Image area reduction for efficient medical image retrieval
Başlık çevirisi mevcut değil.
- Tez No: 912078
- Danışmanlar: Belirtilmemiş.
- Tez Türü: Yüksek Lisans
- Konular: Belirtilmemiş.
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 2015
- Dil: İngilizce
- Üniversite: University of Waterloo
- Enstitü: Yurtdışı Enstitü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 87
Özet
Content-based image retrieval (CBIR) has been one of the most active areas in medical image analysis in the last two decades because of steadily increase in the number of digital images. Efficient diagnosis and treatment planning can be supported with developing retrieval systems to provide high-quality healthcare. Extensive research has attempted to improve the retrieval efficiency. The critical factors when searching in large databases are time and storage requirements. In general, although many methods have been suggested to increase accuracy, fast retrieval has been comparably rather sporadically investigated. In this thesis, two different approaches are proposed to reduce both time and space requirements for medical image retrieval. The IRMA data set is used to validate the methods. Both methods utilized Local Binary Pattern (LBP) histogram features which are extracted from 14,410 X-ray images. The first method is image folding that operates based on salient regions in an image. Saliency is determined by a context-aware saliency algorithm. After the folding process, the reduced image area is used to extract multi-block and multi-scale LBP features and to classify these features by multi-class SVM. The other method consists of classification and distance-based feature similarity. Images are firstly classified into general classes by utilizing LBP features. Subsequently, the retrieval is performed within the class to locate the most similar images. Between the retrieval and classification processes, LBP features are eliminated by employing the error histogram of a shallow (n/p/n) autoencoder to quantify the retrieval relevance of image blocks. If the region is relevant, autoencoder gives large error for its decoding. Hence, via examining the autoencoder error of image blocks, irrelevant regions can be detected and eliminated. In order to calculate similarity within general classes, the distance between the LBP features of relevant regions can be calculated. The results show that the retrieval time can be reduced, and the storage requirements can be lowered without significant decrease in accuracy.
Özet (Çeviri)
Özet çevirisi mevcut değil.
Benzer Tezler
- Denoising and enhancement in medical imaging modalities using deep learning
Medikal görüntüleme sistemlerinde derin öğrenme ile gürültü azaltımı ve görüntü iyileştirme
İREM LOÇ
Doktora
İngilizce
2024
Bilgisayar Mühendisliği Bilimleri-Bilgisayar ve KontrolBoğaziçi ÜniversitesiFizik Ana Bilim Dalı
DOÇ. DR. HAKAN ERKOL
PROF. DR. MEHMET BURÇİN ÜNLÜ
- Çinko nanotanecik içeren polimer nanokompozit malzeme üretimi ve karakterizasyonu
Fabrication and characterization of polymer nanocomposite materials incorporated zno nanoparticles
ALEV AKBAŞ
Yüksek Lisans
Türkçe
2018
Kimya Mühendisliğiİstanbul Teknik ÜniversitesiKimya Mühendisliği Ana Bilim Dalı
PROF. DR. SADRİYE KÜÇÜKBAYRAK OSKAY
- Composite nanofiber patches for topical drug delivery systems
Kompozit nanoliflerin topikal ilaç salım sistemlerinde kullanımı
ZARİFE BARBAK
Doktora
İngilizce
2021
Eczacılık ve Farmakolojiİstanbul Teknik ÜniversitesiTekstil Mühendisliği Ana Bilim Dalı
PROF. DR. HALE KARAKAŞ
- Bor karbür katkılı PMMA (polimetil metakrilat) polimerkompozitlerin ATRP metodu ile sentezi ve karakterizasyonu
Synthesis and characterization of boron carbide additived PMMA (polymethyl methacrylate) polymer composites by ATRP method
DUYGU TULUK TÜRKMANİ
Yüksek Lisans
Türkçe
2023
Metalurji Mühendisliğiİstanbul Teknik ÜniversitesiMetalurji ve Malzeme Mühendisliği Ana Bilim Dalı
PROF. DR. CÜNEYT ARSLAN
PROF. DR. NİLGÜN BAYDOĞAN