CycleSeg: Simultaneous synthetic CT generation and unsupervised segmentation for MR-only radiotherapy treatment planning of prostate cancer

BACKGROUND

MR-only radiotherapy treatment planning is an attractive alternative to conventional workflow, reducing scan time and ionizing radiation. It is crucial to derive the electron density map or synthetic CT (sCT) from MR data to perform dose calculations to enable MR-only treatment planning. Automatic segmentation of relevant organs in MR images can accelerate the process by preventing the time-consuming manual contouring step. However, the segmentation label is available only for CT data in many cases.

PURPOSE

We propose CycleSeg, a unified framework that generates sCT and corresponding segmentation from MR images without access to MR segmentation labels METHODS: CycleSeg utilizes the CycleGAN formulation to perform unpaired synthesis of sCT and image alignment. To enable MR (sCT) segmentation, CycleSeg incorporates unsupervised domain adaptation by using a pseudo-labeling approach with feature alignment in semantic segmentation space. In contrast to previous approaches that perform segmentation on MR data, CycleSeg could perform segmentation on both MR and sCT. Experiments were performed with data from prostate cancer patients, with 78/7/10 subjects in the training/validation/test sets, respectively.

RESULTS

CycleSeg showed the best sCT generation results, with the lowest mean absolute error of 102.2 and the lowest Fréchet inception distance of 13.0. CycleSeg also performed best on MR segmentation, with the highest average dice score of 81.0 and 81.1 for MR and sCT segmentation, respectively. Ablation experiments confirmed the contribution of the proposed components of CycleSeg.

CONCLUSION

CycleSeg effectively synthesized CT and performed segmentation on MR images of prostate cancer patients. Thus, CycleSeg has the potential to expedite MR-only radiotherapy treatment planning, reducing the prescribed scans and manual segmentation effort, and increasing throughput.

SIMPLEX: Multiple phase-cycled bSSFP quantitative magnetization transfer imaging with physic-guided simulation learning of neural network

Most quantitative magnetization transfer (qMT) imaging methods require acquiring additional quantitative maps (such as T1) for data fitting. A method based on multiple phase-cycled bSSFP was recently proposed to enable high-resolution 3D qMT imaging based on least square fitting without any extra acquisition, and thus has high potential for simplifying the qMT procedure. However, the quantification of qMT parameters with this method was suboptimal, limiting its potential for clinical application despite its simpler protocol and higher spatial resolution. To improve the fitting of qMT data obtained with multiple phase-cycled bSSFP, we propose SIMulation-based Physics-guided Learning of neural network for qMT parameters EXtraction, or SIMPLEX. In contrast to previous deep learning supervised approaches for quantitative MR that require the acquisition of input data and corresponding ground truth for training, we leveraged the MR signal model to generate training samples without expensive data curation. The network was trained exclusively with simulation data by predicting the simulation parameters. The same network was applied directly to in-vivo data without additional training. The approach was verified with both simulation and in-vivo data. SIMPLEX showed a decrease in fitting mean squared error for all simulation data compared to the existing least-square fitting method. The in-vivo experiment revealed that the network performed well with the real in vivo data unseen during training. For all experiments, we observed that SIMPLEX consistently improved the quantification quality of the qMT parameters whilst being more robust to noise compared to the prior technique. The proposed SIMPLEX will expedite the routine clinical application of qMT by providing qMT parameters (exchange rate, pool fraction) as well as T1, T2, and ΔB0 maps simultaneously with high spatial resolution, better reliability, and reduced processing time.

Simultaneously optimizing sampling pattern for joint acceleration of multi-contrast MRI using model-based deep learning

BACKGROUND

Acceleration of MR imaging (MRI) is a popular research area, and usage of deep learning for acceleration has become highly widespread in the MR community. Joint acceleration of multiple-acquisition MRI was proven to be effective over a single-acquisition approach. Also, optimization in the sampling pattern demonstrated its advantage over conventional undersampling pattern. However, optimizing the sampling patterns for joint acceleration of multiple-acquisition MRI has not been investigated well.

