arXiv subjects

Xun Jia

Publications and source records attributed to Xun Jia.

At least 37 records · Page 2Linked to original sources

arXiv2021

Interactive Treatment Planning in High Dose-Rate Brachytherapy for Gynecological Cancer

High dose-rate brachytherapy (HDRBT) is widely used for gynecological cancer treatment. Although commercial treatment planning systems (TPSs) have inverse optimization modules, it takes several iterations to adjust planning objectives to achieve a satisfactory plan. Interactive plan-modification modules enable modifying the plan and visualizing results in real time, but they update plans based on simple geometrical or heuristic algorithms, which cannot ensure resulting plan optimality. This project develops an interactive plan optimization module for HDRBT of gynecological cancer. By efficiently solving an optimization problem in real time, it allows a user to visualize a plan and interactively modify it to improve quality. We formulated an optimization problem with an objective function containing a weighted sum of doses to normal organs subject to user-specified target coverage. A user interface was developed that allows a user to adjust organ weights using scroll bars. With a simple mouse click, the optimization problem is solved in seconds with a highly efficient alternating-direction method of multipliers and a warm start optimization strategy. Resulting clinically relevant D2cc of organs are displayed immediately. This allows a user to intuitively adjust plans with satisfactory quality. We tested the effectiveness of our development in cervix cancer cases treated with a tandem-and-ovoid applicator. It took a maximum of 3 seconds to solve the optimization problem in each instance. With interactive optimization capability, a satisfactory plan can be obtained in <1 min. In our clinic, although the time for plan adjustment was typically <10min with simple interactive plan modification tools in TPS, the resulting plans do not ensure optimality. Our plans achieved on average 5% lower D2cc than clinical plans, while maintaining target coverage.

physics.med-ph↗

arXiv2021

Using reinforcement learning to design an AI assistantfor a satisfying co-op experience

In this project, we designed an intelligent assistant player for the single-player game Space Invaders with the aim to provide a satisfying co-op experience. The agent behaviour was designed using reinforcement learning techniques and evaluated based on several criteria. We validate the hypothesis that an AI-driven computer player can provide a satisfying co-op experience.

cs.AI↗

arXiv2021

Image Synthesis for Data Augmentation in Medical CT using Deep Reinforcement Learning

Deep learning has shown great promise for CT image reconstruction, in particular to enable low dose imaging and integrated diagnostics. These merits, however, stand at great odds with the low availability of diverse image data which are needed to train these neural networks. We propose to overcome this bottleneck via a deep reinforcement learning (DRL) approach that is integrated with a style-transfer (ST) methodology, where the DRL generates the anatomical shapes and the ST synthesizes the texture detail. We show that our method bears high promise for generating novel and anatomically accurate high resolution CT images at large and diverse quantities. Our approach is specifically designed to work with even small image datasets which is desirable given the often low amount of image data many researchers have available to them.

cs.CV↗

arXiv2021

Noise Entangled GAN For Low-Dose CT Simulation

We propose a Noise Entangled GAN (NE-GAN) for simulating low-dose computed tomography (CT) images from a higher dose CT image. First, we present two schemes to generate a clean CT image and a noise image from the high-dose CT image. Then, given these generated images, an NE-GAN is proposed to simulate different levels of low-dose CT images, where the level of generated noise can be continuously controlled by a noise factor. NE-GAN consists of a generator and a set of discriminators, and the number of discriminators is determined by the number of noise levels during training. Compared with the traditional methods based on the projection data that are usually unavailable in real applications, NE-GAN can directly learn from the real and/or simulated CT images and may create low-dose CT images quickly without the need of raw data or other proprietary CT scanner information. The experimental results show that the proposed method has the potential to simulate realistic low-dose CT images.

eess.IV↗

arXiv2021

A Feasibility Study on Deep Learning Based Individualized 3D Dose Distribution Prediction

