The ISC is part of the Johns Hopkins Applied Physics Laboratory and will follow all current policies. Please visit the JHU/APL page for more information on the Lab's visitor guidance.

2019

Unsupervised deep novelty detection: application to muscle ultrasound and myositis screening


Abstract

This study investigates unsupervised novelty detection (ND) for screening of rare myopathies and specifically myositis. To support this study we developed from the ground up a novel and fully annotated dataset consisting of 3586 images taken of eighty nine individuals obtained under informed consent during 2016-2017. We developed and compared performance for several ND methods leveraging deep feature embeddings, utilizing generative as well as discriminative deep learning approaches for embeddings, and using various novelty scores. We carried out several performance comparisons including with a clinician, supervised binary classification approaches, and a generative method, demonstrating that our best performing approach is competitive with human performance and other best of breed algorithms.

Citation

@inproceedingsBurlina_2019 doi: 10.1109/isbi.2019.8759565 url: https://doi.org/10.1109/isbi.2019.8759565 year: 2019 month: apr publisher: IEEE author: Burlina P. and Joshi N. and Billings S. and Wang I. J. and Albayda J. title: Unsupervised Deep Novelty Detection: Application To Muscle Ultrasound And Myositis Screening booktitle: 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019)

Citation

@inproceedingsBurlina_2019 doi: 10.1109/isbi.2019.8759565 url: https://doi.org/10.1109/isbi.2019.8759565 year: 2019 month: apr publisher: IEEE author: Burlina P. and Joshi N. and Billings S. and Wang I. J. and Albayda J. title: Unsupervised Deep Novelty Detection: Application To Muscle Ultrasound And Myositis Screening booktitle: 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019)