Enhanced ensemble-based object detection in medical image analysis

A tétel rövid nézete
dc.contributor.advisorHajdu, Lajos
dc.contributor.authorKovács, László
dc.contributor.departmentInformatikai tudományok doktori iskolahu
dc.contributor.submitterdepDE--Informatikai Kar -- Komputergrafika és Képfeldogozás Tanszék
dc.date.accessioned2018-10-02T20:27:39Z
dc.date.available2018-10-02T20:27:39Z
dc.date.defended2018-10-17
dc.date.issued2018
dc.description.abstractMedical imaging has improved considerably and contributed to the creation of remarkable achievements in disease treatment over the past decades. Nowadays it is completely general to use medical imaging in everyday medical practice to support non-invasive or semi-invasive medical interventions. Thanks to the technical progress of computers, digitization of imaging brought in a new era in the field of medical imaging. The stored digitized data and improved technology allow creating more accurate and precise treatments. At the same time, the larger world population causes an increased demand for medical treatment, which stimulates the era of automatization in medical imaging. In recent years the number of the technology-driven developments are increasing for better treatments becoming stoppable in the age of small and mobile computers. Despite of the clear progress, medical image processing is a non-trivial task, and is one of the most widely researched fields. Due to the increasing size and diversity of data obtained by better and better sensors, the most important challenges in this field are the complexity of the image processing problem and the computation time of its algorithmic solutions. To reduce complexity and raise the accuracy of the solutions, splitting the problem into subtasks is a known answer. The detection of the location of an object on an image is one of the most common tasks in the field of medical image processing. In this task a classification procedure of the image pixels is solved by the help of classifiers. The reason for composing ensembles from individual classifiers is to make joint classification from the aggregated data instead of using only one classifier. These ensembles are popular to use since they can raise the accuracy of the final answer and make it possible to apply existing research results. For decision making, majority voting is a usually applied model. Although the ensembles can reduce the complexity in the level of sub-tasks, the complete ensemble-based software solution mostly has great computational demand.hu_HU
dc.description.correctorNE
dc.format.extent124hu_HU
dc.identifier.urihttp://hdl.handle.net/2437/257382
dc.language.isoenhu_HU
dc.subjectensemble based systemhu_HU
dc.subjectensemble based system optimization
dc.subjectdata fusion
dc.subjectmedical image analysis
dc.subjectautomated screening system
dc.subjectHPC
dc.subjecthigh performance computing
dc.subjectdiversity measure
dc.subjectsystem optimization
dc.subjecthybrid parallelization
dc.subjectdistributed systems
dc.subjectmajority voting
dc.subjectoptic disc
dc.subjectmacula
dc.subjectdiabetes retinopathy
dc.subjectmelanoma
dc.subjectliver cancer
dc.subject.disciplineInformatikai tudományokhu
dc.subject.sciencefieldMűszaki tudományokhu
dc.titleEnhanced ensemble-based object detection in medical image analysishu_HU
dc.title.translatedEnhanced ensemble-based object detection in medical image analysishu_HU
dc.typePhD, doktori értekezéshu
Fájlok
Eredeti köteg (ORIGINAL bundle)
Megjelenítve 1 - 2 (Összesen 2)
Bélyegkép
Név:
PhD_dissertation_Laszlo_Kovacs_DEA.pdf
Méret:
12.42 MB
Formátum:
Adobe Portable Document Format
Leírás:
dissertation
Letöltés
Bélyegkép
Név:
PHD_tezis_Laszlo_Kovacs_DEA.pdf
Méret:
772.69 KB
Formátum:
Adobe Portable Document Format
Leírás:
thesis
Letöltés
Engedélyek köteg
Megjelenítve 1 - 1 (Összesen 1)
Nincs kép
Név:
license.txt
Méret:
1.93 KB
Formátum:
Item-specific license agreed upon to submission
Leírás:
Letöltés
Gyűjtemények
Informatikai Tudományok Doktori Iskola