Near-pair patch generative adversarial network for data augmentation of focal pathology object detection models

Ethan Tu; Jonathan Burkow; Andy Tsai; Joseph Junewick; Francisco A. Perez; Jeffrey Otjen; Adam M. Alessio

doi:10.1117/1.JMI.11.3.034505

4 June 2024 Near-pair patch generative adversarial network for data augmentation of focal pathology object detection models

Ethan Tu, Jonathan Burkow, Andy Tsai, Joseph Junewick, Francisco A. Perez, Jeffrey Otjen, Adam M. Alessio

Author Affiliations +

Journal of Medical Imaging, Vol. 11, Issue 3, 034505 (June 2024). https://doi.org/10.1117/1.JMI.11.3.034505

Abstract

Purpose

The limited volume of medical training data remains one of the leading challenges for machine learning for diagnostic applications. Object detectors that identify and localize pathologies require training with a large volume of labeled images, which are often expensive and time-consuming to curate. To reduce this challenge, we present a method to support distant supervision of object detectors through generation of synthetic pathology-present labeled images.

Approach

Our method employs the previously proposed cyclic generative adversarial network (cycleGAN) with two key innovations: (1) use of “near-pair” pathology-present regions and pathology-absent regions from similar locations in the same subject for training and (2) the addition of a realism metric (Fréchet inception distance) to the generator loss term. We trained and tested this method with 2800 fracture-present and 2800 fracture-absent image patches from 704 unique pediatric chest radiographs. The trained model was then used to generate synthetic pathology-present images with exact knowledge of location (labels) of the pathology. These synthetic images provided an augmented training set for an object detector.

Results

In an observer study, four pediatric radiologists used a five-point Likert scale indicating the likelihood of a real fracture (1 = definitely not a fracture and 5 = definitely a fracture) to grade a set of real fracture-absent, real fracture-present, and synthetic fracture-present images. The real fracture-absent images scored 1.7±1.0, real fracture-present images 4.1±1.2, and synthetic fracture-present images 2.5±1.2. An object detector model (YOLOv5) trained on a mix of 500 real and 500 synthetic radiographs performed with a recall of 0.57±0.05 and an F2 score of 0.59±0.05. In comparison, when trained on only 500 real radiographs, the recall and F2 score were 0.49±0.06 and 0.53±0.06, respectively.

Conclusions

Our proposed method generates visually realistic pathology and that provided improved object detector performance for the task of rib fracture detection.

Citation Download Citation

Ethan Tu, Jonathan Burkow, Andy Tsai, Joseph Junewick, Francisco A. Perez, Jeffrey Otjen, and Adam M. Alessio "Near-pair patch generative adversarial network for data augmentation of focal pathology object detection models," Journal of Medical Imaging 11(3), 034505 (4 June 2024). https://doi.org/10.1117/1.JMI.11.3.034505

Received: 23 August 2023; Accepted: 8 May 2024; Published: 4 June 2024

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $24.00

Non-members: $28.00 ADD TO CART

JOURNAL ARTICLE
13 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Education and training

Object detection

Radiography

Sensors

Data modeling

Pathology

Gallium nitride

Purpose

Approach

Results

Conclusions

Show All Keywords

Keywords/Phrases

Search In:

Publication Years