Modeling the efficacy of different anti-angiogenic drugs on treatment of solid tumors using 3D computational modeling and machine learning

An Asynchronous Spiking Neural Membrane System for Edge Detection

Spiking neural membrane systems (SN P systems) are a class of bio-inspired models inspired by the activities and connectivity of neurons. Extensive studies have been made on SN P systems with synchronization-based communication, while further efforts are needed for the systems with rhythm-based communication. In this work, we design an asynchronous SN P system with resonant connections where all the enabled neurons in the same group connected by resonant connections should instantly produce spikes with the same rhythm. In the designed system, each of the three modules implements one type of the three operations associated with the edge detection of digital images, and they collaborate each other through the resonant connections. An algorithm called EDSNP for edge detection is proposed to simulate the working of the designed asynchronous SN P system. A quantitative analysis of EDSNP and the related methods for edge detection had been conducted to evaluate the performance of EDSNP. The performance of the EDSNP in processing the testing images is superior to the compared methods, based on the quantitative metrics of accuracy, error rate, mean square error, peak signal-to-noise ratio and true positive rate. The results indicate the potential of the temporal firing and the proper neuronal connections in the SN P system to achieve good performance in edge detection.

Robust Federated Learning for Heterogeneous Model and Data

Data privacy and security is an essential challenge in medical clinical settings, where individual hospital has its own sensitive patients data. Due to recent advances in decentralized machine learning in Federated Learning (FL), each hospital has its own private data and learning models to collaborate with other trusted participating hospitals. Heterogeneous data and models among different hospitals raise major challenges in robust FL, such as gradient leakage, where participants can exploit model weights to infer data. Here, we proposed a robust FL method to efficiently tackle data and model heterogeneity, where we train our model using knowledge distillation and a novel weighted client confidence score on hematological cytomorphology data in clinical settings. In the knowledge distillation, each participant learns from other participants by a weighted confidence score so that knowledge from clean models is distributed other than the noisy clients possessing noisy data. Moreover, we use symmetric loss to reduce the negative impact of data heterogeneity and label diversity by reducing overfitting the model to noisy labels. In comparison to the current approaches, our proposed method performs the best, and this is the first demonstration of addressing both data and model heterogeneity in end-to-end FL that lays the foundation for robust FL in laboratories and clinical applications.

Optimizing the therapeutic window of sirolimus by monitoring blood concentration for the treatment of immune thrombocytopenia

Previous studies have demonstrated that sirolimus (SRL) is an effective agent for the treatment of refractory/relapsed (R/R) ITP. However, the therapeutic window of sirolimus in the treatment of ITP has not been established. As the toxicity of sirolimus increases with higher blood concentrations, it is crucial to determine the optimal therapeutic concentration of SRL for the treatment of ITP. Thus, in this study, we used a retrospective cohort of ITP patients treated with sirolimus to propose the therapeutic dosage window for sirolimus. A total of 275 laboratory results of SRL blood concentration from 63 ITP patients treated with SRL were analyzed retrospectively. The ITP patients were divided into five groups based on their SRL blood concentration: 0-4 ng/ml, 4-8 ng/ml, 8-12 ng/ml, 12-16 ng/ml and ≥16 ng/ml. In addition to the SRL blood concentration, platelet counts and adverse events that occurred during the first 6 weeks of SRL treatment were analyzed. These findings were then used to establish the decision matrix tables and ROC curves, which helped identify the therapeutic window of SRL. Based on the values and trends of true-positive rate (TPR) and false-positive rate (FPR) in the ROC curve, patients who achieved a SRL blood concentration of 4-12 ng/ml displayed a higher response rate compared to those with a SRL concentration of 0-4 ng/ml or ≥16ng/ml. Additionally, the response rate was better for patients with a SRL concentration of 8-12 ng/ml compared to 4-8 ng/ml. Adverse events were related to the concentration of SRL; however, there was no significant difference in the incidence of adverse events between the concentrations of 4-8 ng/ml and 8-12 ng/ml (P > .05). Regression analysis suggested that the concentration of SRL correlated with the patient's age, PLT count at the start of SRL administration, and the dose of SRL. It is suggested that the optimal blood concentration of SRL monotherapy for managing ITP is 8-12 ng/ml. This range may achieve a favorable balance between clinical efficacy and the severity of adverse events.

