arXiv subjects

Hongming Shan

Publications and source records attributed to Hongming Shan.

116 records · Page 7Linked to original sources

arXiv2019

MRI Super-Resolution with Ensemble Learning and Complementary Priors

Magnetic resonance imaging (MRI) is a widely used medical imaging modality. However, due to the limitations in hardware, scan time, and throughput, it is often clinically challenging to obtain high-quality MR images. The super-resolution approach is potentially promising to improve MR image quality without any hardware upgrade. In this paper, we propose an ensemble learning and deep learning framework for MR image super-resolution. In our study, we first enlarged low resolution images using 5 commonly used super-resolution algorithms and obtained differentially enlarged image datasets with complementary priors. Then, a generative adversarial network (GAN) is trained with each dataset to generate super-resolution MR images. Finally, a convolutional neural network is used for ensemble learning that synergizes the outputs of GANs into the final MR super-resolution images. According to our results, the ensemble learning results outcome any one of GAN outputs. Compared with some state-of-the-art deep learning-based super-resolution methods, our approach is advantageous in suppressing artifacts and keeping more image details.

eess.IV↗

arXiv2018

Can Deep Learning Outperform Modern Commercial CT Image Reconstruction Methods?

Commercial iterative reconstruction techniques on modern CT scanners target radiation dose reduction but there are lingering concerns over their impact on image appearance and low contrast detectability. Recently, machine learning, especially deep learning, has been actively investigated for CT. Here we design a novel neural network architecture for low-dose CT (LDCT) and compare it with commercial iterative reconstruction methods used for standard of care CT. While popular neural networks are trained for end-to-end mapping, driven by big data, our novel neural network is intended for end-to-process mapping so that intermediate image targets are obtained with the associated search gradients along which the final image targets are gradually reached. This learned dynamic process allows to include radiologists in the training loop to optimize the LDCT denoising workflow in a task-specific fashion with the denoising depth as a key parameter. Our progressive denoising network was trained with the Mayo LDCT Challenge Dataset, and tested on images of the chest and abdominal regions scanned on the CT scanners made by three leading CT vendors. The best deep learning based reconstructions are systematically compared to the best iterative reconstructions in a double-blinded reader study. It is found that our deep learning approach performs either comparably or favorably in terms of noise suppression and structural fidelity, and runs orders of magnitude faster than the commercial iterative CT reconstruction algorithms.

cs.CV↗

arXiv2018

Super-resolution MRI through Deep Learning

Magnetic resonance imaging (MRI) is extensively used for diagnosis and image-guided therapeutics. Due to hardware, physical and physiological limitations, acquisition of high-resolution MRI data takes long scan time at high system cost, and could be limited to low spatial coverage and also subject to motion artifacts. Super-resolution MRI can be achieved with deep learning, which is a promising approach and has a great potential for preclinical and clinical imaging. Compared with polynomial interpolation or sparse-coding algorithms, deep learning extracts prior knowledge from big data and produces superior MRI images from a low-resolution counterpart. In this paper, we adapt two state-of-the-art neural network models for CT denoising and deblurring, transfer them for super-resolution MRI, and demonstrate encouraging super-resolution MRI results toward two-fold resolution enhancement.

physics.med-ph↗

arXiv2018

CT Super-resolution GAN Constrained by the Identical, Residual, and Cycle Learning Ensemble(GAN-CIRCLE)

Computed tomography (CT) is widely used in screening, diagnosis, and image-guided therapy for both clinical and research purposes. Since CT involves ionizing radiation, an overarching thrust of related technical research is development of novel methods enabling ultrahigh quality imaging with fine structural details while reducing the X-ray radiation. In this paper, we present a semi-supervised deep learning approach to accurately recover high-resolution (HR) CT images from low-resolution (LR) counterparts. Specifically, with the generative adversarial network (GAN) as the building block, we enforce the cycle-consistency in terms of the Wasserstein distance to establish a nonlinear end-to-end mapping from noisy LR input images to denoised and deblurred HR outputs. We also include the joint constraints in the loss function to facilitate structural preservation. In this deep imaging process, we incorporate deep convolutional neural network (CNN), residual learning, and network in network techniques for feature extraction and restoration. In contrast to the current trend of increasing network depth and complexity to boost the CT imaging performance, which limit its real-world applications by imposing considerable computational and memory overheads, we apply a parallel $1\times1$ CNN to compress the output of the hidden layer and optimize the number of layers and the number of filters for each convolutional layer. Quantitative and qualitative evaluations demonstrate that our proposed model is accurate, efficient and robust for super-resolution (SR) image restoration from noisy LR input images. In particular, we validate our composite SR networks on three large-scale CT datasets, and obtain promising results as compared to the other state-of-the-art methods.

