E-Medicine Development Multi-Agent with Integrated Approach

Reader Impact Factor Score
[Total: 2 Average: 5]

Published on International Journal of Health, Nursing, & Medicine
Publication Date: July 19, 2019

Sunaryo Adi Pranoto, Eka Surahman & Damanik Wahyuni
Graha Edukasi Health Sciences College, Makassar
Nani Hasanuddin College of Health Sciences, Makassar
Lakipadada College of Health Sciences, Tana Toraja
South Sulawesi, Indonesia

Journal Full Text PDF: E-Medicine Development Multi-Agent with Integrated Approach.

E-medicinethe digitization of processes and medical service. Multi-agent is an approach used for the development of e-medicine. Introduced multi-agent approach to build e-medicine system, so the design of multi-agent will be done for the development of e-medicine plan. Next will be explained about the development of e-Medicine with Multi agent integrated approach, by taking the case of telemedicine for diabetes.

Keywords: E-medicine, telemedicine, multi-agent, computer-based patient record, knowledge models.

There are some other known term in the development of e-medicine, namely telemedicine, e-healthcare, e-diagnosis, e-consultation, e-clinic and computer-based patient record (CPR). Telemedicinethe use of communications technology in the area of medical care and treatment. While e-healthcare is one of the activities at the clinic, which specializes in health knowledge and evaluation of health status. E-diagnostics to analyze and determine the health status of the patient, based on medical information. E-consultation particular as giving health consultations and dissemination of health knowledge to the community. E-specialty clinic dealing with diagnosis and treatment to patients of certain disease. Computer-based patient record (CPR) is an electronic data management of health care services to patients.
Multi-agent approach is very effective to deal with the complexities of e-medicine. An agent is a computer system that is able to perform an action freely or stand alone. Multi-agent approach consists of several steps according to Jiang [1]:
a. Identification of the basic rules of each Agent.
b. Identification of the responsibilities of each Agent.
c. Determine the objectives and implementation plans.
d. Determining the structure of e-medicine system.
Step above, can be grouped into two major lines, first, the design agent, second, the design agent society [1].

Applications on the design agent is to identify the agent rules and responsibilities and services in e-medicine system. such applications include: interface agent, broker agent, doctor agent, administration agent, agent controller, monitoring agents, diagnostic agents, consultation agent, traning agents, database wrapper agents, education agents, decision support agent [1].
interface agentdisplays instructions for the implementation of e-medicine system. Broker agent is an agent who knows all about the ability of multi-agent systems. Doctor agent tasked with monitoring the meeting schedule given by the doctors, and the ability of doctors to visit patients. Administration agent implements medical administration. Controller agent controls the entire system of e-medicine. Department agent had knowledge of the medical department, and manage medical problems in departement.Tranning agent displays the instruction given to the patient, such as how to get drugs, or when I can get treatment.
Database wrapper agentsis an agent that controls access to the database contains the medical records of a patient. Education agents to introduce the latest medical technology, as well as in terms of studying it. Decision support agent integrate knowledge and diagnosis, resulting in an effective decision.

Design of multi-agent society focus on the establishment of multi-agent system architecture and the interaction between agents. The model of interaction in multi-agent systems are divided into external model and the internal model. Internal interaction model of e-medicine system not only between the agent with the department, but also with several different departments. External interaction models, for example medical instrument, support scientific fields of psychology, the role of universities and medical institutions, and so forth.

To illustrate the multi-agent approach in the development of e-medicine system. The case study on telemedicine diabetes.
Diabetes is a chronic disease. As a result of this disease is the body’s metabolism can not work well, due to reduced insulin in the body. Therefore patients with diabetes need injections of insulin. In addition, patients should also pay attention to the health of her everyday life, such as measuring blood glucose levels, and so forth. By using telemedicine systems to manage the process of treatment of this disease, it can quickly determine the condition of the patient every time, and evaluate the therapy done. For information telemedicine system has been studied since the 1980s.

Telemedicine system should be able to monitor and services at any time to the patient. Among other things, [1]:
• Visits to patients and the implementation of personalized therapy
• The monitoring of the patient’s condition each particular time and monitoring data.
• Diagnosing a patient, based on the results of health monitoring and patient data.
• Do counseling for patients with diabetes, in order to understand the important things that must be considered in relation to the disease.
• Build a good database system to manage patient data.
• Counseling for patients with diabetes.
•Managing e-medicine interactions with other related systems.

