The Impact of Injection Distance to Bifurcations on Yttrium-90 Distribution in Liver Cancer Radioembolization

Simplification strategies for a patient-specific CFD model of particle transport during liver radioembolization

BACKGROUND

Patient-specific 3D computational fluid dynamics (CFD) simulations have been used previously to identify the impact of injection parameters (e.g. injection location, velocity, etc.) on the particle distribution and the tumor dose during transarterial injection of radioactive microspheres for treatment of hepatocellular carcinoma. However, these simulations are computationally costly, so we aim to evaluate whether these can be reliably simplified.

METHODS

We identified and applied five simplification strategies (i.e. truncation, steady flow modelling, moderate and severe grid coarsening, and reducing the number of cardiac cycles) to a patient-specific CFD setup. Subsequently, we evaluated whether these strategies can be used to (1) accurately predict the CFD output (i.e. particle distribution and tumor dose) and (2) quantify the sensitivity of the model output to a specific injection parameter (injection flow rate).

RESULTS

For both accuracy and sensitivity purposes, moderate grid coarsening is the most reliable simplification strategy, allowing to predict the tumor dose with only a maximal deviation of 1.4 %, and a similar sensitivity (deviation of 0.7 %). The steady strategy performs the worst, with a maximal deviation in the tumor dose of 20 % and a difference in sensitivity of 10 %.

CONCLUSION

The patient-specific 3D CFD simulations of this study can be reliably simplified by coarsening the grid, decreasing the computational time by roughly 45 %, which works especially well for sensitivity studies.

Adaptive design of experiments to fit surrogate Gaussian process regression models allows fast sensitivity analysis of the input waveform for patient-specific 3D CFD models of liver radioembolization

BACKGROUND AND OBJECTIVE

Patient-specific 3D computational fluid dynamics (CFD) models are increasingly being used to understand and predict transarterial radioembolization procedures used for hepatocellular carcinoma treatment. While sensitivity analyses of these CFD models can help to determine the most impactful input parameters, such analyses are computationally costly. Therefore, we aim to use surrogate modelling to allow relatively cheap sensitivity analysis. As an example, we compute Sobol's sensitivity indices for three input waveform shape parameters.

METHODS

We extracted three characteristic shape parameters from our input mass flow rate waveform (peak systolic mass flow rate, heart rate, systolic duration) and defined our 3D input parameter space by varying these parameters within 75 %-125 % of their nominal values. To fit our surrogate model with a minimal number of costly CFD simulations, we developed an adaptive design of experiments (ADOE) algorithm. The ADOE uses 100 Latin hypercube sampled points in 3D input space to define the initial design of experiments (DOE). Subsequently, we re-sample input space with 10,000 Latin Hypercube sampled points and cheaply estimate the outputs using the surrogate model. In each of 27 equivolume bins which divide our input space, we determine the most uncertain prediction of the 10,000 points, compute the true outputs using CFD, and add these points to the DOE. For each ADOE iteration, we calculate Sobol's sensitivity indices, and we continue to add batches of 27 samples to the DOE until the Sobol indices have stabilized.

RESULTS

We tested our ADOE algorithm on the Ishigami function and showed that we can reliably obtain Sobol's indices with an absolute error <0.1. Applying ADOE to our waveform sensitivity problem, we found that the first-order sensitivity indices were 0.0550, 0.0191 and 0.407 for the peak systolic mass flow rate, heart rate, and the systolic duration, respectively.

CONCLUSIONS

Although the current study was an illustrative case, the ADOE allows reliable sensitivity analysis with a limited number of complex model evaluations, and performs well even when the optimal DOE size is a priori unknown. This enables us to identify the highest-impact input parameters of our model, and other novel, costly models in the future.

MIDOS: a novel stochastic model towards a treatment planning system for microsphere dosimetry in liver tumors

PURPOSE

Transarterial radioembolization (TARE) procedures treat liver tumors by injecting radioactive microspheres into the hepatic artery. Currently, there is a critical need to optimize TARE towards a personalized dosimetry approach. To this aim, we present a novel microsphere dosimetry (MIDOS) stochastic model to estimate the activity delivered to the tumor(s), normal liver, and lung.

METHODS

MIDOS incorporates adult male/female liver computational phantoms with the hepatic arterial, hepatic portal venous, and hepatic venous vascular trees. Tumors can be placed in both models at user discretion. The perfusion of microspheres follows cluster patterns, and a Markov chain approach was applied to microsphere navigation, with the terminal location of microspheres determined to be in either normal hepatic parenchyma, hepatic tumor, or lung. A tumor uptake model was implemented to determine if microspheres get lodged in the tumor, and a probability was included in determining the shunt of microspheres to the lung. A sensitivity analysis of the model parameters was performed, and radiation segmentectomy/lobectomy procedures were simulated over a wide range of activity perfused. Then, the impact of using different microspheres, i.e., SIR-Sphere®, TheraSphere®, and QuiremSphere®, on the tumor-to-normal ratio (TNR), lung shunt fraction (LSF), and mean absorbed dose was analyzed.

RESULTS

Highly vascularized tumors translated into increased TNR. Treatment results (TNR and LSF) were significantly more variable for microspheres with high particle load. In our scenarios with 1.5 GBq perfusion, TNR was maximum for TheraSphere® at calibration time in segmentectomy/lobar technique, for SIR-Sphere® at 1-3 days post-calibration, and regarding QuiremSphere® at 3 days post-calibration.

CONCLUSION

This novel approach is a decisive step towards developing a personalized dosimetry framework for TARE. MIDOS assists in making clinical decisions in TARE treatment planning by assessing various delivery parameters and simulating different tumor uptakes. MIDOS offers evaluation of treatment outcomes, such as TNR and LSF, and quantitative scenario-specific decisions.

Deep Learning for Computational Hemodynamics: A Brief Review of Recent Advances

Computational fluid dynamics (CFD) modeling of blood flow plays an important role in better understanding various medical conditions, designing more effective drug delivery systems, and developing novel diagnostic methods and treatments. However, despite significant advances in computational technology and resources, the expensive computational cost of these simulations still hinders their transformation from a research interest to a clinical tool. This bottleneck is even more severe for image-based, patient-specific CFD simulations with realistic boundary conditions and complex computational domains, which make such simulations excessively expensive. To address this issue, deep learning approaches have been recently explored to accelerate computational hemodynamics simulations. In this study, we review recent efforts to integrate deep learning with CFD and discuss the applications of this approach in solving hemodynamics problems, such as blood flow behavior in aorta and cerebral arteries. We also discuss potential future directions in the field. In this review, we suggest that incorporating physiologic understandings and underlying fluid mechanics laws in deep learning models will soon lead to a paradigm shift in the development novel non-invasive computational medical decisions.