Purpose: Radiation therapy treatment planning is a trial-and-error, often time-consuming process. An optimal dose distribution based on a specific anatomy can be predicted by pre-trained deep learning (DL) models. However, dose distributions are often optimized based on not only patient-specific anatomy but also physician preferred trade-offs between planning target volume (PTV) coverage and organ at risk (OAR) sparing. Therefore, it is desirable to allow physicians to fine-tune the dose distribution predicted based on patient anatomy. In this work, we developed a DL model to predict the individualized 3D dose distributions by using not only the anatomy but also the desired PTV/OAR trade-offs, as represented by a dose volume histogram (DVH), as inputs. Methods: The desired DVH, fine-tuned by physicians from the initially predicted DVH, is first projected onto the Pareto surface, then converted into a vector, and then concatenated with mask feature maps. The network output for training is the dose distribution corresponding to the Pareto optimal DVH. The training/validation datasets contain 77 prostate cancer patients, and the testing dataset has 20 patients. Results: The trained model can predict a 3D dose distribution that is approximately Pareto optimal. We calculated the difference between the predicted and the optimized dose distribution for the PTV and all OARs as a quantitative evaluation. The largest average error in mean dose was about 1.6% of the prescription dose, and the largest average error in the maximum dose was about 1.8%. Conclusions: In this feasibility study, we have developed a 3D U-Net model with the anatomy and desired DVH as inputs to predict an individualized 3D dose distribution. The predicted dose distributions can be used as references for dosimetrists and physicians to rapidly develop a clinically acceptable treatment plan.

physics.med-ph↗

arXiv2020

Experimental and numerical studies on kV scattered x-ray imaging for real-time image guidance in radiation therapy

Motion management is a critical component of image guidance radiotherapy for lung cancer. We previously proposed a scheme using kV scattered x-ray photons for marker-less real-time image guidance in lung cancer radiotherapy. This study reports our recently progress using the photon counting detection technique to demonstrate potential feasibility of this method and using Monte Carlo (MC) simulations and ray-tracing calculations to characterize the performance. In our scheme, a thin slice of x-ray beam was directed to the target and we measured the outgoing scattered photons using a photon counting detector with a parallel-hole collimator to establish the correspondence between detector pixels and scatter positions. Image corrections of geometry, beam attenuation and scattering angle were performed to convert the raw image to the actual image of Compton attenuation coefficient. We set up a MC simulation system using an in-house developed GPU-based MC package modeling the image formation process. We also performed ray-tracing calculations to investigate the impacts of imaging system geometry on resulting image resolution. The experiment demonstrated feasibility of using a photon counting detector to measure scattered x-ray photons and generate the proposed scattered x-ray image. After correction, x-ray scattering image intensity and Compton scattering attenuation coefficient were linearly related, with R2=0.91. Contrast to Noise Ratios of different objects were improved and the values in experimental results and MC simulation results agreed with each other. Ray-tracing calculations revealed the dependence of image resolution on imaging geometry. The image resolution increases with reduced source to object distance and increased collimator height. The study demonstrated potential feasibility of using scattered x-ray imaging as a real-time image guidance method in radiation therapy.

physics.med-ph↗

arXiv2020

Topologically Nontrivial Interband Plasmons in Type-II Weyl Semimetal MoTe$_2$

In many realistic topological materials, more than one kind of fermions contribute to the electronic bands crossing the Fermi level, leading to various novel phenomena. Here, using momentum-resolved inelastic electron scattering, we investigate the plasmons and their evolution across the phase transition in a type-II Weyl Semimetal MoTe$_2$, in which both Weyl fermions and trivial nonrelativistic fermions contribute to the Fermi surface in the Td phase. One plasmon mode in the 1T' phase at high temperature and two plasmon modes in the topological T$_d$ phase at low temperature are observed. Combining with first-priciples calculations, we show that all the plasmon modes are dominated by the interband correlations between the inverted bands of MoTe$_2$. Especially in the T$_d$ phase, since the electronic bands split due to inversion symmetry breaking and spin-orbit coupling, the plasmon modes manifest the interband correlation between the topological Weyl fermions and the trivial nonrelativistic electrons. Our work emphasizes the significance of the interplay between different kinds of carriers in plasmons of topological materials.