In diabetes telemedicine system, is a medical service that is included in it is monitoring the patient’s condition every time on an ongoing basis, the delivery of information relating to the patient to those in need, held a communication forum on the development of therapies that do, and has opened counseling for patients. Based on the services mentioned above, it can be defined several agents, as follows: monitoring agent, the data processing agent, agent consultation, traning agent, archical agent, department agent and agent interface. The responsibility of the agent, [1] as follows:
monitoring agent, Responsible for monitoring the health of patients any given moment, as well as transmit data from patient monitoring to the data processing agent. Data processing agent, make statisitk and data integration monitoring results. Diagnosis agent, analyze the situation and condition of the patient, and can determine important matters related to certain diseases. Agent therapy, determine the method of treatment according to the patient’s condition.
Consultation agent, Organized counseling for patients, related to the diagnosis results of diagnostic agent. Decision support agent, to provide support to decision-makers, as well as cooperate with the diagnosis agent. Traning agent, to train the patient to understand the essentials of his illness, as well as how to maintain good condition can be returned. Archival agent, do menambahan or changes to patient data and therapy method that is being executed. Then integrate the individual patient database. Department agent, implement control on the course of the telemedicine system. Interface agent, helping data and information search services.

In this case example, telemedicine services implemented on the diabetes department at a hospital. Telemedicine system above, includes not only health care through remote monitoring, diagnosis, therapy and consultation, but also intergasi with other parts of the e-medicine system, such as education, training, management, security and database.
How to set goals or targets for multi-agent systems. Multi-agent system consists of three groups of interest, namely the interface group, implementation group and the control group. In the implementation group, there are some agents that have different responsibilities.
Below, is shown the architecture of multi-agent systems.

Figure 2. Multi-Agent System Architecture of Telemedical

In the multi-agent system, the control group had a stint as a mediator, in the event of conflict between agent. While the interface group, keeping in touch with the patient’s agent and a cross-section of the e-medicine system. Implementation group to implement the monitoring, diagnosis, therapy, counseling, and to ensure that objectives are met.

First, it must first know the information needed for planning and achieving goals in the interaction among agents. In a multi-agent system, external relations focus on integration efforts portions of e-medicine and the environment. While the focus on the internal relations of cooperation between the occurrence of the interaction agent to realize it.

Below is an overview the interaction that occurs in multiparty systems in telemedicine diabetic agent.

Arrows 1 and 2 illustrate that the telemedicine environment outside the e-medicine and other related issues. Arrow 3 and 4 illustrate the implementation group to interact with the control group and interfaces group, the telemedicine system. Arrow 5, represents the interaction between the agent and control department in e-medicine group, such as the interaction between administrative departments and the control group. Arrows 6, represents the interaction between the agent interface and interphase of e-medicine system, such as doctor agents, personal agent and security agent. In the implementation group, the diagnosis agent plays an important role, because they diagnosis is a complex process, the diagnosis agent not only interact with the agent in the implementation group, but also interact with the group decision, clinic group, education agents,
Arrows 8 and 17, representing the internal interactions between agents in the implementation group on a telemedicine system, in accordance with the matrix tables below. Furthermore, traning agent implements the methods of the treatment agent, it is indicated in the arrow 19.
Below is a table matrix, which describes the interactions that occur between diabetic agent in telemedicine systems.

Table 1. Matrix Interaksi Mulit Agent
Monitoring Agent Monitoring Agent Data Processing Agent diagnosis Agent Agent Therapy Archival Agent
Monitoring Agent 1.2 1.3 1.4 1.5
Data Processing Agent 2.1 2.3 2.4 2.5
diagnosis Agent 3.1 3.2 3.4 3.5
Agent Therapy 4.1 4.2 4.3 4.5
Archival Agent 5.1 5.2 5.3 5.4

Entry <1.5> to show that the monitoring agent sends data to the monitoring results of the archival agent, if needed. The <5.1> to show that knowledge of the monitoring agent obtained from archival agent.
Entry <2.3> to show that the data processing integrated agent will provide the data to the diagnostic agent for the diagnosis needs. While <3.2> to show that a diagnosis agent will ask for specific data, if it considered the data available to him is not complete, so be reprocessed by the data processing agent.
Entry <2.4> to show that the data processing agent will send the data set to therapy agent, if necessary. While <4.2> to show that therapy agent will request certain data reprocessed, if considered no less valid data.
Entry <2.5> to show that the data processing agent sends data to archival agent, if needed. And <5.2> to show that the archival agent will probably ask for some data, to be integrated by the data processing agent.
Entry <3.4> to show that a diagnosis agent will report the results of diagnosis, to use as the basis for the implementation of the therapy by therapy agent. And <4.3> to show that therapy agent might ask for a diagnosis agent, in order to complete the results of the diagnosis, because it is not yet complete.
Entry <4.5> to show that therapy agent may ask for an explanation of the relationships among the data. And <5.4> represents that the archival agent will ask the agent to complete the data therapy treatment.
Each agent has a way to represent knowledge simply by a computer-based.

Can be displayed practical approach in making patient data storage architecture. Computer-based patient data storage, has the ability to share information in relation to a particular situation. So that needs to be added to the knowledge of the work process in the clinic and activities, and issues relating to the activities mentioned above, which must be solved, and what information is needed for each activity.
To overcome the problem of storage of patient data, then there are two frameworks for representing knowledge in e-medicine. The two frameworks, namely, real layer based on experience (facts layer), and a layer of knowledge (knowledge layer).
In the layer of knowledge, there are two models
a. Knowledge Model
Knowledge models of the storage system, both its content and its use, which supports a storage system to recognize the relationship and linkage of data with each other.
b. Relevance Model
Relevance models as the ability to share information based on relevance in a given situation.

Shown below, illustrates the relationship between the model and the relevance of knowledge models.

There are three ontologies at the knowledge model, that process, content, patient.
a. Process ontology
Having understanding of the work flow processes, activities, and interactions that occur with patient data penyipanan system.
b. Content ontology
Content ontology linking concepts in process ontology with concepts in patient records ontology.
c. Patient record (Information ontology)
Has a role to present information about the patient.
While the relevance models share the information presented by the information ontology, based on criteria of relevance / linkages. The final result contains the results of the sharing of information, which is derived from the data storage, according to the relevance of the data requirements of the conditions and the situation at hand.

The advantage of an integrated health system development, namely medical error, outcomes management and disease management, and improved service delivery and patient care and lower costs. The explanation is as follows: Carroll-Barefield [2].

a. Medical Error
In some print media, often we read, there has been a medical error or medical error. This happens because of lack of success of medical employees, to determine / apply the method of treatment is appropriate, to patients based on presenting symptoms and medical history.
The integrated system will be able to reduce or eliminate medical errors that occur. There are several things that can be administered in the presence of an integrated system:
• gather the complete information about the patient, including allergies and medical action ever experienced, as well as the history of diagnosis and reports from the laboratory of him.
• Continuous updating of information about drug use, as well as the prohibition.
• Further clarify the orders given to the patient, such as writing the name of the drug, dose and others.
• And so forth.
b. Outcomes Management and Disease Management
Outcomes Management, That the data resulting from the process is used to analyze the relationship / connection between the medical authorities to fields that exist in the clinic itself, include financial and procurement of technology to support the activities.
The purpose of outcomes management:
• identifying opportunities to improve services to patients.
• Regularly documented development performed by the clinic.
Disease managementis one part of the management outcomes. Which aims to approach to treatment of diseases and conditions, effectively and efficiently.
c. Improvement and Maintenance Service to Patients and Cost Reduction
Lack of good care and service to patients, posing a health risk and financial risk. Due to the reduction in medical error, and the application of appropriate technology, will improve services and care for patients and reduce costs. With the integrated system can support “The right information on the right place and the right time and the right people.”

Multi-agent system can not only integrate all the knowledge and experience of clinical care and support for decision making, but also able to adapt to the changing environment a hospital or clinic. So multi-agent approach is effective, it can overcome the problems of the medical system.

[1] Jiang Tian, and Huaglory, Tianfield, (2013), A Multi-agent Approach to the Design of an E-medicine System, Journal from School of Computing and Mathematical Sciences Glasgow Caledonian University.
[2] Carroll-Barefield, Amanda, and Smith, Sherry P., and Prince, Lori H., and Condon, Jim (2011), Integrated Practice Management Systems, Journal from the Department of Health Information Management, Medical College of Georgia.