Model Parameters Research Articles

Long-term motor training enhances functional connectivity between semantic and motor regions in an effector-specific manner: evidence from elite female football athletes.

The relation between the action verb semantic processing and sensorimotor experience remains controversial. In this study, we examined whether plasticity changes in brain are specifically related to semantic processing of foot action verbs when long-term motor training is mainly aimed at the foot. To address this question, we acquired resting-state functional magnetic resonance imaging scans and behavioral data from a verb two-choice task from female expertise football players and football novices. We compared the resting-state functional connectivity (rsFC) differences between experts and novices using motor execution regions and general semantic regions (left anterior temporal lobe, lATL) as seed, and explored the neural correlates of behavioral performance. Here, the drift rate (v) parameter of the drift diffusion model (DDM) was used to capture the semantic processing capability. We found experts showed increased correlation between lATL subregions and important brain regions for motor processing, including supplementary motor area (SMA), bilateral paracentral lobule (PL), superior parietal lobule and inferior parietal lobule, in contrast to novices. Further predictive model analysis showed the FC found in rsFC analysis can significantly predict drift rate of foot action verb in both experts and novices, but not drift rate of hand action verb. Our findings therefore establish a connection between effector-related semantic processing and the plasticity changes in brain functional connectivity, attributable to long-term foot-related motor training. This provides evidence supporting the view that semantic processing is fundamentally rooted in the sensorimotor system.

Estimation of some AM2 model parameters from a derived empirical logistic function of methane production

Because of its capability to convert organic wastes into renewable energy and into some components useful for agriculture, the anaerobic digestion technology can reduce greenhouse gas emissions in the atmosphere and the pollution. Thus, anaerobic digestion can contribute to achieving some of sustainable development goals. Consequently, many theoretical and empirical approaches are proposed for estimating, predicting and optimizing the methane produced by anaerobic digestion. In this context, the logistic function is a mathematical model that can be used to approximate empirical data of the temporal methane production in anaerobic digestion. In a previous paper, under some appropriate approximations, we have derived from AM2 model a single analytical expression in a form of a logistic function for describing the evolution of methane production in batch bioreactors. In the present paper, by comparing the three standard parameters associated with the classical empirical logistic function with that of the derived one from AM2 model; some relationships between them have been established. These relations are exploited for estimating some coefficients and parameters of AM2 model with respect to empiric logistic function parameters and vice-versa. Moreover, this possibility enables more qualitative insight about the evolution of the methane production and the influence of AM2 parameters and coefficients as well as their interaction over its processes.

Individualized epidemic spreading models predict epilepsy surgery outcomes: A pseudo-prospective study

Abstract Epilepsy surgery is the treatment of choice for drug-resistant epilepsy patients, but up to 50% of patients continue to have seizures one year after the resection. In order to aid presurgical planning and predict postsurgical outcome on a patient-by-patient basis, we developed a framework of individualized computational models that combines epidemic spreading with patient-specific connectivity and epileptogeneity maps: the Epidemic Spreading Seizure and Epilepsy Surgery framework (ESSES). ESSES parameters were fitted in a retrospective study (N = 15) to reproduce invasive electroencephalography (iEEG)-recorded seizures. ESSES reproduced the iEEG-recorded seizures, and significantly better so for patients with good (seizure-free, SF) than bad (nonseizure-free, NSF) outcome. We illustrate here the clinical applicability of ESSES with a pseudo-prospective study (N = 34) with a blind setting (to the resection strategy and surgical outcome) that emulated presurgical conditions. By setting the model parameters in the retrospective study, ESSES could be applied also to patients without iEEG data. ESSES could predict the chances of good outcome after any resection by finding patient-specific model-based optimal resection strategies, which we found to be smaller for SF than NSF patients, suggesting an intrinsic difference in the network organization or presurgical evaluation results of NSF patients. The actual surgical plan overlapped more with the model-based optimal resection, and had a larger effect in decreasing modeled seizure propagation, for SF patients than for NSF patients. Overall, ESSES could correctly predict 75% of NSF and 80.8% of SF cases pseudo-prospectively. Our results show that individualised computational models may inform surgical planning by suggesting alternative resections and providing information on the likelihood of a good outcome after a proposed resection. This is the first time that such a model is validated with a fully independent cohort and without the need for iEEG recordings.

Comparação das distribuições Normal, Logística, Laplace e t de Student para o erro experimental na descrição bayesiana do acúmulo de matéria seca de Allium sativum

ABSTRACT: This study assessed distributions associated with Bayesian nonlinear modeling error in the description of total plant dry matteraccumulation (TDMA) of Allium sativumas a function of days after planting (DAP). According to the DIC criterion, Logistic and Gompertzmodels that use student’s t distribution error exhibited the highest DIC with logistic error distribution. In general, the difference of DIC in all the scenarios was not more than 5.The Bayes factor (BF) criterion showed no difference in the Logistic and Gompertzmodel fit when four distributions are used for the errors, where BF values do not exceed 2. Posterior distributions and the usual estimators of Logistic and Gompertz model parameters were similar even forvaried error distribution. In summary, there was no difference in the use of 4 distributions associated with the modeling error of garlic plant growth by the Bayes factor, whereby the results showed that alternating between error distributions significantly changes the number of Markov Chain Monte Carlo (MCMC) iterations.

Reducing Energy Consumption in the Synthesis of Dimethyl Ether (DME) from Methanol Dehydration by Modifying Heat Transfer Unit Using Aspen HYSYS

In the pursuit of a transport sector free from fossil fuels, the incorporation of Dimethyl ether (DME) emerges as a commendable ecological substitute. The DME is a synthetically produced serves as a viable alternative to conventional fuels such as diesel and liquified petroleum gas (LPG). The Dimethyl Ether (DME) production is carried out by catalytic dehydration of methanol over an acidic zeolite-based catalyst. The technological process for the DME synthesis was simulated using Aspen HYSYS based on the combined operating parameters of the reaction dynamic model for the methanol dehydration reaction. This paper attempts to evaluate a modification in dehydration of methanol process to reducing energy consumption by modifying heat transfer unit. The heater and cooler heat transfer units were converted into heat exchangers (HE) by utilizing the output of the process that can be used for other processes so that energy consumption is reduced. The temperature of reaction and the heat transfer unit are modified to reduce energy consumption from 11.850 MMBtu/h to 7.6291 MMBtu/h by changing one heater and one cooler with two heat exchangers. Copyright © 2024 by Authors, Published by Universitas Diponegoro and BCREC Publishing Group. This is an open access article under the CC BY-SA License (https://creativecommons.org/licenses/by-sa/4.0).

Black and White Image Colorization with Deep Learning

Abstract: The project "Black and White Image Colorization with Deep Learning" develops a system to automatically infuse colors into grayscale images through cutting-edge deep learning approaches. Utilizing convolutional neural networks (CNNs) and generative adversarial networks (GANs), the system aims to deliver both realistic and visually pleasing colorization outcomes. This process includes training the deep learning models on diverse image datasets, optimizing the model parameters, and using advanced loss functions to ensure precise color representation.This project addresses the complex challenges associated with adding colors to monochrome images and also seeks to improve user experience by introducing customizable options such as style modification and color tweaks. By merging deep learning with computer vision techniques, this initiative is poised to contribute significantly to various image processing fields, especially in the restoration of historical images and the enhancement of visual content production. With thorough experimentation and ongoing refinements, the project aspires to establish a reliable and easy-to-use system that advances the state of image colorization technology in the modern digital age..

PSF and MTF from a bar pattern in digital mammography

Background. The MTF has difficulties being determined (according to the provisions of the IEC standards) in the hospital setting due to the lack of resources. Purpose. The objective of this work is to propose a quantitative method for obtaining the point spread function (PSF) and the modulation transfer function (MTF) of a digital mammography system from an image of a bar pattern. Methods. The method is based on the measurement of the contrast transfer function (CTF) of the system over the image of the bar pattern. In addition, a theoretical model for the PSF is proposed, from which the theoretical CTF of the system is obtained by means of convolution with a square wave (mathematical simulation of the bar pattern). Through an iterative process, the free parameters of the PSF model are varied until the experimental CTF coincides with the one calculated by convolution. Once the PSF of the system is obtained, we calculate the MTF by means of its Fourier transform. The MTF calculated from the model PSF have been compared with those calculated from an image of a 65 μm diameter gold wire using an oversampling process. Results. The CTF has been calculated for three digital mammographic systems (DMS 1, DMS 2 and DMS 3), no differences of more than 5 % were found with the CTF obtained with the PSF model. The comparison of the MTF shows us the goodness of the PSF model. Conclusions. The proposed method for obtaining PSF and MTF is a simple and accessible method, which does not require a complex configuration or the use of phantoms that are difficult to access in the hospital world. In addition, it can be used to calculate other magnitudes of interest such as the normalized noise power spectrum (NNPS) and the detection quantum efficiency (DQE).