cond-mat.mtrl-sci↗

arXiv2020

Improving Efficiency of Training a Virtual Treatment Planner Network via Knowledge-guided Deep Reinforcement Learning for Intelligent Automatic Treatment Planning of Radiotherapy

We previously proposed an intelligent automatic treatment planning framework for radiotherapy, in which a virtual treatment planner network (VTPN) was built using deep reinforcement learning (DRL) to operate a treatment planning system (TPS). Despite the success, the training of VTPN via DRL was time consuming. Also the training time is expected to grow with the complexity of the treatment planning problem, preventing the development of VTPN for more complicated but clinically relevant scenarios. In this study we proposed a knowledge-guided DRL (KgDRL) that incorporated knowledge from human planners to guide the training process to improve the training efficiency. Using prostate cancer intensity modulated radiation therapy as a testbed, we first summarized a number of rules of operating our in-house TPS. In training, in addition to randomly navigating the state-action space, as in the DRL using the epsilon-greedy algorithm, we also sampled actions defined by the rules. The priority of sampling actions from rules decreased over the training process to encourage VTPN to explore new policy that was not covered by the rules. We trained a VTPN using KgDRL and compared its performance with another VTPN trained using DRL. Both VTPNs trained via KgDRL and DRL spontaneously learned to operate the TPS to generate high-quality plans, achieving plan quality scores of 8.82 and 8.43, respectively. Both VTPNs outperformed treatment planning purely based on the rules, which had a plan score of 7.81. VTPN trained with 8 episodes using KgDRL was able to perform similarly to that trained using DRL with 100 episodes. The training time was reduced from more than a week to 13 hours. The proposed KgDRL framework accelerated the training process by incorporating human knowledge, which will facilitate the development of VTPN for more complicated treatment planning scenarios.

physics.med-ph↗

arXiv2020

Development of a Real-time Indoor Location System using Bluetooth Low Energy Technology and Deep Learning to Facilitate Clinical Applications

An indoor, real-time location system (RTLS) can benefit both hospitals and patients by improving clinical efficiency through data-driven optimization of procedures. Bluetooth-based RTLS systems are cost-effective but lack accuracy and robustness because Bluetooth signal strength is subject to fluctuation. We developed a machine learning-based solution using a Long Short-Term Memory (LSTM) network followed by a Multilayer Perceptron classifier and a posterior constraint algorithm to improve RTLS performance. Training and validation datasets showed that most machine learning models perform well in classifying individual location zones, although LSTM was most reliable. However, when faced with data indicating cross-zone trajectories, all models showed erratic zone switching. Thus, we implemented a history-based posterior constraint algorithm to reduce the variability in exchange for a slight decrease in responsiveness. This network increases robustness at the expense of latency. When latency is less of a concern, we computed the latency-corrected accuracy which is 100% for our testing data, significantly improved from LSTM without constraint which is 96.2%. The balance between robustness and responsiveness can be considered and adjusted on a case-by-case basis, according to the specific needs of downstream clinical applications. This system was deployed and validated in an academic medical center. Industry best practices enabled system scaling without substantial compromises to performance or cost.

eess.SP↗

arXiv2019

Incorporating human and learned domain knowledge into training deep neural networks: A differentiable dose volume histogram and adversarial inspired framework for generating Pareto optimal dose distributions in radiation therapy

We propose a novel domain specific loss, which is a differentiable loss function based on the dose volume histogram, and combine it with an adversarial loss for the training of deep neural networks to generate Pareto optimal dose distributions. The mean squared error (MSE) loss, dose volume histogram (DVH) loss, and adversarial (ADV) loss were used to train 4 instances of the neural network model: 1) MSE, 2) MSE+ADV, 3) MSE+DVH, and 4) MSE+DVH+ADV. 70 prostate patients were acquired, and the dose influence arrays were calculated for each patient. 1200 Pareto surface plans per patient were generated by pseudo-randomizing the tradeoff weights (84,000 plans total). We divided the data into 54 training, 6 validation, and 10 testing patients. Each model was trained for 100,000 iterations, with a batch size of 2. The prediction time of each model is 0.052 seconds. Quantitatively, the MSE+DVH+ADV model had the lowest prediction error of 0.038 (conformation), 0.026 (homogeneity), 0.298 (R50), 1.65% (D95), 2.14% (D98), 2.43% (D99). The MSE model had the worst prediction error of 0.134 (conformation), 0.041 (homogeneity), 0.520 (R50), 3.91% (D95), 4.33% (D98), 4.60% (D99). For both the mean dose PTV error and the max dose PTV, Body, Bladder and rectum error, the MSE+DVH+ADV outperformed all other models. All model's predictions have an average mean and max dose error less than 2.8% and 4.2%, respectively. Expert human domain specific knowledge can be the largest driver in the performance improvement, and adversarial learning can be used to further capture nuanced features. The real-time prediction capabilities allow for a physician to quickly navigate the tradeoff space, and produce a dose distribution as a tangible endpoint for the dosimetrist to use for planning. This can considerably reduce the treatment planning time, allowing for clinicians to focus their efforts on challenging cases.

physics.med-ph↗

arXiv2019

Electron-phonon Coupling and Kohn Anomaly due to the Floating 2D Electronic Bands on the Surface of ZrSiS

ZrSiS, an intriguing candidate of topological nodal line semimetals, was discovered to have exotic surface floating two-dimensional (2D) electrons [Phys. Rev. X 7, 041073 (2017)], which are likely to interact with surface phonons. Here, we reveal a prominent Kohn anomaly in a surface phonon branch by mapping out the surface phonon dispersions of ZrSiS using high-resolution electron energy loss spectroscopy. Theoretical analysis via an electron-phonon coupling (EPC) model attributes the strong renormalization of the surface phonon branch to the interactions with the surface floating 2D electrons. With the random phase approximation, we calculate the phonon self-energy and evaluate the mode-specific EPC constant by fitting the experimental data. The EPC picture provided here may be important for potential applications of topological nodal line semimetals.

cond-mat.mtrl-sci↗

arXiv2019

Generating Pareto optimal dose distributions for radiation therapy treatment planning

Radiotherapy treatment planning currently requires many trail-and-error iterations between the planner and treatment planning system, as well as between the planner and physician for discussion/consultation. The physician's preferences for a particular patient cannot be easily quantified and precisely conveyed to the planner. In this study we present a real-time volumetric Pareto surface dose generation deep learning neural network that can be used after segmentation by the physician, adding a tangible and quantifiable end-point to portray to the planner. From 70 prostate patients, we first generated 84,000 intensity modulated radiation therapy plans (1,200 plans per patient) sampling the Pareto surface, representing various tradeoffs between the planning target volume (PTV) and the organs-at-risk (OAR), including bladder, rectum, left femur, right femur, and body. We divided the data to 10 test patients and 60 training/validation patients. We then trained a hierarchically densely connected convolutional U-net (HD U-net), to take the PTV and avoidance map representing OARs masks and weights, and predict the optimized plan. The HD U-net is capable of accurately predicting the 3D Pareto optimal dose distributions, with average [mean, max] dose errors of [3.4%, 7.7%](PTV), [1.6%, 5.6%](bladder), [3.7%, 4.2%](rectum), [3.2%, 8.0%](left femur), [2.9%, 7.7%](right femur), and [0.04%, 5.4%](body) of the prescription dose. The PTV dose coverage prediction was also very similar, with errors of 1.3% (D98) and 2.0% (D99). Homogeneity was also similar, differing by 0.06 on average. The neural network can predict the dose within 1.7 seconds. Clinically, the optimization and dose calculation is much slower, taking 5-10 minutes.

physics.med-ph↗

arXiv2019

Three-Dimensional Radiotherapy Dose Prediction on Head and Neck Cancer Patients with a Hierarchically Densely Connected U-net Deep Learning Architecture