PURPOSE

To develop a model-based deep learning scheme to optimize sampling patterns for a joint acceleration of multi-contrast MRI.

METHODS

The proposed scheme combines sampling pattern optimization and multi-contrast MRI reconstruction. It was extended from the physics-guided method of the joint model-based deep learning (J-MoDL) scheme to optimize the separate sampling pattern for each of multiple contrasts simultaneously for their joint reconstruction. Tests were performed with three contrasts of T2-weighted, FLAIR, and T1-weighted images. The proposed multi-contrast method was compared to (i) single-contrast method with sampling optimization (baseline J-MoDL), (ii) multi-contrast method without sampling optimization, and (iii) multi-contrast method with single common sampling optimization for all contrasts. The optimized sampling patterns were analyzed for sampling location overlap across contrasts. The scheme was also tested in a data-driven scenario, where the inversion between input and label was learned from the under-sampled data directly and tested on knee datasets for generalization test.

RESULTS

The proposed scheme demonstrated a quantitative and qualitative advantage over the single-contrast scheme with sampling pattern optimization and the multi-contrast scheme without sampling pattern optimization. Optimizing the separate sampling pattern for each of the multi-contrasts was superior to optimizing only one common sampling pattern for all contrasts. The proposed scheme showed less overlap in sampling locations than the single-contrast scheme. The main hypothesis was also held in the data-driven situation as well. The brain-trained model worked well on the knee images, demonstrating its generalizability.

CONCLUSION

Our study introduced an effective scheme that combines the sampling optimization and the multi-contrast acceleration. The seamless combination resulted in superior performance over the other existing methods.

Varying undersampling directions for accelerating multiple acquisition magnetic resonance imaging

In this study, we propose a new sampling strategy for efficiently accelerating multiple acquisition MRI. The new sampling strategy is to obtain data along different phase-encoding directions across multiple acquisitions. The proposed sampling strategy was evaluated in multicontrast MR imaging (T1, T2, proton density) and multiple phase-cycled (PC) balanced steady-state free precession (bSSFP) imaging by using convolutional neural networks with central and random sampling patterns. In vivo MRI acquisitions as well as a public database were used to test the concept. Based on both visual inspection and quantitative analysis, the proposed sampling strategy showed better performance than sampling along the same phase-encoding direction in both multicontrast MR imaging and multiple PC-bSSFP imaging, regardless of sampling pattern (central, random) or datasets (public, retrospective and prospective in vivo). For the prospective in vivo applications, acceleration was performed by sampling along different phase-encoding directions at the time of acquisition with a conventional rectangular field of view, which demonstrated the advantage of the proposed sampling strategy in the real environment. Preliminary trials on compressed sensing (CS) also demonstrated improvement of CS with the proposed idea. Sampling along different phase-encoding directions across multiple acquisitions is advantageous for accelerating multiacquisition MRI, irrespective of sampling pattern or datasets, with further improvement through transfer learning.

qMTNet: Accelerated quantitative magnetization transfer imaging with artificial neural networks

PURPOSE

To develop a set of artificial neural networks, collectively termed qMTNet, to accelerate data acquisition and fitting for quantitative magnetization transfer (qMT) imaging.

METHODS

Conventional and interslice qMT data were acquired with two flip angles at six offset frequencies from seven subjects for developing the networks and from four young and four older subjects for testing the generalizability. Two subnetworks, qMTNet-acq and qMTNet-fit, were developed and trained to accelerate data acquisition and fitting, respectively. qMTNet-2 is the sequential application of qMTNet-acq and qMTNet-fit to produce qMT parameters (exchange rate, pool fraction) from undersampled qMT data (two offset frequencies rather than six). qMTNet-1 is one single integrated network having the same functionality as qMTNet-2. qMTNet-fit was compared with a Gaussian kernel-based fitting. qMT parameters generated by the networks were compared with those from ground truth fitted with a dictionary-driven approach.