Mechanical properties of recycled aggregate concrete containing brick particles

To investigate the mechanical properties of recycled aggregate concrete containing brick particles (RACB), compressive strength tests were conducted on RACBs with different water/cement ratios and recycled fine aggregate substitution rates. The changes in compressive strength, elastic modulus and peak strain of RACB were analysed as a function of the proportion of substituted recycled fine aggregate, and a stress–strain equation was established. A damage model was established using ultrasonic non-destructive testing technology to reveal the damage mechanism of RACB. Results show that the basic mechanical properties of RACB are best when the replacement proportion of recycled fine aggregate (r) is 30%. The model parameter a of the stress–strain curve equation is consistent with the change of elastic modulus with r. The model parameter b has a quadratic function relationship with r, reflecting the brittleness of RACB. The initial damage of RACB is relatively large, and cracks penetrate the recycled coarse aggregate, leading to specimen failure.

Tensor low-rank and sparse decomposition and its application in bearing fault information separation

Abstract Properly separating fault information from noisy measured signals is crucial for effective bearing health sensing. However, conventional fault information separation methods face challenges such as predefined model parameters and poor noise robustness. Additionally, with the advent of Industry Big Data, multichannel monitoring signals present significant challenges for traditional single decomposition approaches. To address these challenges and fully extract potential fault information, this paper introduces a tensor low-rank and sparse decomposition (tensor LRSD) approach for multichannel signal processing. Inspired by matrix LRSD, we construct a tensor LRSD model that adaptively decomposes the signal into a tensor sparse term containing fault information and a low-rank term representing the intrinsic signal pattern. To further enhance the decomposition performance, a maximum correlation-based selection strategy is designed. This strategy evaluates the correlation between each tensor slice and selects appropriate tensor sparse terms for fault information extraction. Simulation analysis and two experimental studies involving typical bearing failures are implemented to verify the capability and superiority of the presented tensor LRSD approach. The consequences demonstrate that the presented method outperforms conventional techniques, showcasing its capability to effectively separate fault information from noisy signals.

Deciphering oxygen distribution and hypoxia profiles in the tumor microenvironment: a data-driven mechanistic modeling approach

Objective. The distribution of hypoxia within tissues plays a critical role in tumor diagnosis and prognosis. Recognizing the significance of tumor oxygenation and hypoxia gradients, we introduce mathematical frameworks grounded in mechanistic modeling approaches for their quantitative assessment within a tumor microenvironment. By utilizing known blood vasculature, we aim to predict hypoxia levels across different tumor types. Approach. Our approach offers a computational method to measure and predict hypoxia using known blood vasculature. By formulating a reaction-diffusion model for oxygen distribution, we derive the corresponding hypoxia profile. Main results. The framework successfully replicates observed inter- and intra-tumor heterogeneity in experimentally obtained hypoxia profiles across various tumor types (breast, ovarian, pancreatic). Additionally, we propose a data-driven method to deduce partial differential equation models with spatially dependent parameters, which allows us to comprehend the variability of hypoxia profiles within tissues. The versatility of our framework lies in capturing diverse and dynamic behaviors of tumor oxygenation, as well as categorizing states of vascularization based on the dynamics of oxygen molecules, as identified by the model parameters. Significance. The proposed data-informed mechanistic method quantitatively assesses hypoxia in the tumor microenvironment by integrating diverse histopathological data and making predictions across different types of data. The framework provides valuable insights from both modeling and biological perspectives, advancing our comprehension of spatio-temporal dynamics of tumor oxygenation.