The treatment planning process for patients with head and neck (H&N) cancer is regarded as one of the most complicated due to large target volume, multiple prescription dose levels, and many radiation-sensitive critical structures near the target. Treatment planning for this site requires a high level of human expertise and a tremendous amount of effort to produce personalized high quality plans, taking as long as a week, which deteriorates the chances of tumor control and patient survival. To solve this problem, we propose to investigate a deep learning-based dose prediction model, Hierarchically Densely Connected U-net, based on two highly popular network architectures: U-net and DenseNet. We find that this new architecture is able to accurately and efficiently predict the dose distribution, outperforming the other two models, the Standard U-net and DenseNet, in homogeneity, dose conformity, and dose coverage on the test data. Averaging across all organs at risk, our proposed model is capable of predicting the organ-at-risk max dose within 6.3% and mean dose within 5.1% of the prescription dose on the test data. The other models, the Standard U-net and DenseNet, performed worse, having an averaged organ-at-risk max dose prediction error of 8.2% and 9.3%, respectively, and averaged mean dose prediction error of 6.4% and 6.8%, respectively. In addition, our proposed model used 12 times less trainable parameters than the Standard U-net, and predicted the patient dose 4 times faster than DenseNet.

physics.med-ph↗

arXiv2018

A feasibility study for predicting optimal radiation therapy dose distributions of prostate cancer patients from patient anatomy using deep learning

With the advancement of treatment modalities in radiation therapy for cancer patients, outcomes have improved, but at the cost of increased treatment plan complexity and planning time. The accurate prediction of dose distributions would alleviate this issue by guiding clinical plan optimization to save time and maintain high quality plans. We have modified a convolutional deep network model, U-net (originally designed for segmentation purposes), for predicting dose from patient image contours of the planning target volume (PTV) and organs at risk (OAR). We show that, as an example, we are able to accurately predict the dose of intensity-modulated radiation therapy (IMRT) for prostate cancer patients, where the average Dice similarity coefficient is 0.91 when comparing the predicted vs. true isodose volumes between 0% and 100% of the prescription dose. The average value of the absolute differences in [max, mean] dose is found to be under 5% of the prescription dose, specifically for each structure is [1.80%, 1.03%](PTV), [1.94%, 4.22%](Bladder), [1.80%, 0.48%](Body), [3.87%, 1.79%](L Femoral Head), [5.07%, 2.55%](R Femoral Head), and [1.26%, 1.62%](Rectum) of the prescription dose. We thus managed to map a desired radiation dose distribution from a patient's PTV and OAR contours. As an additional advantage, relatively little data was used in the techniques and models described in this paper.

physics.med-ph↗

arXiv2018

Intelligent Inverse Treatment Planning via Deep Reinforcement Learning, a Proof-of-Principle Study in High Dose-rate Brachytherapy for Cervical Cancer

Inverse treatment planning in radiation therapy is formulated as optimization problems. The objective function and constraints consist of multiple terms designed for different clinical and practical considerations. Weighting factors of these terms are needed to define the optimization problem. While a treatment planning system can solve the optimization problem with given weights, adjusting the weights for high plan quality is performed by human. The weight tuning task is labor intensive, time consuming, and it critically affects the final plan quality. An automatic weight-tuning approach is strongly desired. The weight tuning procedure is essentially a decision making problem. Motivated by the tremendous success in deep learning for decision making with human-level intelligence, we propose a novel framework to tune the weights in a human-like manner. Using treatment planning in high-dose-rate brachytherapy as an example, we develop a weight tuning policy network (WTPN) that observes dose volume histograms of a plan and outputs an action to adjust organ weights, similar to the behaviors of a human planner. We train the WTPN via end-to-end deep reinforcement learning. Experience replay is performed with the epsilon greedy algorithm. Then we apply the trained WTPN to guide treatment planning of testing patient cases. The trained WTPN successfully learns the treatment planning goals to guide the weight tuning process. On average, the quality score of plans generated under the WTPN's guidance is improved by ~8.5% compared to the initial plan with arbitrary weights, and by 10.7% compared to the plans generated by human planners. To our knowledge, this is the first tool to adjust weights for the treatment planning in a human-like fashion based on learnt intelligence. The study demonstrates potential feasibility to develop intelligent treatment planning system via deep reinforcement learning.

