Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Mammography Classification and Nodule Detection using Deep Neural Networks
KTH, School of Engineering Sciences (SCI), Mathematics (Dept.), Numerical Analysis, NA.
2017 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Mammografiklassificering och Nodulär Detektion med Djupa Neurala Nätverk (Swedish)
Abstract [en]

Mammographic screenings are the most common modality for an early detection of breast cancer, but a robust annotation of the depicted breast tissue presents an ongoing challenge, even for well-experienced radiologists. Computer-aided diagnosis systems can support the human classification. Modern systems therefore often rely on deep-learning based methods. This thesis investigates the fully automatic on-image classification of mammograms into one of the classes benign, malignant (cancerous) or normal. In this context, we compare two different design paradigms, one straightforward end-to-end model with a more complex decomposition hierarchy.

While the end-to-end model consists mainly of the deep-learning based classifier, the decomposition pipeline incorporates multiple stages i.e. a region of interest detection (realized as a fully convolutional architecture) followed by a classification stage. Contrary to initial expectations, the end-to-end classifier turned out to obtain a superior performance in terms of accuracy (end-to-end: 76.57 %, decomposition: 65.66 %, computed as mean over all three classes in a one vs. all evaluation) and an improved area under receiver operating characteristic-score. All discussed parametric models were trained from scratch without using pre-trained network weights. Therefore we discuss the choice of hyper-parameters, initialization, and choice of a feasible cost function. For a successful feature extraction, in the region of interest detection stage, the negative dice coefficient proved itself to be a more robust cost function than the also investigated sensitivity-specificity loss.

Abstract [sv]

Mammografiscreening är den vanligaste modaliteten för tidig detektion av bröstcancer, men en robust annotering av den avbildade bröstvävnaden innebär en fortgående utmaning, till och med för en erfaren radiolog. Datorstödda diagnossystem kan bistå den mänskliga klassifikationen. Moderna system lutar sig därför ofta på datorbaserade metoder för djupinlärning. Den här avhandlingen undersöker den fullt automatiserade ”on-image” klassifikationen av mammogram i klasserna benign, malignt (cancer) eller normal. I denna kontext undersöker vi två olika design paradigm, en direkt end-to-end modell med en mer komplex dekomposition-hierarki.

Medan end-to-end-modellen främst består av en deep-learning-baserad klassificerare består decomposition-pipelinen av flera steg, d.v.s. en detektion av en intresseregion (implementerad som en fullt faltningsoperations- neuralnätverk), följt av ett klassificeringsstadium. Till skillnad från initiala förväntningar visade sig det att end-to-end-klassificeraren erhöll en överlägsen prestanda när det gäller noggrannhet (end-to-end: 76.57 %, dekomposition: 65.66 %, mätvärdena är beräknade som medel av alla tre klasser i en en-mot-alla-utvärdering) och en förbättrad area-under-mottagare-operations karaktäristik. Alla behandlade parametriska modeller tränades initialt utan användning av förtränade nätverksvariabler. Därför diskuteras valet av hyper-parametrar, initiering, och val av rimlig kostnadsfunktion. För en funktionsextraktion, vid detektionsstadium i regionen av intresse, visade sig den negativa dice koefficienten vara en mer robust kostnadsfunktion än den också undersökta sensitivity-specifity loss.

Place, publisher, year, edition, pages
2017.
Series
TRITA-MAT-E ; 2017:78
National Category
Computational Mathematics
Identifiers
URN: urn:nbn:se:kth:diva-220319OAI: oai:DiVA.org:kth-220319DiVA, id: diva2:1167376
Subject / course
Scientific Computing
Educational program
Master of Science - Applied and Computational Mathematics
Supervisors
Examiners
Available from: 2017-12-18 Created: 2017-12-18 Last updated: 2017-12-18Bibliographically approved

Open Access in DiVA

fulltext(1726 kB)84 downloads
File information
File name FULLTEXT01.pdfFile size 1726 kBChecksum SHA-512
b9f0665b5526a1ec12c5ad540e6d7bcb6678bbb10993d2074ec6bf333acca658f2af6ddcb9d6e602cd932613c7ed97c44762518ba9e1b4fa6a2a58e1c54f4a76
Type fulltextMimetype application/pdf

By organisation
Numerical Analysis, NA
Computational Mathematics

Search outside of DiVA

GoogleGoogle Scholar
Total: 84 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 1977 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf