Integrirani dijagnostički postupak za posttraumatski stresni poremećaj [Integrative diagnostic algorithm for posttraumatic stress disorder]

Marinić, Igor (2014) Integrirani dijagnostički postupak za posttraumatski stresni poremećaj [Integrative diagnostic algorithm for posttraumatic stress disorder]. PhD thesis, Sveučilište u Zagrebu.

[img]
Preview
PDF
Download (1MB) | Preview

Abstract

The diagnosis of post-traumatic stress disorder (PTSD) is a complex process, which is reflected in clinical and forensic work and research. Because of the complexity of diagnosing PTSD, there is a need for additional and more comprehensive diagnostic models. The intention of this work was to use data mining methods for integration of data related to psychiatric patients with PTSD (personal history, data from psychiatric and psychological questionnaires and evaluation), to select significant variables and rules, and form an algorithm that would contribute to improvement and objectivization of diagnostic procedure. After performing the data analysis, we have extracted significant variables and rules that characterize patients with PTSD, conducted regression analysis and then formed diagnostic algorithm for PTSD. This is the first study of this type in psychiatry, using data mining methods, which is focusing on the integration of a broad set of data and formation of models for diagnosing PTSD. Of importance for clinical practice is that the variables that are found significant may help in the diagnostic process in patients with PTSD. In addition, we have developed a system that allows the collection of a large number of variables and their structuring in a way that is suitable for further processing using data mining methods, which can serve as a basis for the development of similar models for other psychiatric disorders.

Abstract in Croatian

Dijagnosticiranje posttraumatskog stresnog poremećaja (PTSP) je složen postupak, što se očituje u kliničkom i forenzičkom radu te istraživanjima. Zbog složenosti dijagnosticiranja PTSP-a postoji potreba za formiranjem dodatnih i sveobuhvatnijih dijagnostičkih modela. Namjera rada je bila da se primjenom postupaka inteligentne analize podataka (eng. data mining) ostvari integracija podataka vezanih uz psihijatrijske bolesnike s PTSP-om (anamnestički podaci, podaci iz psihijatrijskih i psihologijskih upitnika i obrade), da se iz tih podataka izdvoje značajne varijable i pravila te formira postupnik koji bi doprinjeo unaprijeđenju i objektivizaciji dijagnostičkog postupka. Provedenom analizom podataka izdvojene su značajne varijable i pravila koji karakteriziraju bolesnike s PTSP-om, provedena je regresijska analiza rezultata i zatim je korištenjem značajnih varijabli i pravila formiran dijagnostički postupnik za PTSP. Ovo je prvo istraživanje ovog tipa u psihijatriji, korištenjem metoda inteligentne analize, usmjereno na integraciju širokog skupa podataka i formiranje modela za dijagnosticiranje PTSP-a. Od značaja za kliničku praksu je to što su izdvojene varijable koje mogu pomoći u dijagnostičkom procesu kod bolesnika sa PTSP-om. Uz to, razvijen je sustav koji omogućuje prikupljanje velikog broja varijabli ispitanika i njihovo strukturiranje na način koji je pogodan za daljnju obradu korištenjem metoda inteligentne analize podataka, a koji može poslužiti kao osnova za izradu sličnih modela u psihijatriji.

Item Type: Thesis (PhD)
Mentors:
Mentor
Kozarić-Kovačić, Dragica
Departments: Izvan medicinskog fakulteta
Depositing User: Marijan Šember
University: Sveučilište u Zagrebu
Institution: Medicinski fakultet
Number of Pages: 188
Status: Unpublished
Creators:
CreatorsEmail
Marinić, IgorUNSPECIFIED
Date: 9 September 2014
Date Deposited: 13 Oct 2014 10:17
Last Modified: 13 Oct 2014 10:17
Subjects: /
Related URLs:
    URI: http://medlib.mef.hr/id/eprint/2162

    Actions (login required)

    View Item View Item

    Downloads

    Downloads per month over past year