or manipulation of

information.

There are two opinions on the application of DM

to the problem of detecting duplicate records in DBs.

In the first case, there are advantages and benefits

that this technique can contribute to the problem and

secondly, the complexity and relevance that results

from its application.

In the case study it was determined the

application of algorithms for the detection of

duplicities, as Soundex, with optimization and

clustering techniques to reduce the execution and

detection times. Additionally, DM is used to analyze

the results in duplicities in order to prevent further

duplication or fraud attempts entering invalid

records system records.

Mining on the results obtained can also generate

constructive knowledge of DBs, as advanced

statistics and forecasts hedging, investment plans in

the healthcare system, affiliation programs to the

health care system, among others.

As future research directions it is important to

take into account different DM algorithms and

compare their results in this specific field of

application to evaluate their performance,

effectiveness and appropriateness of these, enabling

support implementation decisions to the solution

presented.

ACKNOWLEDGEMENTS

This research was supported by the Colombian

Institute for Development in Science and

Technology Francisco José de Caldas”

(COLCIENCIAS) through the project titled

“Desarrollo de un sistema de información para la

gestión de usuarios en la Secretaria de Salud del

Distrito de Barranquilla” No. 3351-604-37776.

REFERENCES

Batra, S., Parashar , H., Sachdeva, S., and Mehndiratta, P.

(2013). Applying data mining techniques to

standardized electronic health records for decision

support. ieeexplore, 510-515.

Bellazzi, R., and Zupan, B. (2008). Predictive data mining

in clinical medicine: current issues and guidelines.

International journal of medical informatics, 77(2),

81-97.

DANE. (2013). Departamento Administrativo Nacional de

Estadística. Obtenido de https://www.dane.gov.co/

index.php/estadisticas-sociales/encuesta-longitudinal-

de-proteccion-social.

Date, C., and Date, C. (1990). An introduction to database

systems. Addison-wesley Reading, MA.

Davila Hernandez , F., and Sanchez Corrales, Y. (2012).

Técnicas de minería de datos aplicadas al diagnóstico

de entidades clínicas. Revista Cubana de Informática

Médica, 174-183.

DeBariloche. (09 de 15 de 2014). Portal Rio negro.

Obtenido de Buscan mejorar la Tish con la base de

datos de AFIP: http://www.rionegro.com.ar/diario/

buscan-mejorar-la-tish-con-la-base-de-datos-de-afip-

4365198-9862-nota_cordillera.aspx.

Elmagarmid, A. K., Ipeirotis, P. G., and Verykios, V. S.

(2007). Duplicate record detection: A survey.

Knowledge and Data Engineering, 19(1), 1-16.

Esp, I., and Ramírez, C. (2009). Hacia una metodología

para la selección de técnicas de depuración de datos.

Rev. Av. En Sist. E Informática 6.

Han, J., and Gao, J. (2008). Research challenges for data

mining in science and engineering. Next Generation of

Data Mining. . Chapman and Hall.

Harrison, T. (07 de 2013). PRWeb. Obtenido de

http://www.prweb.com/releases/2014/08/prweb12084

827.htm.

Holzinger, A., and Jurisica, I. (2014). Knowledge

discovery and data mining in biomedical informatics:

The future is in integrative, interactive machine

learning solutions. Interactive Knowledge Discovery

and Data Mining in Biomedical Informatics, 1-18.

ICORES2015-InternationalConferenceonOperationsResearchandEnterpriseSystems

216

Hsu-Hao, T. (2012). Global data mining: An empirical

study of current trends, future forecasts and

technology diffusions. Expert Systems with

Applications, 39(9), 8172-8181. Obtenido de

http://www.sciencedirect.com/science/article/pii/S095

7417412001704.

Kaur, H., and Wasan, S. K. (2006). Empirical Study on

Applications of Data Mining Techniques in

Healthcare. Journal of Computer Science, 194-200.

Marcano Aular, Y. J., and Talavera , R. (2007). Minería de

Datos como soporte a la toma de decisiones

empresariales. Opción, 10-118.

McClellan, M. A. (2009). Duplicate medical records: a

survey of twin cities healthcare organizations. AMIA

Annual Symposium Proceedings, 421.

McCoy, A. B., Wright, A., Kahn, M. G., Shapiro, J. S.,

Bernstam, E. V., and Sittig, D. F. (2013). Matching

identifiers in electronic health records: implications for

duplicate records and patient safety. BMJ quality and

safety, 22(3), 219-224.

MinSalud. (09 de October de 2014). FOSYGA. Obtenido

de http://www.fosyga.gov.co/

Oviedo, E., and Fernández, A. (2010). Tecnologías de la

información y la comunicación en el sector salud:

oportunidades y desafíos para reducir inequidades en

América Latina y el Caribe. CEPAL.

Pagliery, J. (09 de 09 de 2014). Home Depot confirms

months-long hack. Obtenido de http://money.cnn.com/

2014/09/08/technology/security/home-depot-breach/

Shu-Hsien, L., Pei-Hui, C., and Pei-Yuan, H. (2012). Data

mining techniques and applications – A decade review

from 2000 to 2011. Expert Systems with Applications,

39(12), 11303-11311.

Teorey, T. J., Lightstone, S. S., Nadeau, T., and Jagadish,

H. V. (2011). Database modeling and design: logical

design. Elsevier.

Viveros, M., Nearhos, J., and Rothman , M. (1996).

Applying Data Mining Techniques to a Health

Insurance Information System. Proceedings of the

22th International Conference on Very Large Data

Bases (págs. 286-294). San Francisco, CA, USA:

Morgan Kaufmann Publishers Inc.

Wai Lup, L., Mong Li, L., and Tok Wang, L. (2001). A

knowledge-based approach for duplicate elimination

in data cleaning. Information Systems, 26(8), 585-606.

Washington, A. P. (10 de 23 de 2014). newspaper the

guardian. Obtenido de http://www.theguardian.com/

technology/2014/aug/23/homeland-security-25000-

employees-hacked.

Xintong, G., Hongzhi, W., Song, Y., and Hong, G. (2014).

Brief survey of crowdsourcing for data mining. Expert

Systems with Applications, 41(17), 7987-7994.

Obtenido de http://www.sciencedirect.com/science/

article/pii/S0957417414003984.

Yucatan, D. d. (17 de 09 de 2014). Diario de Yucatan.

Obtenido de http://yucatan.com.mx/merida/policia/

recibe-reconocimiento-fge-por-sistema-forense.

DataMiningfortheUniqueIdentificationofPatientsintheNationalHealthcareSystems

217