Accelerating antibody discovery and design with artificial intelligence: Recent advances and prospects

Therapeutic antibodies are the largest class of biotherapeutics and have been successful in treating human diseases. However, the design and discovery of antibody drugs remains challenging and time-consuming. Recently, artificial intelligence technology has had an incredible impact on antibody design and discovery, resulting in significant advances in antibody discovery, optimization, and developability. This review summarizes major machine learning (ML) methods and their applications for computational predictors of antibody structure and antigen interface/interaction, as well as the evaluation of antibody developability. Additionally, this review addresses the current status of ML-based therapeutic antibodies under preclinical and clinical phases. While many challenges remain, ML may offer a new therapeutic option for the future direction of fully computational antibody design.

Multi-dimensional modeling of nanoparticles transportation from capillary bed into the tumor microenvironment

In this study, we examined the extravasation of pharmaceutical inorganic nanoparticles (NPs) with a new approach from the leaky endothelium of tumor microvasculature (TMV) into the tumor microenvironment (TME) multi-dimensionally. We proposed a combination of prevailing macroscopic and microscopic methods and addressed the effect of interstitial fluid (IF) retention in solid tumor as an imperative parameter in drug delivery modeling. The Navier-Stokes equations and Darcy's law were utilized for blood flow and porous media, and the Starling's law was brought in for coupling effect. The blood flow was simulated as a non-Newtonian fluid alongside the Newtonian IF. We applied the Galerkin finite element method for the simulations. Our parametric study includes examining the effect of IF retention and TMV pressure on the distribution of tumor interstitial fluid pressure (TIFP), NPs concentration, and diameter on the penetration process, together with the time effect, on two-dimensional (2D) delivery of NPs. Our findings indicate that the IF retention in tumor cells increases TIFP depending on the amount of TMV pressure and IF retained. In addition to doubling pressure in the tumor necrotic region rather than the rest of TME, it enhances the TIFP which is an important parameter in drug delivery to solid tumors. By decreasing pressure drop within the TMV, pressure distribution within the TME becomes more uniform, creating a better condition for homogeneous penetration of NPs. Increasing both inlet pressure and NPs concentration leads to a nonlinear increase in the average concentration of tumor. Decreasing the diameter of NPs increases the penetration of NPs with a higher ratio in the TME.

Convection-Enhanced Delivery of Antiangiogenic Drugs and Liposomal Cytotoxic Drugs to Heterogeneous Brain Tumor for Combination Therapy

Simple Summary

Although developed anticancer drugs have shown desirable effects in preclinical trials, the clinical efficacy of chemotherapy against brain cancer remains disappointing. One of the important obstacles is the highly heterogeneous environment in tumors. This study aims to evaluate the performance of an emerging treatment using antiangiogenic and cytotoxic drugs. Our mathematical modelling confirms the advantage of this combination therapy in homogenizing the intratumoral environment for better drug delivery outcomes. In addition, the effects of local microvasculature and cell density on this therapy are also discussed. The results would contribute to the development of more effective treatments for brain cancer.

