Encoding CT Anatomy Knowledge for Unpaired Chest X-ray Image Decomposition

Li, Zeju; Han, Li; Han, Hu; Shi, Gonglei; Wang, Jiannan; Zhou, S. Kevin

doi:10.1007/978-3-030-32226-7_31

Cited by 20 publications

(20 citation statements)

References 10 publications

(16 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Convolutions Neural Networks (CNNs) have been widely used in CAD systems [26], e.g., for classification of diseases, segmentation of organs and lesions, and detection of lesions. In order to improve the accuracy of disease classification, prior works focus on improve the models [14,16] from multiple perspectives, e.g., incorporating attention [50], adopting self-training [29,43], or utilizing medical knowledge [22]. For the segmentation of organs and lesions, UNet [39] is one classic network, which has inspired many follow-up variants, such as Attention U-Net [34] and mUNet [41].…”

Section: Related Workmentioning

confidence: 99%

FIBA: Frequency-Injection based Backdoor Attack in Medical Image Analysis

Feng¹,

Ma²,

Zhang³

et al. 2021

Preprint

View full text Add to dashboard Cite

In recent years, the security of AI systems has drawn increasing research attention, especially in the medical imaging realm. To develop a secure medical image analysis (MIA) system, it is a must to study possible backdoor attacks (BAs), which can embed hidden malicious behaviors into the system. However, designing a unified BA method that can be applied to various MIA systems is challenging due to the diversity of imaging modalities (e.g., X-Ray, CT, and MRI) and analysis tasks (e.g., classification, detection, and segmentation). Most existing BA methods are designed to attack natural image classification models, which apply spatial triggers to training images and inevitably corrupt the semantics of poisoned pixels, leading to the failures of attacking dense prediction models. To address this issue, we propose a novel Frequency-Injection based Backdoor Attack method (FIBA) that is capable of delivering attacks in various MIA tasks. Specifically, FIBA leverages a trigger function in the frequency domain that can inject the lowfrequency information of a trigger image into the poisoned image by linearly combining the spectral amplitude of both images. Since it preserves the semantics of the poisoned image pixels, FIBA can perform attacks on both classification and dense prediction models. Experiments on three benchmarks in MIA (i.e., ISIC-2019[4] for skin lesion classification, for kidney tumor segmentation, and EAD-2019 [1] for endoscopic artifact detection), validate the effectiveness of FIBA and its superiority over state-ofthe-art methods in attacking MIA models as well as bypassing backdoor defense. The code is available at here.

show abstract

Section: Related Workmentioning

confidence: 99%

FIBA: Frequency-Injection based Backdoor Attack in Medical Image Analysis

Feng¹,

Ma²,

Zhang³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Knowledge arises from various sources such as imaging physics, statistical constraints, and task specifics and ways of embedding into a DL approach vary too. For chest x-ray disease classification, Li et al [65] encode anatomy knowledge embedded in unpaired CT into a deep network that decomposes a chest xray into lung, bone and the remaining structures (see Fig. 2).…”

Section: E Emerging Deep Learning Approachesmentioning

confidence: 99%

“…Most publications use a standard approach of inputting the entire image in a popular convolutional network architecture. Methodological contributions include novel ways of preprocessing the images, handling the label uncertainty and the large number of classes, suppressing the bones [65], and exploiting self-supervised learning as a way of pretraining. So far, only few publications analyze multiple exams of the same patient to detect interval change or analyze the lateral views.…”

Section: A Deep Learning In Thoracic Imagingmentioning

confidence: 99%

A review of deep learning in medical imaging: Imaging traits, technology trends, case studies with progress highlights, and future promises

Zhou,

Greenspan,

Davatzikos

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Since its renaissance, deep learning has been widely used in various medical imaging tasks and has achieved remarkable success in many medical imaging applications, thereby propelling us into the so-called artificial intelligence (AI) era. It is known that the success of AI is mostly attributed to the availability of big data with annotations for a single task and the advances in high performance computing. However, medical imaging presents unique challenges that confront deep learning approaches. In this survey paper, we first highlight both clinical needs and technical challenges in medical imaging and describe how emerging trends in deep learning are addressing these issues. We cover the topics of network architecture, sparse and noisy labels, federating learning, interpretability, uncertainty quantification, etc. Then, we present several case studies that are commonly found in clinical practice, including digital pathology and chest, brain, cardiovascular, and abdominal imaging. Rather than presenting an exhaustive literature survey, we instead describe some prominent research highlights related to these case study applications. We conclude with a discussion and presentation of promising future directions.

show abstract

“…Bulat et al [10] applied the cycle architecture to generate super-resolved facial images. Li et al [53] proposed knowledge transfer between unpaired CT and X-ray images based on cycle-consistency loss, facilitating chest X-ray image decomposition. Jeong et al [54] adopted the structure of Cycle-GAN and explored the use of cross-spectral correspondence between visible and infrared images in an unpaired setting.…”

Section: B Domain Adaptationmentioning

confidence: 99%

Coupled Real-Synthetic Domain Adaptation for Real-World Deep Depth Enhancement

Guo

Deligianni

et al. 2020

IEEE Trans. on Image Process.

View full text Add to dashboard Cite

Advances in depth sensing technologies have allowed simultaneous acquisition of both color and depth data under different environments. However, most depth sensors have lower resolution than that of the associated color channels and such a mismatch can affect applications that require accurate depth recovery. Existing depth enhancement methods use simplistic noise models and cannot generalize well under real-world conditions. In this paper, a coupled real-synthetic domain adaptation method is proposed, which enables domain transfer between high-quality depth simulators and real depth camera information for super-resolution depth recovery. The method first enables the realistic degradation from synthetic images, and then enhances degraded depth data to high quality with a color-guided sub-network. The key advantage of the work is that it generalizes well to real-world datasets without further training or fine-tuning. Detailed quantitative and qualitative results are presented, and it is demonstrated that the proposed method achieves improved performance compared to previous methods fine-tuned on the specific datasets.

show abstract

Encoding CT Anatomy Knowledge for Unpaired Chest X-ray Image Decomposition

Cited by 20 publications

References 10 publications

FIBA: Frequency-Injection based Backdoor Attack in Medical Image Analysis

FIBA: Frequency-Injection based Backdoor Attack in Medical Image Analysis

A review of deep learning in medical imaging: Imaging traits, technology trends, case studies with progress highlights, and future promises

Coupled Real-Synthetic Domain Adaptation for Real-World Deep Depth Enhancement

Contact Info

Product

Resources

About