Fast Calculation of Gaussian Process Multiple-Fold Cross-Validation Residuals and their Covariances

We generalize fast Gaussian process leave-one-out formulas to multiple-fold cross-validation, highlighting in turn the covariance structure of cross-validation residuals in simple and universal kriging frameworks. We illustrate how resulting covariances affect model diagnostics. We further establish in the case of noiseless observations that correcting for covariances between residuals in cross-validation-based estimation of the scale parameter leads back to maximum likelihood estimation. Also, we highlight in broader settings how differences between pseudo-likelihood and likelihood methods boil down to accounting or not for residual covariances. The proposed fast calculation of cross-validation residuals is implemented and benchmarked against a naive implementation, all in R. Numerical experiments highlight the substantial speed-ups that our approach enables. However, as supported by a discussion on main drivers of computational costs and by a numerical benchmark, speed-ups steeply decline as the number of folds (say, all sharing the same size) decreases. An application to a contaminant localization test case illustrates that the way of grouping observations in folds may affect model assessment and parameter fitting compared to leave-one-out. Overall, our results enable fast multiple-fold cross-validation, have consequences in model diagnostics, and pave the way to future work on hyperparameter fitting as well as on goal-oriented fold design. Supplementary materials for this article are available online.

Identifying low-dimensional trajectories of mechanically-ventilated patient systems: Empirical phenotypes of joint patient+care processes to enhance temporal analysis in ARDS research.

Refined management of mechanically ventilation is an obvious target for improving patient outcomes, but is impeded by the nature of data for study and hypothesis generation. The connections between clinical outcomes and temporal development of iatrogenic injuries current lung-protective ventilator settings remain poorly understood. Analysis of lung-ventilator system (LVS) evolution at relevant timescales is frustrated by data volume and multiple sources of heterogeneity. This work motivates, presents, and validates a computational pipeline for resolving LVS systems into the joint evolution of data-conditioned model parameters and ventilator information. Applied to individuals, the workflow yields a concise low-dimensional representation of LVS behavior expressed in phenotypic breath waveforms suitable for analysis. The effectiveness of this approach is demonstrated through application to multi-day observational series of 35 patients. Individual patient analyses reveal multiple types of patient-oriented dynamics and breath behavior to expose the complexity of LVS evolution; less than 10% of phenotype changes related to ventilator settings changes. Dynamics are shown to including both stable and unstable phenotype transitions as well as both discrete and continuous changes unrelated to ventilator settings. At a cohort scale, 721 phenotypes constructed from individual data are condensed into a set of 16 groups that empirically organize around certain settings (positive end-expository pressure and ventilator mode) and structurally similar pressure-volume loop characterizations. Individual and cohort scale phenotypes, which may be refined by hypothesis-specific constructions, provide a common framework for ongoing temporal analysis and investigation of LVS dynamics.

Diagnostically distinct resting state fMRI energy distributions: A subject-specific maximum entropy modeling study.

Objective Existing neuroimaging studies of psychotic and mood disorders have reported brain activation differences (first-order properties) and altered pairwise correlation based functional connectivity (second-order properties). However, both approaches have certain limitations that can be overcome by integrating them in a pairwise maximum entropy model (MEM) that better represents a comprehensive picture of fMRI signal patterns and provides a system-wide summary measure called energy. This study examines the applicability of individual-level MEM for psychiatry and identifies image-derived model coefficients related to model parameters. Method MEMs are fit to resting state fMRI data from each individual with schizophrenia/schizoaffective disorder, bipolar disorder, and major depression (n=132) and demographically matched healthy controls (n=132) from the UK Biobank to different subsets of the default mode network (DMN) regions. Results The model satisfactorily explained observed brain energy state occurrence probabilities across all participants, and model parameters were significantly correlated with image-derived coefficients for all groups. Within clinical groups, averaged energy level distributions were higher in schizophrenia/schizoaffective disorder but lower in bipolar disorder compared to controls for both bilateral and unilateral DMN. Major depression energy distributions were higher compared to controls only in the right hemisphere DMN. Conclusions Diagnostically distinct energy states suggest that probability distributions of temporal changes in synchronously active nodes may underlie each diagnostic entity. Subject-specific MEMs allow for factoring in the individual variations compared to traditional group-level inferences, offering an improved measure of biologically meaningful correlates of brain activity that may have potential clinical utility.

Investigating the confinement effects on masonry behavior based on friction interface

As a material masonry is anisotropic in nature as it is constituted of various components that differ each other in many ways. Though, the ideas of modeling of concrete are also valid for masonry, still there are difficulties in determining the exact behavior of masonry. The micro modeling (simplified) approach has been adopted in this study for modeling masonry samples subjected to various nature of loadings. The brick masonry units were modeled as discrete elements and linked together by interface element. The coulomb friction model was used to describe the interface element as mortar joint. Damage was assumed to be governed by the compressive strength of both the constituents i.e., mortar and brick. The mechanical properties of the masonry samples were determined by means of variety of tests and those properties then served as input parameters in the constitutive model. The results clearly indicate that numerical model predicted the behavior of the experimentally tested samples, in compression as well as tension prims for both unconfined and confined cases. The proposed model is suitable for modeling any type of brick masonry walls.