Abstract

Although convection-enhanced delivery can successfully bypass the blood-brain barrier, its clinical performance remains disappointing. This is primarily attributed to the heterogeneous intratumoral environment, particularly the tumor microvasculature. This study investigates the combined convection-enhanced delivery of antiangiogenic drugs and liposomal cytotoxic drugs in a heterogeneous brain tumor environment using a transport-based mathematical model. The patient-specific 3D brain tumor geometry and the tumor’s heterogeneous tissue properties, including microvascular density, porosity and cell density, are extracted from dynamic contrast-enhanced magnetic resonance imaging data. Results show that antiangiogenic drugs can effectively reduce the tumor microvascular density. This change in tissue structure would inhibit the fluid loss from the blood to prevent drug concentration from dilution, and also reduce the drug loss by blood drainage. The comparisons between different dosing regimens demonstrate that the co-infusion of liposomal cytotoxic drugs and antiangiogenic drugs has the advantages of homogenizing drug distribution, increasing drug accumulation, and enlarging the volume where tumor cells can be effectively killed. The delivery outcomes are susceptible to the location of the infusion site. This combination treatment can be improved by infusing drugs at higher microvascular density sites. In contrast, infusion at a site with high cell density would lower the treatment effectiveness of the whole brain tumor. Results obtained from this study can deepen the understanding of this combination therapy and provide a reference for treatment design and optimization that can further improve survival and patient quality of life.

Replacement in angiogenesis research: Studying mechanisms of blood vessel development by animal-free in vitro, in vivo and in silico approaches

Angiogenesis, the development of new blood vessels from pre-existing ones, is an essential process determining numerous physiological and pathological conditions. Accordingly, there is a high demand for research approaches allowing the investigation of angiogenic mechanisms and the assessment of pro- and anti-angiogenic therapeutics. The present review provides a selective overview and critical discussion of such approaches, which, in line with the 3R principle, all share the common feature that they are not based on animal experiments. They include in vitro assays to study the viability, proliferation, migration, tube formation and sprouting activity of endothelial cells in two- and three-dimensional environments, the degradation of extracellular matrix compounds as well as the impact of hemodynamic forces on blood vessel formation. These assays can be complemented by in vivo analyses of microvascular network formation in the chorioallantoic membrane assay and early stages of zebrafish larvae. In addition, the combination of experimental data and physical laws enables the mathematical modeling of tissue-specific vascularization, blood flow patterns, interstitial fluid flow as well as oxygen, nutrient and drug distribution. All these animal-free approaches markedly contribute to an improved understanding of fundamental biological mechanisms underlying angiogenesis. Hence, they do not only represent essential tools in basic science but also in early stages of drug development. Moreover, their advancement bears the great potential to analyze angiogenesis in all its complexity and, thus, to make animal experiments superfluous in the future.

Comparative Analysis of Machine Learning and Numerical Modeling for Combined Heat Transfer in Polymethylmethacrylate

This study has compared different methods to predict the simultaneous effects of conductive and radiative heat transfer in a polymethylmethacrylate (PMMA) sample. PMMA is a type of polymer utilized in various sensors and actuator devices. One-dimensional combined heat transfer is considered in numerical analysis. Computer implementation was obtained for the numerical solution of the governing equation with the implicit finite difference method in the case of discretization. Kirchhoff transformation was used to obtain data from a non-linear equation of conductive heat transfer by considering monochromatic radiation intensity and temperature conditions applied to the PMMA sample boundaries. For the deep neural network (DNN) method, the novel long short-term memory (LSTM) method was introduced to find accurate results in the least processing time compared to the numerical method. A recent study derived the combined heat transfer and temperature profiles for the PMMA sample. Furthermore, the transient temperature profile was validated by another study. A comparison proves the perfect agreement. It shows the temperature gradient in the primary positions, which provides a spectral amount of conductive heat transfer from the PMMA sample. It is more straightforward when they are compared with the novel DNN method. Results demonstrate that this artificial intelligence method is accurate and fast in predicting problems. By analyzing the results from the numerical solution, it can be understood that the conductive and radiative heat flux are similar in the case of gradient behavior, but the amount is also twice as high approximately. Hence, total heat flux has a constant value in an approximated steady-state condition. In addition to analyzing their composition, the receiver operating characteristic (ROC) curve and confusion matrix were implemented to evaluate the algorithm’s performance.