397

Views

Developing A Guideline-based Decision Support System to Diagnosis of Primary Immunodeficiency Diseases

Fateme Sepehri1, 2; Mostafa Langarizadeh3; Laleh Sharifi4; Gholamreza Azizi5; Reza Safdari2; Asghar Aghamohammadi6

1. Department of Health Information Technology, Mousavi Hospital, Zanjan University of Medical Sciences, Zanjan, Iran., 2. Department of Health Information Management, School of Allied Medical Sciences, Tehran University of Medical Sciences, Tehran, Iran., 3. Department of Health Information Management, School of Health Management and Information Science, Iran University of Medical Sciences, Tehran, Iran., 4. Uro-Oncology Research Center, Tehran University of Medical Sciences, Tehran, Iran., 5. Non-Communicable Diseases Research Center, Alborz University of Medical Sciences, Karaj, Iran., 6. Research Center for Immunodeficiencies, Pediatrics Center of Excellence, Children’s Medical Center, Tehran University of Medical Sciences, Tehran, Iran.

Correspondence: *. Corresponding author: Reza Safdari , Department of Health Information Management, School of Allied Medical Sciences, Tehran University of Medical Sciences, Tehran, Iran.


Abstract

Introduction:

Primary immunodeficiency diseases (PID) are generally rare genetic disorders affecting the immune system. Overlapping PIDs symptoms and signs is a challenge to diagnosis and treatment. On the other hand, remembering of all diagnosis criteria is difficult for practitioners. The purpose of this research is developing guideline-based clinical decision support system for diagnosis of primary immune deficiency diseases, to assist practitioner in order to diagnose of disease in early stage and to minimize complications of such diseases.

Material and Methods:

To provide data a checklist was used and most important demographic information, symptoms, family history, physical findings and laboratory findings to diagnose eight common PIDs extracted from guidelines and literature under specialists opinion. The diagnosis inference model design and develop in Protégé (version 3.4.8) frame based ontology modeling using "Noy and McGuinness" method. Then the mobile based inference model in Eclipse (SDK version 3.7.1) software has been developed and clinical decision support system of primary immunodeficiency has been created.

Results:

To design the diagnosis inference model in Protégé software, data were classified in 5 main classes and 24 subclasses as hierarchical. Then, specific properties of each class, and determine the value of each property. Then define Instances of each class and initialized instance properties. Then use this model to develop CDSS based on mobile in Eclipse software. At the end, the inference model and the CDSS test with 110 patient’s record data and both of them recognized all 110 patient correctly such as specialist recognition.

Conclusion:

Guideline-based decision support systems help to detect diseases correctly, quickly and early. Guideline-based decision support systems are very reliable to practitioner, because guidelines are accepted to their. These systems reduce the forget probability of diagnosis stages and percentage error of diagnosis by practitioner and increase the accuracy of diagnosis.

Received: 2019 March 15; Revision Received: 2019 April 25; Accepted: 2019 May 21

Iran J Med Inform. 2019 ; 8(1): e11
doi: 10.30699/fhi.v8i1.184

Keywords: Key Words Primary Immunodeficiency Diseases (PID), Inference Model, Protégé, Frame-Based Ontology, Guideline-Based Clinical Decision Support System.

INTRODUCTION

Primary immunodeficiency diseases (PID) are a genetically heterogeneous group disorders that affect distinct components of both humoral and cellular arms of the immune system [1-5]. Overlapping signs and symptoms of these diseases is a challenge for diagnosis and treatment [6-9]. Awareness of the symptoms and considering the possibility of a PID in differential diagnosis as the diagnosis key in many cases, help to rapid recognition and more appropriate treatment of patients [10, 11]. Timely recognition and treatment reduced mortality and increased lifespan and quality of life of the patients [6, 12-15]. For practitioners to memorize all effective criteria to diagnosis is difficult, so developing computerized program based on diagnosis criteria, improves significantly the quality of care [16-18].

Ontology is a tool for representation and share the knowledge of a field through modeling and create a framework of concepts and relations among them. Ontology used in many medical fields and clinical decision support systems, medical expert systems and biomedical literature data bases [19].

In this study ontology was used to develop the inference model in order to diagnosis of PIDs. Then the inference model was developed using Eclipse (SDK version 3.7.1) software and clinical decision support system of primary immunodeficiency based on mobile has been created.