Determining tissue distribution of the oral antileishmanial agent miltefosine: a physiologically-based pharmacokinetic modeling approach.

Miltefosine (MTS) is the only approved oral drug for treating leishmaniasis caused by intracellular Leishmania parasites that localize in macrophages of the liver, spleen, skin, bone marrow, and lymph nodes. MTS is extensively distributed in tissues and has prolonged elimination half-lives due to its high plasma protein binding, slow metabolic clearance, and minimal urinary excretion. Thus, understanding and predicting the tissue distribution of MTS help assess therapeutic and toxicologic outcomes of MTS, especially in special populations, e.g., pediatrics. In this study, a whole-body physiologically-based pharmacokinetic (PBPK) model of MTS was built on mice and extrapolated to rats and humans. MTS plasma and tissue concentration data obtained by intravenous and oral administration to mice were fitted simultaneously to estimate model parameters. The resulting high tissue-to-plasma partition coefficient values corroborate extensive distribution in all major organs except the bone marrow. Sensitivity analysis suggests that plasma exposure is most susceptible to changes in fraction unbound in plasma. The murine oral-PBPK model was further validated by assessing overlay of simulations with plasma and tissue profiles obtained from an independent study. Subsequently, the murine PBPK model was extrapolated to rats and humans based on species-specific physiological and drug-related parameters, as well as allometrically scaled parameters. Fold errors for pharmacokinetic parameters were within acceptable range in both extrapolated models, except for a slight underprediction in the human plasma exposure. These animal and human PBPK models are expected to provide reliable estimates of MTS tissue distribution and assist dose regimen optimization in special populations.

Significance of Dynamic Axial Stretching on Estimating Biomechanical Behavior and Properties of the Human Ascending Aorta.

Contrary to most vessels, the ascending thoracic aorta (ATA) not only distends but also elongates in the axial direction. The purpose of this study is to investigate the biomechanical behavior of the ascending thoracic aorta (ATA) in response to dynamic axial stretching during the cardiac cycle. In addition, the implications of neglecting this dynamic axial stretching when estimating the constitutive model parameters of the ATA are investigated. The investigations were performed through in silico simulations by assuming a Gasser-Ogden-Holzapfel (GOH) constitutive model representative of ATA tissue material. The GOH model parameters were obtained from biaxial tests performed on four human ATA tissues in a previous study. Pressure-diameter curves were simulated as synthetic data to assess theeffect of neglecting dynamic axial stretching on estimating constitutive model parameters. Our findings reveal a significant increase in axial stress (~ 16%) and stored strain energy (~ 18%) in the vessel when dynamic axial stretching is considered, as opposed to assuming a fixed axial stretch. All but one artery showed increased volume compliance while considering a dynamic axial stretching condition. Furthermore, we observe a notable difference in the estimated constitutive model parameters when dynamic axial stretching of the ATA is neglected, compared to the ground truth model parameters. These results underscore the critical importance of accounting for axial deformations when conducting in vivo biomechanical characterization of the ascending thoracic aorta.

Dynamic simulation-assisted Gaussian mixture alignment approach for fault diagnosis of rotation machinery under small samples

Abstract Obtaining a substantial number of actual samples for rotating machinery in an industrial setting can be challenging, particularly when faulty samples are acquired under hazardous working conditions. The issue of insufficient samples hinders the effective training of reliable fault diagnosis models, impeding the industrial implementation of advanced intelligent methods. This study proposes an innovative Dynamic Simulation-assisted Gaussian Mixture Alignment model (DSGMA) to address the challenge of applying fault diagnosis technologies, with its performance mined by advanced transfer algorithms. Specifically, we establish a fault dynamics model for rotating machinery and acquire a substantial amount of simulated data as the source domain to facilitate the training of the deep neural network model. Subsequently, we propose a Gaussian mixture-guided domain alignment approach that assigns a domain-independent Gaussian distribution to each category as prior knowledge, with the parameters calculated using limited actual samples. Diagnostic knowledge is transferred from the source domain to the target domain by minimizing the Kullback-Leibler divergence between the features of the simulated samples and the Gaussian mixture priors. Furthermore, the DSGMA model incorporates Gaussian clustering loss to augment the clustering capability of samples belonging to the same category from real devices and enhances the computational stability of the parameters in the Gaussian mixture model. The efficacy of the DSGMA method is validated using three publicly available datasets and compared against five widely adopted methods. The experimental findings illustrate that DSGMA exhibits superior diagnostic and robust capabilities, facilitating efficient fault diagnosis under scenarios of small samples.