eess.IV↗

arXiv2018

Structure-sensitive Multi-scale Deep Neural Network for Low-Dose CT Denoising

Computed tomography (CT) is a popular medical imaging modality in clinical applications. At the same time, the x-ray radiation dose associated with CT scans raises public concerns due to its potential risks to the patients. Over the past years, major efforts have been dedicated to the development of Low-Dose CT (LDCT) methods. However, the radiation dose reduction compromises the signal-to-noise ratio (SNR), leading to strong noise and artifacts that down-grade CT image quality. In this paper, we propose a novel 3D noise reduction method, called Structure-sensitive Multi-scale Generative Adversarial Net (SMGAN), to improve the LDCT image quality. Specifically, we incorporate three-dimensional (3D) volumetric information to improve the image quality. Also, different loss functions for training denoising models are investigated. Experiments show that the proposed method can effectively preserve structural and texture information from normal-dose CT (NDCT) images, and significantly suppress noise and artifacts. Qualitative visual assessments by three experienced radiologists demonstrate that the proposed method retrieves more detailed information, and outperforms competing methods.

cs.CV↗

arXiv2018

A Synergized Pulsing-Imaging Network (SPIN)

Currently, the deep neural network is the mainstream for machine learning, and being actively developed for biomedical imaging applications with an increasing emphasis on tomographic reconstruction for MRI, CT, and other imaging modalities. Multiple deep-learning-based approaches were applied to MRI image reconstruction from k-space samples to final images. Each of these studies assumes a given pulse sequence that produces incomplete and/or inconsistent data in the Fourier space, and targets a trained neural network that recovers an underlying image as close as possible to the ground truth. For the first time, in this paper we view data acquisition and the image reconstruction as the two key parts of an integrated MRI process, and optimize both the pulse sequence and the reconstruction scheme seamlessly in the machine learning framework. Our pilot simulation results show an exemplary embodiment of our new MRI strategy. Clearly, this work can be extended to other imaging modalities and their combinations as well, such as ultrasound imaging, and also potentially simultaneous emission-transmission tomography aided by polarized radiotracers.

physics.med-ph↗

arXiv2018

3D Convolutional Encoder-Decoder Network for Low-Dose CT via Transfer Learning from a 2D Trained Network

Low-dose computed tomography (CT) has attracted a major attention in the medical imaging field, since CT-associated x-ray radiation carries health risks for patients. The reduction of CT radiation dose, however, compromises the signal-to-noise ratio, and may compromise the image quality and the diagnostic performance. Recently, deep-learning-based algorithms have achieved promising results in low-dose CT denoising, especially convolutional neural network (CNN) and generative adversarial network (GAN). This article introduces a Contracting Path-based Convolutional Encoder-decoder (CPCE) network in 2D and 3D configurations within the GAN framework for low-dose CT denoising. A novel feature of our approach is that an initial 3D CPCE denoising model can be directly obtained by extending a trained 2D CNN and then fine-tuned to incorporate 3D spatial information from adjacent slices. Based on the transfer learning from 2D to 3D, the 3D network converges faster and achieves a better denoising performance than that trained from scratch. By comparing the CPCE with recently published methods based on the simulated Mayo dataset and the real MGH dataset, we demonstrate that the 3D CPCE denoising model has a better performance, suppressing image noise and preserving subtle structures.

cs.CV↗

arXiv2017

Population Density-based Hospital Recommendation with Mobile LBS Big Data

The difficulty of getting medical treatment is one of major livelihood issues in China. Since patients lack prior knowledge about the spatial distribution and the capacity of hospitals, some hospitals have abnormally high or sporadic population densities. This paper presents a new model for estimating the spatiotemporal population density in each hospital based on location-based service (LBS) big data, which would be beneficial to guiding and dispersing outpatients. To improve the estimation accuracy, several approaches are proposed to denoise the LBS data and classify people by detecting their various behaviors. In addition, a long short-term memory (LSTM) based deep learning is presented to predict the trend of population density. By using Baidu large-scale LBS logs database, we apply the proposed model to 113 hospitals in Beijing, P. R. China, and constructed an online hospital recommendation system which can provide users with a hospital rank list basing the real-time population density information and the hospitals' basic information such as hospitals' levels and their distances. We also mine several interesting patterns from these LBS logs by using our proposed system.

cs.SI↗