MARTIAL AND METHODS

The study focused on eight common disease of primary immunodeficiency include Common Variable Immune Deficiency (CVID), X-Linked Agammaglobulinemia (Bruton’s) (XLA), Selective IgA Deficiency (SIgA), CD40L deficiency, UNG deficiency, Isolated immunoglobulin (Ig) G Subclass deficiency, Specific antibody deficiency (SAD) with normal Ig concentrations and normal numbers of B cells and Transient hypogammaglobulinemia (THI) of infancy with normal numbers of B cells. Based on clinical guidelines and medical literatures in PID [20-22], we designed a checklist to extract and classified most important signs and symptoms, family history, and laboratory data for eight main type of primary antibody deficiencies (PADs) (Table 1). To evaluate the quality of checklist, data for 10 cases in different type of PIDs were tested.

Table 1. most important signs and symptoms, family history, and laboratory data for eight selected of PIDs
Disease Symptoms Family history Signs Lab Data
1. CVID 1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
8. meningitis
9. Onset Age>4
10. Malignnancies
11. Granulomatosis
12. Vitilogo
1. Parents relative
2. male or female
1. lymph nodes: Lymphadenophathy or Normal
2. tonsil: Normal
3. Spleen: Large or Normal
4. Liver: Large or Normal
5. Clubbing: has or not
6. Auscultation of Lungs: has or not
1. IgG: low
2. Isohemagglutinin: low
3. B lymphocyte: normal
4. T lymphocyte: normal
5. Specific AB: normal or low
2. XLA 1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
8. Dermatomiositis
9. Encephalitis
10. EnteroViral
1. Parents relative or not
2. male
3. Mother grandfather or uncle have XLA
1. lymph nodes: Hasn't
2. tonsil: Atrophy
3. Spleen: small
1. IgG: low
2. IgM: low
3. IgA: low
4. Isohemagglutinin: low
5. B lymphocyte: low
6. T lymphocyte: normal or high
7. Btk Mutation
8. Specific AB: low
3. IGA 1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
8. Onset Age>4
9. Allergies
10. Anaphylaxis
1. Parents relative
2. male or female
1. lymph nodes: Normal
2. tonsil: Normal
3. Spleen: Normal or Large
4. Liver: Normal or Large
1. IgG: normal
2. IgM: normal
3. IgA: low
4. B lymphocyte: normal
5. T lymphocyte: normal
6. Specific AB: normal
4. HIGM
CD40ligand
1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
8. Pneumcystis Serovici Pneumonia
1. Parents not relative
2. male
3. Mother grandfather or uncle have HIGM CD 40ligand
1. lymph nodes: Hasn't
2. tonsil: Atrophy
3. Spleen: small
1. IgG: low
2. IgM: normal or high
3. IgA: low
4. Isohemagglutinin: low
5. Specific AB: low
5. HIGM
UNG, AID
1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
1. Parents relative
2. male or female
1. lymph nodes: Hypertrophy
2. tonsil: Hypertrophy
3. Spleen: Large
4. Liver: Large
1. IgG: low
2. IgM: normal or high
3. IgA: low
4. Isohemagglutinin: low
5. Specific AB: low
6. IGsub class 1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
1. Parents relative
2. male or female
1. lymph nodes: Normal
2. tonsil: Normal
3. Spleen: Normal
4. Liver: Normal
1. IgG: normal
2. IgM: normal
3. IgA: normal
4. IgG1:low
5. IgG2: low
6. IgG3: low
7. IgG4: normal
8. B lymphocyte: normal
9. T lymphocyte: normal
10. Isohemagglutinin: normal
11. Specific AB: normal or low
7. SAD 1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
1. Parents relative
2. male or female
1. lymph nodes: Normal
2. tonsil: Normal
3. Spleen: Normal
4. Liver: Normal
1. IgG: normal
2. IgM: normal
3. IgA: normal
4. Isohemagglutinin: low
5. B lymphocyte: normal
6. T lymphocyte: normal
7. Antipneumococys AB: low
8. THI 1. Recurrent infections
2. Pneumonia
3. Acute Otitis media
4. Auto immunity
5. Diarrhea
6. Osteomyelitis
7. Sinusitis
8. Onset Age < 5
1. Parents relative or not
2. male or female
1. lymph nodes: Normal
2. tonsil: Normal
3. Spleen: Normal
4. Liver: Normal
1. IgG: low
2. IgM: normal
3. IgA: normal
4. B lymphocyte: normal
5. T lymphocyte: normal
6. Isohemagglutinin: normal
7. Specific AB: normal
8. IgG1, IgG2, IgG3: normal