Superexchange Mechanism in Coupled Triangulenes Forming Spin-1 Chains.

We show that the origin of the antiferromagnetic coupling in spin-1 triangulene chains, which were recently synthesized and measured by Mishra et al. ( Nature 2021, 598, 287-292), originates from a superexchange mechanism. This process, mediated by intertriangulene states, opens the possibility to control parameters in the effective bilinear-biquadratic spin model. We start from the derivation of an effective tight-binding model for triangulene chains using a combination of tight-binding and Hartree-Fock methods fitted to hybrid density functional theory results. Next, correlation effects are investigated within the configuration interaction method. Our low-energy many-body spectrum for NTr = 2 and NTr = 4 triangulene chains agree well with the bilinear-biquadratic spin-1 chain antiferromagnetic model when indirect coupling processes and superexchange coupling between triangulene spins are taken into account.

Membrane models as a means of propulsion in general relativity: super-luminal warp-drive that satisfies the weak energy condition

Presented are toy-models for sub-luminal and super-luminal warp-drives in 3+1 dimensions. The models are constructed in a chimeric manner—as different bulk space-times separated by thin membranes. The membranes contain perfect-fluid-like stress-energy tensors. The Israel junction conditions relate this stress-energy to a jump in extrinsic curvature across the brane, which in turn manifests as apparent acceleration in the bulk space-times. The acceleration on either side of the brane may be set individually by choice of model parameters. The Weak Energy Condition (WEC) is shown to be satisfied everywhere in both models. Although the branes in these toy models are not compact, it is demonstrated that super-luminal warp-drive is possible that satisfies the WEC. Additionally, the nature of these models provides framework for speculation on a mechanism for transition from sub-luminal to super-luminal warp. It is shown that the difference in extrinsic curvature across a thin membrane can yield a positive contribution to the Landau–Raychaudhuri equation, thus providing a means to evade some super-luminal warp-drive no-go theorems. Neither quantum effects nor stability of the models is considered.

Predicting treatment response in multicenter non-small cell lung cancer patients based on federated learning

BackgroundMulticenter non-small cell lung cancer (NSCLC) patient data is information-rich. However, its direct integration becomes exceptionally challenging due to constraints involving different healthcare organizations and regulations. Traditional centralized machine learning methods require centralizing these sensitive medical data for training, posing risks of patient privacy leakage and data security issues. In this context, federated learning (FL) has attracted much attention as a distributed machine learning framework. It effectively addresses this contradiction by preserving data locally, conducting local model training, and aggregating model parameters. This approach enables the utilization of multicenter data with maximum benefit while ensuring privacy safeguards. Based on pre-radiotherapy planning target volume images of NSCLC patients, a multicenter treatment response prediction model is designed by FL for predicting the probability of remission of NSCLC patients. This approach ensures medical data privacy, high prediction accuracy and computing efficiency, offering valuable insights for clinical decision-making.MethodsWe retrospectively collected CT images from 245 NSCLC patients undergoing chemotherapy and radiotherapy (CRT) in four Chinese hospitals. In a simulation environment, we compared the performance of the centralized deep learning (DL) model with that of the FL model using data from two sites. Additionally, due to the unavailability of data from one hospital, we established a real-world FL model using data from three sites. Assessments were conducted using measures such as accuracy, receiver operating characteristic curve, and confusion matrices.ResultsThe model’s prediction performance obtained using FL methods outperforms that of traditional centralized learning methods. In the comparative experiment, the DL model achieves an AUC of 0.718/0.695, while the FL model demonstrates an AUC of 0.725/0.689, with real-world FL model achieving an AUC of 0.698/0.672.ConclusionsWe demonstrate that the performance of a FL predictive model, developed by combining convolutional neural networks (CNNs) with data from multiple medical centers, is comparable to that of a traditional DL model obtained through centralized training. It can efficiently predict CRT treatment response in NSCLC patients while preserving privacy.