RESULTS

The proposed networks achieved high peak signal-to-noise ratio (>30) and structural similarity index (>97) in reference to the ground truth. qMTNet-fit produced qMT parameters in concordance with the ground truth with better performance than the Gaussian kernel-based fitting. qMTNet-2 and qMTNet-1 could accelerate data acquisition at threefold and accelerate fitting at 5800- and 4218-fold, respectively. qMTNet-1 showed slightly better performance than qMTNet-2, whereas qMTNet-2 was more flexible for applications.

CONCLUSION

The proposed single (qMTNet-1) and two joint neural networks (qMTNet-2) can accelerate qMT workflow for both data acquisition and fitting significantly. qMTNet has the potential to accelerate qMT imaging for clinical applications, which warrants further investigation.

Bioavailability of octamethylcyclotetrasiloxane (D(4)) after exposure to silicones by inhalation and implantation

We developed a physiologically based pharmacokinetic (PBPK) model to predict the target organ doses of octamethylcyclotetrasiloxane (D(4)) after intravenous (IV), inhalation, or implantation exposures. The model used (14)C-D(4) IV disposition data in rats to estimate tissue distribution coefficients, metabolism, and excretion parameters. We validated the model by comparing the predicted blood and tissues concentrations of D(4) after inhalation to experimental results in both rats and humans. We then used the model to simulate D(4) kinetics after single and/or repeated D(4) exposures in rats and humans. The model predicted bioaccumulation of D(4) in fatty tissues (e.g., breast), especially in women. Because of its high lipid solubility (Log P(oct/water) = 5.1), D(4) persisted in fat with a half life of 11.1 days after inhalation and 18.2 days after breast implant exposure. Metabolism and excretion remained constant with repeated exposures, larger doses, and/or different routes of exposure. The accumulation of D(4) in fatty tissues should play an important role in the risk assessment of D(4) especially in women exposed daily to multiple personal care products and silicone breast implants.

Prediction of the shelf life of cellulose acetate hemodialyzers by Monte Carlo simulation

A previous investigation by our laboratory linked cellulose acetate degradation with adverse health effects in hemodialysis patients. To establish the accumulation of degradation products with time, a Monte Carlo model of degradation kinetics was developed. The model tracks changes in a population of molecules representative of the dialyzer membrane during the degradation process. The degradation calculation is a two step process: First, the model uses a random number to select an individual polymer molecule out of the population, and then a second random number is used to identify a site on the selected molecule for the degradation reaction to occur. After the reaction calculation, the resulting degraded molecules are redistributed into the population. The course of the reaction is determined by recalculating the molecular weight averages in the changing population as the calculations proceed. The model was validated using gel permeation chromatography molecular weight results and total acetyl content measurements on dialyzers stored up to 13.3 years after manufacture. It was found that the degradation reactions can be accurately modeled as random events and that the chain scissions and deacetylation events occur at constant rates. The shelf life of these devices was estimated using the model predictions and animal test results.

Physiologically based pharmacokinetic model for fluorocarbon elimination after the administration of an octafluoropropane-albumin microsphere sonographic contrast agent

A physiologically based pharmacokinetic model was developed to evaluate the kinetics of one of the newest sonographic contrast agents available, FS069 or Optison. This material consists of octafluoropropane gas encapsulated in proteinaceous microspheres, injected intravenously for use as a myocardial contrast agent in humans. This model has six compartments: two lung compartments (alveolar and dead volume), and compartments for the heart, slowly perfused tissue, richly perfused tissue, and gastrointestinal tract. The model was developed to determine the distribution and excretion of the octafluoropropane in the body. Despite the high affinity of octafluoropropane for tissue, the model predicted that nearly 100% of the material would be exhaled from the lungs within 6 min. The model verified the results of a phase I clinical trial with 10 healthy subjects. Ventilation rate was found to play a critical role in the complete excretion of this contrast agent. The physiologically based pharmacokinetic model was a useful tool for evaluating the safety of FS069. This model can be used a basis for developing similar models for other types of contrast agents.