Frame based ontology modeling (to create the inference model) and "Noy and McGuinness" method was used to develop the inference model. Noy and McGuinness method includes seven stages as followings:

1. Determine the domain and scope of the ontology

2. Consider reusing existing ontologies

3. Enumerate important terms in the ontology

4. Define the classes and the class hierarchy

5. Define the properties of classes-slots

6. Define the facets of the slots

7. Create instances [23].

Determine the domain and scope of the ontology

The aim of this study is developing inference model that could help physicians in diagnose of eight selected disease according to the signs and symptoms of patients. The study area is eight primary immunodeficiency diseases of and its signs and symptoms.

Consider reusing existing ontologies

Today, rarely an ontology is developed from scratch [24]. In this study, expert’s knowledge is used to develop the ontology.

Enumerate important terms in the ontology

Make a list of related terms that are expected to be in ontology is the first step of ontology definition [24]. In this study, a list of related terms including eight primary immunodeficiency diseases and their symptoms, signs and lab data to diagnosis were extracted from clinical guidelines and medical literature under the immunodeficiency specialist supervision.

Define the classes and the class hierarchy

After identifying related terms, such terms should be classified as a hierarchy [24]. Classes define a set of instances that have common features [25]. In this study, based on this definition, classes was determined and entered in classes tab of protégé software (Fig 1). In Fig 1 ANTIBODY_PRODUCTION_DEFECTS class is the main class that contains eight PID diseases as instances.


[Figure ID: F1] Fig 1. Classes and the class hirarcy of inference model

Define the properties of classes-slots

Properties are links and relationships between classes [19, 24]. Object type Property is used for the relationship between instances (individuals) of two class. Data type Property define data type values for instances (individuals) of a class [25]. Properties (slots) of ANTIBODY_PRODUCTION_DEFECTS class is shown in Fig 1. For example, LabDatas_Info slot is an Object type Property that made instances of LabDatas class usable in instances of ANTIBODY _PRODUCTION_DEFECTS class. OnsetAge slot is a Data type Property with value type of integer.

Define the facets of the slots

For properties (slots) can define facets. Slots can have different facets describing the value type, allowed values, the number of the values (cardinality), and other features of the values the slot can take [23]. For example, the OnsetAge slot can have multiple values because some of the primary immunodeficiency diseases occurs in particular time range. The slot Symptoms_Info can have multiple values and the values are instances of the class Symptoms.

Create instances

Instances are considered as members of a class [25-28]. An ontology with individual instances of classes constitute a knowledge base [23]. Actually ontology is used to organize the instances [24]. Defining an individual instance of a class requires 1) choosing a class, 2) creating an individual instance of that class, and 3) filling in the slot values [19, 23]. For example, the eight selected primary immunodeficiency diseases are instances of ANTIBODY_PRODUCTION_DEFECTS class with them slots such as symptoms, family history, signs and laboratory finding.

Whereas the model is not useable for all, After developing of diagnosis inference model in Protégé (version 3.4.8) frame based ontology modeling using "Noy and McGuinness" method, develop the inference model in Eclipse (SDK version 3.7.1) software and development mobile application of Clinical decision support system of primary immunodeficiency (Android 2.1) that physicians can use this application.

RESULTS

The diagnosis inference model developed in 5 main class and 24 subclass as hierarchical. Eight primary immunodeficiency diseases as instances of ANTIBODY_PRODUCTION_DEFECTS class, with their properties that is their symptoms, signs, family history and laboratory data, consist the knowledge base, so question and answer done on ANTIBODY_PRODUCTION_DEFECTS class. To execute the inference model use the query tab of protégé software. The symptoms and signs and laboratory data is entered as input and the diseases that have related symptoms is returned as output by model. Whereas this model is developed with deterministic data and information and based on knowledge, the model responds correctly to the question that related knowledge exist in the knowledge base and did not return any answer where the related knowledge not exist in knowledge base. In Fig 2 a query of the inference model is shown.


