Automating Labeling of Fetal Ultrasound Views Using Machine Learning
Document Type
Dissertation
Abstract
Ultrasound is the primary imaging modality used in clinical practice during pregnancy. More than 140M fetuses are born yearly, resulting in numerous scans. The availability of a large volume of data presents the opportunity to train machine learning models. However, this abundance of scans also poses challenges, as manual labeling of each image is needed for supervised methods. Labeling is typically labor-intensive and requires expertise to annotate the images accurately. This study introduces a way to generate pseudo-labeling for fetal ultrasound views and presents an unsupervised approach for automatically clustering images into a large range of fetal views, reducing or eliminating the need for manual labeling. Our Fetal Ultrasound Semantic Clustering (FUSC) method is developed using a large dataset of 88,063 images and further evaluated on an additional unseen dataset of 8,187 images achieving over 94% clustering purity. The results of our investigation hold the potential to significantly impact the field of fetal imaging and pave the way for more advanced automated labeling solutions.
First Page
i
Last Page
39
Publication Date
6-2023
Recommended Citation
H. Alasmawi, "Automating Labeling of Fetal Ultrasound Views Using Machine Learning", M.S. Thesis, Machine Learning, MBZUAI, Abu Dhabi, UAE, 2023.
Comments
Thesis submitted to the Deanship of Graduate and Postdoctoral Studies
In partial fulfillment of the requirements for the M.Sc degree in Machine Learning
Advisors: Dr. Mohammad Yaqub, Dr. Karthik Nandakumar
Online access available for MBZUAI patrons