A physiologically based pharmacokinetic model for 2,4-toluenediamine leached from polyurethane foam-covered breast implants

Physiologically based pharmacokinetic (PBPK) modeling was used to assess the low-dose exposure of patients to the carcinogen 2, 4-toluenediamine (2,4-TDA) released from the degradation of the polyester urethane foam (PU) used in Meme silicone breast implants. The tissues are represented as five compartments: liver, kidney, gastrointestinal tract, slowly perfused tissues (e.g., fat), and richly perfused tissues (e.g., muscle). The PBPK model was fitted to the plasma and urine concentrations of 2,4-TDA and its metabolite 4-AAT (4-N-acetyl-2-amino toluene) in rats given low doses of 2, 4-TDA intravenously and subcutaneously. The rat model was extrapolated to simulate oral and implant routes in rats. After adjusting for human physiological parameters, the model was then used to predict the bioavailability of 2,4-TDA released from a typical 4.87-g polyester urethane foam implant found in a patient who weighed 58 kg with the Meme and had the breast implant for 10 years. A quantitative risk assessment for 2,4-TDA was performed and the polyester urethane foam did present an unreasonable risk to health for the patient.

Characterization of [3H]ouabain binding sites in human brain, platelet, and erythrocyte

[3H]Ouabain binding was investigated in membranes prepared from human brain, erythrocyte, and platelet. Scatchard analysis of [3H]ouabain binding to human hypothalamic membranes revealed a single class of noninteracting binding sites with an apparent affinity constant (KD) of 21 nM. Though the number of [3H]ouabain binding sites was lower in human platelets than in erythrocytes, both tissues exhibited a single class of high-affinity binding sites with an apparent KD similar to that found in human brain. Specific [3H]ouabain binding in basal ganglia tissue from patients with Huntington's disease was more than 50% lower than in tissue from age- and sex-matched controls. These results, along with previous findings in rat brain, suggest that high-affinity [3H]ouabain binding labels the neuronal form of Na, K-ATPase in human brain, and may prove useful in quantitating this enzyme in postmortem brain samples.

Characterization of "high-affinity" [3H]ouabain binding in the rat central nervous system

The characteristics of [3H]ouabain binding were examined in various areas of rat brain. In the striatum, Scatchard analysis revealed a single class of "high-affinity" binding sites with an apparent binding affinity (KD) of 10.4 +/- 0.9 nM and an estimated binding capacity (Bmax) of 7.6 +/- 1.9 pmol/mg protein. Similar monophasic Scatchard plots were found in the brainstem, cerebellum, hypothalamus, and frontal cerebral cortex. [3H]Ouabain binding to rat brain was sodium- and ATP-dependent and strongly inhibited by potassium. Proscillariden A was the most potent cardiac glycoside tested in inhibiting specific [3H]ouabain binding to brain membranes, and the rank order of inhibitory potencies for a series of cardiac glycosides was similar to that previously reported for inhibition of heart Na,K-ATPase. To assess whether the high-affinity binding sites for [3H]ouabain were localized to neuronal or nonneuronal membranes, the effect of discrete kainic acid lesions on striatal [3H]ouabain binding was examined. Kainic acid lesions of the striatum reduced [3H]ouabain binding to striatal homogenates by 79.6 +/- 1.6%. This suggests that the "high-affinity" [3H]ouabain binding sites measured in our experiments are localized to neuronal elements. Thus, the high-affinity binding of [3H]ouabain to brain membranes may selectively label a neuronal form or conformation of Na,K-ATPase.