Classification of Human Blastocyst Quality Using Wavelets and Transfer Learning

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Embryo culture and transfer are the procedure of maturation and transmission of the embryo into the uterus. This procedure is one of a stage in the series of in vitro fertilization processes, better known as IVF. The selection of good quality embryos to be implanted presents a problem because of the blastocyst image. Blastocyst image is a very intricate texture to be visually determined, which is good or poor quality. This research aims to implement the pre-trained Inception-v3 network to predict blastocyst quality with add image pre-processing using wavelets. Using only 249 of human blastocyst microscope images, we developed an accurate classifier that can classify blastocyst quality with a transfer learning. The experiment with twenty epochs, the accuracy of training for only raw blastocyst images is 95%, and the best training accuracy uses a pre-processing image with Daubechies 6-tap of 99.29%. Our model was then tested on the 14 of blastocyst images and classified the images of two kinds of grade with the best accuracy of around 64.29%.

Original languageEnglish
Title of host publicationAdvances in Computer, Communication and Computational Sciences - Proceedings of IC4S 2019
EditorsSanjiv K. Bhatia, Shailesh Tiwari, Su Ruidan, Munesh Chandra Trivedi, K. K. Mishra
PublisherSpringer Science and Business Media Deutschland GmbH
Pages965-974
Number of pages10
ISBN (Print)9789811544088
DOIs
Publication statusPublished - 2021
EventInternational Conference on Computer, Communication and Computational Sciences, IC4S 2019 - Bangkok, Thailand
Duration: 11 Oct 201912 Oct 2019

Publication series

NameAdvances in Intelligent Systems and Computing
Volume1158
ISSN (Print)2194-5357
ISSN (Electronic)2194-5365

Conference

ConferenceInternational Conference on Computer, Communication and Computational Sciences, IC4S 2019
Country/TerritoryThailand
CityBangkok
Period11/10/1912/10/19

Keywords

  • Human blastocyst
  • Quality classification
  • Transfer learning
  • Wavelets

Fingerprint

Dive into the research topics of 'Classification of Human Blastocyst Quality Using Wavelets and Transfer Learning'. Together they form a unique fingerprint.

Cite this