[Figure ID: F2] Fig 2. Execute the inference model in query tab by entering symptoms, signs and laboratory data

At the end, evaluate functionality of the inference model and application to investigate that could them correctly analyzed and recognized eight PID diseases correctly or not (Fig 3 and 4). Both inference model and mobile application were tested with 10 patient’s data. Results showed in Table 2.


[Figure ID: F3] Fig 3. Execute the mobile application of CDSS by entering symptoms, signs and laboratory data

[Figure ID: F4] Fig 4. Result of executemobile application of CDSS
Table 2. Inference Model and Software Test Results with Ten Patients Case
Patient ID 1 2 3 4 5
Patient Symptoms 1. age: 2
2. Acute Otitis media
3. IgG : Low
4. SpecificAB : Normal
1. age: 3
2. male
3. Recurrent infections
4. tonsil : small
5. IgG : Low
6. BCell : Low
1. age:6
2. Pneumonia 3. IgG : Normal
4. IgM : Normal
5. IgA : Low
1. age: 7
2. Pneumonia
3. Diarrhea
4. IgG : Low
5. BCell : Normal
1. age: 5
2. male
3. Recurrent infections
4. tonsil : small
5. IgG : Low
6. IgA : Low
7. IgM : Normal
Diagnosis by model and System THI XLA IgA CVID CD40Ligand
Diagnosis by specialist Correct Correct Correct Correct Correct
Patient ID 6 7 8 9 10
Patient Symptoms 1. age: 6
2. Pneumonia
3. IgG : Normal
4. IgA : Normal
5. IgM : Normal
6. BCell : Normal
1. age: 8
2. Pneumonia
3. IgG : Normal
4. IgA : Normal
5. IgM : Normal
6. IgG2: low
7. BCell: Normal
1. age: 4
2. female
3. Pneumonia
3. IgG : Low
4. IgA : Normal
5. IgM : Normal
6. BCell : Normal
7. sub class IgG : Normal
8. SpecificAB: Normal
1. age: 7
2. Pneumonia
3. IgG: Low
4. IgA: Low
5. IgM: high
6. Specific AB: Low
7. Diarrhea
1. age: 4
2. male
3. Pneumonia
4. IgG : Normal
5. IgA : Normal
6. IgM : Normal
7. BCell : Normal
8. IgG3: low
Diagnosis by model and System SAD, IgSubClass IgSubClass THI UNG-AID, CD40Ligand IgSubClass
Diagnosis by specialist Correct Correct Correct Correct Correct

DISCUSSION

in this study data is limit and follow the Closed World Assumption (in the closed world assumption is true if its negation is not provable [24] and terms should be considered unique, so the best method is frame-based modeling that used in artificial intelligence, expert systems and knowledge bases [29].

Eslami and et al. developed a computerized program designed to diagnose primary headache based on international classification of headache disorders, 2nd edition (ICHD-II) criteria for use by physicians. Of 80 patients, the software was able to come up with correct results in 78 cases. In other words, this software permitted the diagnosis of more than 97% of the patients similar to the physician's [16].

Yin and et al. proposed a computerized headache guideline method using SAGE module and developed a decision support system for headache diagnosis. 282 previously diagnosed cases from EMR were used to evaluate the diagnostic accuracy of the system, the result is: migraine 94.1%, tension-type headache 89%, cluster headache 90.9% and chronic daily headache 93.0% [17].

Dong and et al. developed a headache CDSS based on ICHD-3 beta and validated it in a prospective study that included 543 headache patients. The CDSS correctly recognized 99.4% of migraine without aura, 100% of migraine with aura, 95.2% of chronic migraine, and 62.7% of probable migraine. This system also correctly identified 87.2% of patients with tension-type headache (TTH), which infrequent episodic TTH was diagnosed in 92.3%, frequent episodic TTH was diagnosed in 98.0%, chronic TTH in 90.0%, and probable TTH in 60.9%. The correct diagnostic rates of cluster headache and new daily persistent headache (NDPH) were 90.0% and 100%, respectively. In addition, the system recognized 100% of patients with medication overuse headache [30].

CONCLUSION

In this study and previous studies, software's are developing with deterministic data, they recognized correctly with high accuracy. Also integrating clinical guidelines and diagnostic criteria with information systems and electronic medical records, increase following clinical guideline, improve the quality of patient care, ensure patient safety and reduce costs.


References
1. Mohammadinejad, P. Pourhamdi, S. Abolhassani, H. Mirminachi, B. Havaei, A. Masoom, SN. Primary antibody deficiency in a tertiary referral hospital: A 30-year experiment. J Investig Allergol Clin Immunol 2015 25(6):416–25. [PubMed]
2. Mohammadinejad, P. Mirminachi, B. Sadeghi, B. Movahedi, M. Gharagozlou, M. Mohammadi, J. Distribution of primary immunodeficiency disorders diagnosed in a tertiary referral center, Tehran, Iran (2006-2013). Iran J Immunol 2014 11(4):282–91. [PubMed]
3. Aghamohammadi, A. Mohammadinejad, P. Abolhassani, H. Mirminachi, B. Movahedi, M. Gharagozlou, M. Primary immunodeficiency disorders in Iran: update and new insights from the third report of the national registry. J clin immuno 2014 34(4):478–90. [CrossRef]
4. Abolhassani, H. Aghamohammadi, A. Pourjabbar, S. Salehi, SM. Nikayin, S. Rabiee, A. Psychiatric aspects of primary immunodeficiency diseases: The parental study. Iran J Allergy Asthma Immunol 2013 12(2):176–81. [PubMed]
5. Mir, SGB. Aghamohammadi, A. Kouhi, A. Farhoudi, A. Moin, M. Rezaei, N. Mortality in primary immunodeficient patients, registered in Iranian primary immunodeficiency registry. Iran J Allergy Asthma Immunol 2004 3(1):31–6. [PubMed]
6. Nabavi, M. Arshi, S. Bemanian, MH. Aghamohammadi, A. Mansouri, D. Hedayat, M. Long-term follow-up of ninety eight Iranian patients with primary immune deficiency in a single tertiary centre. Allergol Immunopathol (Madr) 2016 44(4):322–30. [PubMed] [CrossRef]
7. Mohammadinejad, P. Ataeinia, B. Kaynejad, K. Zeinoddini, A. Sadeghi, B. Hosseini, M. Antibiotic resistance in patients with primary immunodeficiency disorders versus immunocompetent patients. Expert Rev Clin Immunol 2015 11(10):1163–72. [PubMed] [CrossRef]
8. Aghamohammadi, A. Moghaddam, ZG. Abolhassani, H. Hallaji, Z. Mortazavi, H. Pourhamdi, S. Investigation of underlying primary immunodeficiencies in patients with severe atopic dermatitis. Allergol Immunopathol (Madr) 2014 42(4):336–41. [PubMed] [CrossRef]
9. Rezaei, N. Mohammadinejad, P. Aghamohammadi, A. The demographics of primary immunodeficiency diseases across the unique ethnic groups in Iran, and approaches to diagnosis and treatment. Ann N Y Acad Sci 2011 1238:24–32. [PubMed] [CrossRef]
10. Abolhassani, H. Mirminachi, B. Daryabeigi, M. Agharahimi, Z. Aghamohammadi, A. Rabbani, A. Evaluation of physicians' awareness of pediatric diseases in Iran. Iran J Pediatr 2014 24(1):87–92. [PubMed]
11. Azizi, G. Abolhassani, H. Asgardoon, MH. Shaghaghi, S. Negahdari, B. Mohammadi, J. Managing patients with side effects and adverse events to immunoglobulin therapy. Expert Rev Clin Pharmacol 2016 9(1):91–102. [PubMed] [CrossRef]
12. Abolhassani, H. Rezaei, N. Mohammadinejad, P. Mirminachi, B. Hammarstrom, L. Aghamohammadi, A. Important differences in the diagnostic spectrum of primary immunodeficiency in adults versus children. Expert Rev Clin Immunol 2015 11(2):289–302. [PubMed] [CrossRef]
13. Abolhassani, H. Akbari, F. Mirminachi, B. Bazregari, S. Hedayat, E. Rezaei, N. Morbidity and mortality of Iranian patients with hyper IgM syndrome: a clinical analysis. Iran J Immunol 2014 11(2):123–33. [PubMed]
