Detekcija leukemije primjenom metoda digitalne obrade slike

undergraduate thesis

• Title (english): Leukemia detection using digital image processing methods
• Author: Mia Gmiza
• Mentor: Sonja Grgić (mentor)
• Committee member: Sonja Grgić (predsjednik povjerenstva)
• Committee member: Jelena Božek (član povjerenstva)
• Committee member: Ivana Žeger (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2023-09-14, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Rana dijagnoza leukemije ključna je u ozdravljenju osoba oboljelih od ove teške bolesti. Zbog toga se razvijaju brojne automatske metode kod detekcije. Ovaj rad proučavao je mikroskopske slike iz ALL-IDB baze. To su slike mikroskopskih uzoraka krvi ljudi oboljelih od akutne limfocitne leukemije. Proučen je postupak digitalne obrade slike za izdvajanje limfocitnih jezgri iz znanstvenog rada „Diagnosis of Leukemia and its types Using Digital Image Processing Techniques“ u literaturi označen pod [2]. Primijenile su se metode digitalne obrade slike spomenute u navedenom radu te su se za slike iz ALL-IDB2 baze izračunale referentne vrijednosti okruglosti nezrelih i zrelih limfocita, odnosno njihovih jezgri. Uočeno je da se prema vrijednostima okruglosti ne može potpuno odrediti radi li se o abnormalnom limfocitu. Zbog toga su uvedeni dodatni kriteriji kod detekcije, kao što su vrijednosti površine limfocitnih jezgri te omjer ukupnog broja limfocita i eritrocita na mikroskopskoj slici. Također je kod nekih slika uočena nepotpuna segmentacija limfocita te su te vrijednosti zanemarene dodatnim kriterijem kod same detekcije. Postupak detekcije je u 66.67% slučajeva uspješno detektirao leukemiju.

Abstract (english)

Early diagnosis of leukemia is crucial in the recovery of people suffering from this serious disease. This is why numerous automatic methods are being developed for the detection of this disease. This paper studied microscopic images from the ALL-IDB database. These are pictures of microscopic blood samples from people suffering from acute lymphocytic leukemia. The procedure of digital image processing for extracting lymphocyte nuclei from the scientific work "Diagnosis of Leukemia and its types Using Digital Image Processing Techniques" in the literature marked under [2] was studied. The methods of digital image processing mentioned in the mentioned paper were applied, and reference values of the roundness of immature and mature lymphocytes, i.e. their nuclei, were calculated for the images from the ALL-IDB2 database. It was observed that according to the roundness values it is not possible to completely determine whether it is an abnormal lymphocyte. For this reason, additional detection criteria were introduced, such as values of the are of lymphocyte nuclei and the ratio of the total number of lymphocytes and erythrocytes on the microscopic image. Incomplete segmentation of lymphocytes was also observed in some images, and these values were ignored by an additional criterion in the detection itself. The detection procedure successfully detected leukemia in 66.67% of cases.

• Keywords: leukemija; limfociti; digitalna obrada slike; okruglost; izdvajanje jezgri
• Keywords (english): leukemia; lymphocytes; digital image processing; roundness; nuclei extraction
• Language: croatian
• URN:NBN: urn:nbn:hr:168:435721
• Study programme: Title: Computing; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-10-11 08:53:51