arXiv subjects

Pingkun Yan

Publications and source records attributed to Pingkun Yan.

59 records · Page 4Linked to original sources

arXiv2018

Multifactorial cancer treatment outcome prediction through multifaceted radiomics

Accurately predicting the treatment outcome plays a greatly important role in tailoring and adapting a treatment planning in cancer therapy. Although the development of different modalities and personalized medicine can greatly improve the accuracy of outcome prediction, they also bring the three mainly simultaneous challenges including multi-modality, multi-classifier and multi-criteria, which are summarized as multifactorial outcome prediction (MFOP) in this paper. Compared with traditional outcome prediction, MFOP is a more generalized problem. To handle this novel problem, based on the recent proposed radiomics, we propose a new unified framework termed as multifaceted radiomics (M-radiomics). M-radiomics trains multiple modality-specific classifiers first and then optimally combines the output from the outputs of different classifiers which are trained according to multiple different criteria such as sensitivity and specificity. It considers multi-modality, multi-classifier and multi-criteria into a unified framework, which makes the prediction more accurate. Furthermore, to obtain the more reliable predictive performance which is to maximize the similarity between predicted output and labelled vector, a new validation set based reliable fusion (VRF) strategy and reliable optimization models as well as a new recursive two stage hybrid optimization algorithm (RTSH) were also developed. Two clinical problems for predicting distant metastasis and locoregional recurrence in head & neck cancer were investigated to validate the performance and reliability of the proposed M-radiomics. By using the proposed RF strategy and RTSH optimization algorithm, the experimental results demonstrated that M-radiomics performed better than current radiomic models that rely on a single objective, modality or classifier.

physics.med-ph↗

arXiv2018

Low Dose CT Image Denoising Using a Generative Adversarial Network with Wasserstein Distance and Perceptual Loss

In this paper, we introduce a new CT image denoising method based on the generative adversarial network (GAN) with Wasserstein distance and perceptual similarity. The Wasserstein distance is a key concept of the optimal transform theory, and promises to improve the performance of the GAN. The perceptual loss compares the perceptual features of a denoised output against those of the ground truth in an established feature space, while the GAN helps migrate the data noise distribution from strong to weak. Therefore, our proposed method transfers our knowledge of visual perception to the image denoising task, is capable of not only reducing the image noise level but also keeping the critical information at the same time. Promising results have been obtained in our experiments with clinical CT images.

cs.CV↗

arXiv2018

Deep Compressive Macroscopic Fluorescence Lifetime Imaging

Compressive Macroscopic Fluorescence Lifetime Imaging (MFLI) is a novel technical implementation that enables monitoring multiple molecular interactions in macroscopic scale. Especially, we reported recently on the development of a hyperspectral wide-field time-resolved single-pixel imaging platform that facilitates whole-body in vivo lifetime imaging in less than 14 minutes. However, despite efficient data acquisition, the data processing of a Compressed Sensing (CS) based inversion plus lifetime fitting remain very time consuming. Herein, we propose to investigate the potential of deep learning for fast and accurate image formation. More precisely we developed a Convolutional Neural Network (CNN) called Net-FLICS (Network for Fluorescence Lifetime Imaging with Compressive Sensing) that reconstructs both intensity and lifetime images directly from raw CS measurements. Results show that better quality reconstruction can be obtained using Net-FLICS, for both simulation and experimental dataset, with almost negligible time compared to the traditional analytic methods. This first investigation suggests that Net-FLICS may be a powerful tool to enable CS-based lifetime imaging for real-time applications.

physics.med-ph↗

arXiv2017

Deeply-Supervised CNN for Prostate Segmentation

Prostate segmentation from Magnetic Resonance (MR) images plays an important role in image guided interven- tion. However, the lack of clear boundary specifically at the apex and base, and huge variation of shape and texture between the images from different patients make the task very challenging. To overcome these problems, in this paper, we propose a deeply supervised convolutional neural network (CNN) utilizing the convolutional information to accurately segment the prostate from MR images. The proposed model can effectively detect the prostate region with additional deeply supervised layers compared with other approaches. Since some information will be abandoned after convolution, it is necessary to pass the features extracted from early stages to later stages. The experimental results show that significant segmentation accuracy improvement has been achieved by our proposed method compared to other reported approaches.

cs.CV↗

arXiv2017

CT Image Denoising with Perceptive Deep Neural Networks

Increasing use of CT in modern medical practice has raised concerns over associated radiation dose. Reduction of radiation dose associated with CT can increase noise and artifacts, which can adversely affect diagnostic confidence. Denoising of low-dose CT images on the other hand can help improve diagnostic confidence, which however is a challenging problem due to its ill-posed nature, since one noisy image patch may correspond to many different output patches. In the past decade, machine learning based approaches have made quite impressive progress in this direction. However, most of those methods, including the recently popularized deep learning techniques, aim for minimizing mean-squared-error (MSE) between a denoised CT image and the ground truth, which results in losing important structural details due to over-smoothing, although the PSNR based performance measure looks great. In this work, we introduce a new perceptual similarity measure as the objective function for a deep convolutional neural network to facilitate CT image denoising. Instead of directly computing MSE for pixel-to-pixel intensity loss, we compare the perceptual features of a denoised output against those of the ground truth in a feature space. Therefore, our proposed method is capable of not only reducing the image noise levels, but also keeping the critical structural information at the same time. Promising results have been obtained in our experiments with a large number of CT images.

cs.CV↗