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.

Adversarial cycle-consistent synthesis of cerebral microbleeds for data augmentation

Medical Image Synthesis for Data Augmentation and Anonymization using Generative Adversarial Networks

Red-GAN: Attacking Class Imbalance Via Conditioned Generation. Yet Another Perspective on Medical Image Synthesis for Skin Lesion Dermoscopy and Brain Tumor MRI

Self-supervised Brain Lesion Generation for Effective Data Augmentation of Medical Images

Adversarial Pseudo Healthy Synthesis Needs Pathology Factorization

Label-informed cardiac magnetic resonance image synthesis through conditional generative adversarial networks

An Adversarial Learning Approach to Medical Image Synthesis for Lesion Detection

Red blood cell image generation for data augmentation using Conditional Generative Adversarial Networks

Selective Synthetic Augmentation with Quality Assurance

CycleSGAN: A cycle-consistent and semantics-preserving generative adversarial network for unpaired MR-to-CT image synthesis

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

An Adversarial Learning Approach to Medical Image Synthesis for Lesion Removal.

Image Translation for Medical Image Generation -- Ischemic Stroke Lesions

An Attentive-based Generative Model for Medical Image Synthesis

Image translation for medical image generation: Ischemic stroke lesion segmentation

Unpaired Brain MR-to-CT Synthesis using a Structure-Constrained CycleGAN

Subject-Specific Lesion Generation and Pseudo-Healthy Synthesis for Multiple Sclerosis Brain Images

Synthesis of Brain Tumor MR Images for Learning Data Augmentation

Lesion Mask-based Simultaneous Synthesis of Anatomic and MolecularMR Images using a GAN

Unsupervised learning for cross-domain medical image synthesis using deformation invariant cycle consistency networks

TBI-GAN: An Adversarial Learning Approach for Data Synthesis on Traumatic Brain Segmentation