| fullrecord |
[{"key": "dc.contributor.advisor", "value": "\u00c4yr\u00e4m\u00f6, Sami", "language": "", "element": "contributor", "qualifier": "advisor", "schema": "dc"}, {"key": "dc.contributor.advisor", "value": "P\u00f6l\u00f6nen, Ilkka", "language": "", "element": "contributor", "qualifier": "advisor", "schema": "dc"}, {"key": "dc.contributor.author", "value": "Lindroos, Jari", "language": "", "element": "contributor", "qualifier": "author", "schema": "dc"}, {"key": "dc.date.accessioned", "value": "2022-06-06T10:30:32Z", "language": null, "element": "date", "qualifier": "accessioned", "schema": "dc"}, {"key": "dc.date.available", "value": "2022-06-06T10:30:32Z", "language": null, "element": "date", "qualifier": "available", "schema": "dc"}, {"key": "dc.date.issued", "value": "2022", "language": "", "element": "date", "qualifier": "issued", "schema": "dc"}, {"key": "dc.identifier.uri", "value": "https://jyx.jyu.fi/handle/123456789/81505", "language": null, "element": "identifier", "qualifier": "uri", "schema": "dc"}, {"key": "dc.description.abstract", "value": "Rintasy\u00f6p\u00e4 on maailmanlaajuisesti naisten yleisin sy\u00f6p\u00e4, sen varhainen havaitseminen voi merkitt\u00e4v\u00e4sti v\u00e4hent\u00e4\u00e4 siihen liittyv\u00e4\u00e4 kuolleisuutta. Histopatologista analyysi\u00e4 tarvitaan kasvainten laadun m\u00e4\u00e4ritt\u00e4miseksi solutasolla. Histopatologisten kuvien manuaalinen analyysi vie kuitenkin aikaa ja on altis virheille. Syv\u00e4oppimiseen pohjautuvassa tutkimuksessa on esitetty menetelmi\u00e4 rintasy\u00f6v\u00e4n tunnistamiseen, jotka voivat auttaa patologeja diagnosoimisessa. Konvoluutioneuroverkot ovat pitk\u00e4\u00e4n olleet k\u00e4ytetyin menetelm\u00e4 rintasy\u00f6v\u00e4n luokitteluun syv\u00e4oppimisessa, mutta ne ovat enimm\u00e4kseen rajoittuneet keskittym\u00e4\u00e4n kuvien paikallisiin ominaisuuksiin. Vision Transformer on osoittautunut suoriutumaan konvoluutioneuroverkkoja paremmin useissa kuvanluokitteluteht\u00e4viss\u00e4, koska se pystyy keskittym\u00e4\u00e4n kuvien pitk\u00e4n matkan riippuvuuksiin. T\u00e4m\u00e4n tutkielman tavoitteena on arvioida Vision Transformer -pohjaisten mallien suorituskyky\u00e4 vertaamalla niit\u00e4 yleisesti k\u00e4ytettyyn konvoluutioneuroverkkoon ResNet-50, kokeilut suoritetaan PCam-aineistolla. Mallien koulutuksessa hy\u00f6dynn\u00e4mme sek\u00e4 perinteist\u00e4 siirto-oppimiseen perustuvaa l\u00e4hestymistapaa ett\u00e4 my\u00f6s toimialuekohtaiseen esikoulutukseen perustuvaa l\u00e4hestymistapaa. Osoitamme, ett\u00e4 implementoiduilla Vision Transformer -malleilla saadaan parempia tuloksia kuin ResNet-50 -mallilla. Parhaalla mallilla B/16 saavutettiin paras AUC-tulos arvolla 0.97315. Toimialuekohtaisen esikoulutuksen k\u00e4ytt\u00f6 parantaa suorituskyky\u00e4 kaikissa malleissa paitsi Ti/16 malleissa.", "language": "fi", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.abstract", "value": "Breast cancer is the most common cancer worldwide in females apart from non-melanoma skin cancer. Detecting breast cancer as early as possible could significantly reduce its death rates. Histopathological analysis of the breast tissues is needed for determining the malignancy of the tumor on a cellular level. Manual analysis of histopathological images is time consuming and sensitive to human errors. Deep learning has introduced methods for recognizing breast cancer to assist pathologists in their diagnostic workflow. The convolutional neural networks have for long been the bandwagon deep learning model for breast cancer classification, but they are mostly limited at focusing on local variations in image patterns. The Vision Transformer, which originated from the dominant Transformer architecture in natural language processing has shown to outperform convolutional neural networks on several image classification benchmarks, due to its ability to focus on long range dependencies in images. In this thesis we aim to evaluate the performance of Vision Transformer based models by comparing them to the commonly used convolutional neural network ResNet-50 on the PCam-dataset. For training we utilize both the conventional transfer learning based approach and also an pre-training approach based on domain adaptation. We demonstrate the effectiveness of the implemented Vision Transformer models in the medical domain, by obtaining better results than the ResNet-50 on the PCam-dataset, with the best model B/16 achieving the best AUC score of 0.97315. The use of domain-based pre-training shows a performance gain for every model except the Ti/16-family models.", "language": "en", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Submitted by Miia Hakanen (mihakane@jyu.fi) on 2022-06-06T10:30:32Z\nNo. of bitstreams: 0", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Made available in DSpace on 2022-06-06T10:30:32Z (GMT). No. of bitstreams: 0\n Previous issue date: 2022", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.format.extent", "value": "54", "language": "", "element": "format", "qualifier": "extent", "schema": "dc"}, {"key": "dc.format.mimetype", "value": "application/pdf", "language": null, "element": "format", "qualifier": "mimetype", "schema": "dc"}, {"key": "dc.language.iso", "value": "eng", "language": null, "element": "language", "qualifier": "iso", "schema": "dc"}, {"key": "dc.rights", "value": "In Copyright", "language": "en", "element": "rights", "qualifier": null, "schema": "dc"}, {"key": "dc.subject.other", "value": "image classification", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "vision transformer", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "transformer", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "convolutional neural network", "language": "", "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.title", "value": "Transformers for breast cancer classification", "language": "", "element": "title", "qualifier": null, "schema": "dc"}, {"key": "dc.type", "value": "master thesis", "language": null, "element": "type", "qualifier": null, "schema": "dc"}, {"key": "dc.identifier.urn", "value": "URN:NBN:fi:jyu-202206063122", "language": "", "element": "identifier", "qualifier": "urn", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Pro gradu -tutkielma", "language": "fi", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Master\u2019s thesis", "language": "en", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Informaatioteknologian tiedekunta", "language": "fi", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Faculty of Information Technology", "language": "en", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.department", "value": "Informaatioteknologia", "language": "fi", "element": "contributor", "qualifier": "department", "schema": "dc"}, {"key": "dc.contributor.department", "value": "Information Technology", "language": "en", "element": "contributor", "qualifier": "department", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "Jyv\u00e4skyl\u00e4n yliopisto", "language": "fi", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "University of Jyv\u00e4skyl\u00e4", "language": "en", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.subject.discipline", "value": "Tietotekniikka", "language": "fi", "element": "subject", "qualifier": "discipline", "schema": "dc"}, {"key": "dc.subject.discipline", "value": "Mathematical Information Technology", "language": "en", "element": "subject", "qualifier": "discipline", "schema": "dc"}, {"key": "yvv.contractresearch.funding", "value": "0", "language": "", "element": "contractresearch", "qualifier": "funding", "schema": "yvv"}, {"key": "dc.type.coar", "value": "http://purl.org/coar/resource_type/c_bdcc", "language": null, "element": "type", "qualifier": "coar", "schema": "dc"}, {"key": "dc.rights.accesslevel", "value": "openAccess", "language": null, "element": "rights", "qualifier": "accesslevel", "schema": "dc"}, {"key": "dc.type.publication", "value": "masterThesis", "language": null, "element": "type", "qualifier": "publication", "schema": "dc"}, {"key": "dc.subject.oppiainekoodi", "value": "602", "language": "", "element": "subject", "qualifier": "oppiainekoodi", "schema": "dc"}, {"key": "dc.subject.yso", "value": "syv\u00e4oppiminen", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "rintasy\u00f6p\u00e4", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "sy\u00f6p\u00e4taudit", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "neuroverkot", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "konen\u00e4k\u00f6", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "deep learning", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "breast cancer", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "cancerous diseases", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "neural networks (information technology)", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "computer vision", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.format.content", "value": "fulltext", "language": null, "element": "format", "qualifier": "content", "schema": "dc"}, {"key": "dc.rights.url", "value": "https://rightsstatements.org/page/InC/1.0/", "language": null, "element": "rights", "qualifier": "url", "schema": "dc"}, {"key": "dc.type.okm", "value": "G2", "language": null, "element": "type", "qualifier": "okm", "schema": "dc"}]
|