physics.med-ph↗

arXiv2018

Low dose CT reconstruction assisted by an image manifold prior

X-ray Computed Tomography (CT) is an important tool in medical imaging to obtain a direct visualization of patient anatomy. However, the x-ray radiation exposure leads to the concern of lifetime cancer risk. Low-dose CT scan can reduce the radiation exposure to patient while the image quality is usually degraded due to the appearance of noise and artifacts. Numerous studies have been conducted to regularize CT image for better image quality. Yet, exploring the underlying manifold where real CT images residing on is still an open problem. In this paper, we propose a fully data-driven manifold learning approach by incorporating the emerging deep-learning technology. An encoder-decoder convolutional neural network has been established to map a CT image to the inherent low-dimensional manifold, as well as to restore the CT image from its corresponding manifold representation. A novel reconstruction algorithm assisted by the leant manifold prior has been developed to achieve high quality low-dose CT reconstruction. In order to demonstrate the effectiveness of the proposed framework, network training, testing, and comprehensive simulation study have been performed using patient abdomen CT images. The trained encoder-decoder CNN is capable of restoring high-quality CT images with average error of ~20 HU. Furthermore, the proposed manifold prior assisted reconstruction scheme achieves high-quality low-dose CT reconstruction, with average reconstruction error of < 30 HU, more than five times and two times lower than that of filtered back projection method and total-variation based iterative reconstruction method, respectively.

physics.med-ph↗

arXiv2018

Lattice Dynamics of Ultrathin FeSe Films on SrTiO3

Charge transfer and electron-phonon coupling (EPC) are proposed to be two important constituents associated with enhanced superconductivity in the single unit cell FeSe films on oxide surfaces. Using high-resolution electron energy loss spectroscopy combined with first-principles calculations, we have explored the lattice dynamics of ultrathin FeSe films grown on SrTiO3. We show that, despite the significant effect from the substrate on the electronic structure and superconductivity of the system, the FeSe phonons in the films are unaffected. The energy dispersion and linewidth associated with the Fe- and Se-derived vibrational modes are thickness- and temperature-independent. Theoretical calculations indicate the crucial role of antiferromagnetic correlation in FeSe to reproduce the experimental phonon dispersion. Importantly, the only detectable change due to the growth of FeSe films is the broadening of the Fuchs-Kliewer (F-K) phonons associated with the lattice vibrations of SrTiO$_3$(001) substrate. If EPC plays any role in the enhancement of film superconductivity, it must be the interfacial coupling between the electrons in FeSe film and the F-K phonons from substrate rather than the phonons of FeSe.

cond-mat.supr-con↗

arXiv2017

Intelligent Parameter Tuning in Optimization-based Iterative CT Reconstruction via Deep Reinforcement Learning

A number of image-processing problems can be formulated as optimization problems. The objective function typically contains several terms specifically designed for different purposes. Parameters in front of these terms are used to control the relative weights among them. It is of critical importance to tune these parameters, as quality of the solution depends on their values. Tuning parameter is a relatively straightforward task for a human, as one can intelligently determine the direction of parameter adjustment based on the solution quality. Yet manual parameter tuning is not only tedious in many cases, but becomes impractical when a number of parameters exist in a problem. Aiming at solving this problem, this paper proposes an approach that employs deep reinforcement learning to train a system that can automatically adjust parameters in a human-like manner. We demonstrate our idea in an example problem of optimization-based iterative CT reconstruction with a pixel-wise total-variation regularization term. We set up a parameter tuning policy network (PTPN), which maps an CT image patch to an output that specifies the direction and amplitude by which the parameter at the patch center is adjusted. We train the PTPN via an end-to-end reinforcement learning procedure. We demonstrate that under the guidance of the trained PTPN for parameter tuning at each pixel, reconstructed CT images attain quality similar or better than in those reconstructed with manually tuned parameters.

physics.med-ph↗