A neural ordinary differential equation model for visualizing deep neural network behaviors in multi-parametric MRI-based glioma segmentation.

PURPOSE: To develop a neural ordinary differential equation (ODE) model for visualizing deep neural network behavior during multi-parametric MRI-based glioma segmentation as a method to enhance deep learning explainability. METHODS: By hypothesizing that deep feature extraction can be modeled as a spatiotemporally continuous process, we implemented a novel deep learning model, Neural ODE, in which deep feature extraction was governed by an ODE parameterized by a neural network. The dynamics of (1) MR images after interactions with the deep neural network and (2) segmentation formation can thus be visualized after solving the ODE. An accumulative contribution curve (ACC) was designed to quantitatively evaluate each MR image's utilization by the deep neural network toward the final segmentation results. The proposed Neural ODE model was demonstrated using 369 glioma patients with a 4-modality multi-parametric MRI protocol: T1, contrast-enhanced T1 (T1-Ce), T2, and FLAIR. Three Neural ODE models were trained to segment enhancing tumor (ET), tumor core (TC), and whole tumor (WT), respectively. The key MRI modalities with significant utilization by deep neural networks were identified based on ACC analysis. Segmentation results by deep neural networks using only the key MRI modalities were compared to those using all four MRI modalities in terms of Dice coefficient, accuracy, sensitivity, and specificity. RESULTS: All Neural ODE models successfully illustrated image dynamics as expected. ACC analysis identified T1-Ce as the only key modality in ET and TC segmentations, while both FLAIR and T2 were key modalities in WT segmentation. Compared to the U-Net results using all four MRI modalities, the Dice coefficient of ET (0.784→0.775), TC (0.760→0.758), and WT (0.841→0.837) using the key modalities only had minimal differences without significance. Accuracy, sensitivity, and specificity results demonstrated the same patterns. CONCLUSION: The Neural ODE model offers a new tool for optimizing the deep learning model inputs with enhanced explainability. The presented methodology can be generalized to other medical image-related deep-learning applications.

Post-Radiotherapy PET Image Outcome Prediction by Deep Learning Under Biological Model Guidance: A Feasibility Study of Oropharyngeal Cancer Application.

Purpose: To develop a method of biologically guided deep learning for post-radiation 18FDG-PET image outcome prediction based on pre-radiation images and radiotherapy dose information. Methods: Based on the classic reaction-diffusion mechanism, a novel biological model was proposed using a partial differential equation that incorporates spatial radiation dose distribution as a patient-specific treatment information variable. A 7-layer encoder-decoder-based convolutional neural network (CNN) was designed and trained to learn the proposed biological model. As such, the model could generate post-radiation 18FDG-PET image outcome predictions with breakdown biological components for enhanced explainability. The proposed method was developed using 64 oropharyngeal patients with paired 18FDG-PET studies before and after 20-Gy delivery (2 Gy/day fraction) by intensity-modulated radiotherapy (IMRT). In a two-branch deep learning execution, the proposed CNN learns specific terms in the biological model from paired 18FDG-PET images and spatial dose distribution in one branch, and the biological model generates post-20-Gy 18FDG-PET image prediction in the other branch. As in 2D execution, 718/233/230 axial slices from 38/13/13 patients were used for training/validation/independent test. The prediction image results in test cases were compared with the ground-truth results quantitatively. Results: The proposed method successfully generated post-20-Gy 18FDG-PET image outcome prediction with breakdown illustrations of biological model components. Standardized uptake value (SUV) mean values in 18FDG high-uptake regions of predicted images (2.45 ± 0.25) were similar to ground-truth results (2.51 ± 0.33). In 2D-based Gamma analysis, the median/mean Gamma Index (<1) passing rate of test images was 96.5%/92.8% using the 5%/5 mm criterion; such result was improved to 99.9%/99.6% when 10%/10 mm was adopted. Conclusion: The developed biologically guided deep learning method achieved post-20-Gy 18FDG-PET image outcome predictions in good agreement with ground-truth results. With the breakdown biological modeling components, the outcome image predictions could be used in adaptive radiotherapy decision-making to optimize personalized plans for the best outcome in the future.

Capillary-Driven Rise of Well-Wetting Liquid on the Outer Surface of Cylindrical Nozzles.

Well-wetting liquids exiting small-diameter nozzles in the dripping regime can partially rise up along the outer nozzle surfaces. This is problematic for fuel injectors and other devices such as direct-contact heat and mass exchangers that incorporate arrays of nozzles to distribute liquids. We report our experimental and numerical study of the rising phenomenon for wide ranges of parameters. Our study shows that the interplay of three dimensionless numbers (the Bond number, the Weber number, and the Ohnesorge number) governs the capillary-driven rise dynamics. In general, as the flow rate or the viscosity increases, the capillary-driven rise height over each dripping period becomes smaller. We identify liquid flow rates below which the temporal evolution of the meniscus positions can be well approximated by a quasistatic model based on the Young-Laplace equation. Our analysis reveals two critical Bond numbers that give nozzle sizes, which correspond to the maximum meniscus rise and the onset of capillary-driven rise cessation. These critical Bond numbers are characterized as a function of the contact angle, regardless of the fluid type. Our study leads to a more efficient and optimized nozzle design in systems using wetting liquids by reducing both the risks of contamination and high pressure drop in such devices.

Experimental study of a string-based counterflow wet electrostatic precipitator for collection of fine and ultrafine particles.

Wet electrostatic precipitators (WESP) have been widely studied for collecting fine and ultrafine particles, such as diesel particulate matter (DPM), which have deleterious effects on human health. Here, we report an experimental and numerical simulation study on a novel string-based two-stage WESP. Our new design incorporates grounded vertically aligned polymer strings, along which thin films of water flow down. The water beads, generated by intrinsic flow instability, travel down the strings and collect charged particles in the counterflowing gas stream. We performed experiments using two different geometric configurations of WESP: rectangular and cylindrical. We examined the effects of the WESP electrode bias voltage, air stream velocity, and water flow rate on the number-based fractional collection efficiency for particles of diameters ranging from 10 nm to 2.5 µm. The collection efficiency improves with increasing bias voltages or decreasing airflow rates. At liquid-to-gas (L/G) as low as approximately 0.0066, our design delivers a collection efficiency over 70% even for fine and ultrafine particles. The rectangular and cylindrical configurations exhibit similar collection efficiencies under nominally identical experimental conditions. We also compare the water-to-air mass flow rate ratio, air flow rate per unit collector volume, and collection efficiency of our string-based design with those of previously reported WESPs. The present work demonstrates a promising design for a highly efficient, compact, and scalable two-stage WESPs with minimal water consumption.Implications: Wet Electrostatic Precipitators (WESPs) are highly effective for collecting fine particles in exhaust air streams from various sources such as diesel engines, power plants, and oil refineries. However, their large-scale adoption has been limited by high water usage and reduced collection efficiencies for ultrafine particles. We perform experimental and numerical investigation to characterize the collection efficiency and water flow rate-dependence of a new design of WESP. The string-based counterflow WESP reported in this study offers number-based collection efficiencies >70% at air flow rates per collector volume as high as 4.36 (m3/s)/m3 for particles of diameters ranging from 10 nm - 2.5 µm, while significantly reducing water usage. Our work provides a basis for the design of more compact and water-efficient WESPs.