14. Abolhassani, H. Hirbod-Mobarakeh, A. Shahinpour, S. Panahi, M. Mohammadinejad, P. Mirminachi, B. Mortality and morbidity in patients with X-linked agammaglobulinaemia. Allergol Immunopathol (Madr) 2015 43(1):62–6. [PubMed] [CrossRef]
15. Tavakol, M. Mohammadinejad, P. Baiardini, I. Braido, F. Gharagozlou, M. Aghamohammadi, A. The persian version of the chronic urticaria quality of life questionnaire: factor analysis, validation, and initial clinical findings. Iran J Allergy Asthma Immunol 2014 13(4):278–85. [PubMed]
16. Eslami, V. Rouhani-Esfahani, S. Hafezi-Nejad, N. Refaeian, F. Abdi, S. Togha, M. A computerized expert system for diagnosing primary headache based on International Classification of Headache Disorder (ICHD-II). SpringerPlus 2013 2(1):199–202. [PubMed] [CrossRef]
17. Yin, Z. Dong, Z. Yu, S. Lu, X. Feng, G. Duan, H. A guideline-based decision support system for headache diagnosis. Stud Health Technol Inform 2013 192:1022–3. [PubMed]
18. Maizels, M. Wolfe, WJ. An expert system for headache diagnosis: The Computerized Headache Assessment tool (CHAT). Headache 2008 48(1):72–8. [PubMed] [CrossRef]
19. Zahedi, R.; Alibeik, M.; Amin, G.; Karimi, M. Developement of ontology representation of medicinal plants in Iran based on Unified Medical Language System (UMLS). [Thesis]. Tehran University of Medical Sciences: 2013.
20. Bonilla, FA. Khan, DA. Ballas, ZK. Chinen, J. Frank, MM. Hsu, JT. Practice parameter for the diagnosis and management of primary immunodeficiency. J Allergy Clin immunol 2015 136(5):1186–205. [PubMed] [CrossRef]
21. de, VE. Patient-centred screening for primary immunodeficiency, a multi-stage diagnostic protocol designed for non-immunologists: 2011 update. Clin Exp Immunol 2012 167(1):108–19. [PubMed] [CrossRef]
22. Bousfiha, AA. Jeddane, L. Ailal, F. Al, HW. Conley, ME. Cunningham-Rundles, C. A phenotypic approach for IUIS PID classification and diagnosis: guidelines for clinicians at the bedside. J Clin Immunol 2013 33(6):1078–87. [PubMed] [CrossRef]
23. Noy, NF. McGuinness, DL. Ontology development 101: A guide to creating your first ontology. Stanford knowledge systems laboratory technical report [Report No: KSL-01-05]; 2001
24. Shadgar, B.; Osare, A.; Haratian, NA. Semantic Web- Concepts and techniques. Tehtan: Armaghan; 2014.
25. Sanat, JA. Fathian, A. The methodology for design, construction and implementation of ontology: Approaches, languages and tools (Case study in the field of library and information science ASFAONT design ontology). Journal of Librarianship and Information Sciences 2012 15(1):113–142.
26. Asooshe, A.; Mehrasa, Sh; Khatibi, T.; Khakshour, SP. Ontology: Methodology, tools and development languages. Tehran: Tarbyat Modares University Publication Center; 2011.
27. Ghaffari, J. Karami, H. Khanian, A. Mohammadzadeh, E. Primary immuno-deficiencies (PID). J Mazandaran Univ Med Sci 2009 19(70)
28. Chen, CC. Chen, K. Hsu, CY. Li, YC. Developing guideline-based decision support systems using protégé and jess. Comput Methods Programs Biomed 2011 102(3):288–94. [PubMed] [CrossRef]
29. Behkamal, B. Semantic Web Tutorial. Ferdowsi University: Web Technology Lab; 2012.
30. Dong, Z. Yin, Z. He, M. Chen, X. Lv, X. Yu, S. Validation of a guideline-based decision support system for the diagnosis of primary headache disorders based on ICHD-3 beta. J Headache Pain 2014 15(1):40–7. [PubMed] [CrossRef]

Refbacks

  • There are currently no refbacks.