Patient-specific Predictions Research Articles

Computational modelling of valvular heart disease: haemodynamic insights and clinical implications.

An aging population and an increasing incidence of cardiovascular risk factors form the basis for a global rising prevalence of valvular heart disease (VHD). Research to further our understanding of the pathophysiology of VHD is often confined to the clinical setting. However, in recent years, sophisticated computational models of the cardiovascular system have been increasingly used to investigate a variety of VHD states. Computational modelling provides new opportunities to gain insights into pathophysiological processes that may otherwise be difficult, or even impossible, to attain in human or animal studies. Simulations of co-existing cardiac pathologies, such as heart failure, atrial fibrillation, and mixed valvular disease, have unveiled new insights that can inform clinical research and practice. More recently, advancements have been made in using models for making patient-specific diagnostic predictions. This review showcases valuable insights gained from computational studies on VHD and their clinical implications.

One-year employment outcome prediction after traumatic brain injury: A CENTER-TBI study

BackgroundTraumatic brain injury (TBI) can come with long term consequences for functional outcome that can complicate return to work. ObjectivesThis study aims to make accurate patient-specific predictions on one-year return to work after TBI using machine learning algorithms. Within this process, specific research questions were defined: 1 How can we make accurate predictions on employment outcome, and does this require follow-up data beyond hospitalization? 2 Which predictors are required to make accurate predictions? 3 Are predictions accurate enough for use in clinical practice? MethodsThis study used the core CENTER-TBI observational cohort dataset, collected across 18 European countries between 2014 and 2017. Hospitalized patients with sufficient follow-up data were selected for the current analysis (N = 586). Data regarding hospital stay and follow-up until three months post-injury were used to predict return to work after one year. Three distinct algorithms were used to predict employment outcomes: elastic net logistic regression, random forest and gradient boosting. Finally, a reduced model and corresponding ROC-curve was created. ResultsFull models without follow-up achieved an area under the curve (AUC) of about 81 %, which increased up to 88 % with follow-up data. A reduced model with five predictors achieved similar results with an AUC of 90 %. ConclusionThe addition of three-month follow-up data causes a notable increase in model performance. The reduced model - containing Glasgow Outcome Scale Extended, pre-injury job class, pre-injury employment status, length of stay and age - matched the predictive performance of the full models. Accurate predictions on post-TBI vocational outcomes contribute to realistic prognosis and goal setting, targeting the right interventions to the right patients.

Pharmacogenomic augmented machine learning in electronic health record alerts: A health system-wide usability survey of clinicians.

Pharmacogenomic (PGx) biomarkers integrated using machine learning can be embedded within the electronic health record (EHR) to provide clinicians with individualized predictions of drug treatment outcomes. Currently, however, drug alerts in the EHR are largely generic (not patient-specific) and contribute to increased clinician stress and burnout. Improving the usability of PGx alerts is an urgent need. Therefore, this work aimed to identify principles for optimal PGx alert design through a health-system-wide, mixed-methods study. Clinicians representing multiple practices and care settings (N = 1062) in urban, rural, and underserved regions were invited to complete an electronic survey comparing the usability of three drug alerts for citalopram, as a case study. Alert 1 contained a generic warning of pharmacogenomic effects on citalopram metabolism. Alerts 2 and 3 provided patient-specific predictions of citalopram efficacy with varying depth of information. Primary outcomes included the System's Usability Scale score (0-100 points) of each alert, the perceived impact of each alert on stress and decision-making, and clinicians' suggestions for alert improvement. Secondary outcomes included the assessment of alert preference by clinician age, practice type, and geographic setting. Qualitative information was captured to provide context to quantitative information. The final cohort comprised 305 geographically and clinically diverse clinicians. A simplified, individualized alert (Alert 2) was perceived as beneficial for decision-making and stress compared with a more detailed version (Alert 3) and the generic alert (Alert 1) regardless of age, practice type, or geographic setting. Findings emphasize the need for clinician-guided design of PGx alerts in the era of digital medicine.

Machine Learning versus Cox Models for Predicting Overall Survival in Patients with Osteosarcoma: A Retrospective Analysis of the EURAMOS-1 Clinical Trial Data.

Since the mid-1980s, there has been little progress in improving survival of patients diagnosed with osteosarcoma. Survival prediction models play a key role in clinical decision-making, guiding healthcare professionals in tailoring treatment strategies based on individual patient risks. The increasing interest of the medical community in using machine learning (ML) for predicting survival has sparked an ongoing debate on the value of ML techniques versus more traditional statistical modelling (SM) approaches. This study investigates the use of SM versus ML methods in predicting overall survival (OS) using osteosarcoma data from the EURAMOS-1 clinical trial (NCT00134030). The well-established Cox proportional hazard model is compared to the extended Cox model that includes time-varying effects, and to the ML methods random survival forests and survival neural networks. The impact of eight variables on OS predictions is explored. Results are compared on different model performance metrics, variable importance, and patient-specific predictions. The article provides comprehensive insights to aid healthcare researchers in evaluating diverse survival prediction models for low-dimensional clinical data.

A Computational Pipeline to Investigate Longitudinal Blood Flow Changes in the Circle of Willis of Patients with Stable and Growing Aneurysms

Changes in cerebral blood flow are often associated with the initiation and development of different life-threatening medical conditions including aneurysm rupture and ischemic stroke. Nevertheless, it is not fully clear how haemodynamic changes in time across the Circle of Willis (CoW) are related with intracranial aneurysm (IA) growth. In this work, we introduced a novel reduced-order modelling strategy for the systematic quantification of longitudinal blood flow changes across the whole CoW in patients with stable and unstable/growing aneurysm. Magnetic Resonance Angiography (MRA) images were converted into one-dimensional (1-D) vessel networks through a semi-automated procedure, with a level of geometric reconstruction accuracy controlled by user-dependent parameters. The proposed pipeline was used to systematically analyse longitudinal haemodynamic changes in seven different clinical cases. Our preliminary simulation results indicate that growing aneurysms are not necessarily associated with significant changes in mean flow over time. A concise sensitivity analysis also shed light on which modelling aspects need to be further characterized to have reliable patient-specific predictions. This study poses the basis for investigating how time-dependent changes in the vasculature affect the haemodynamics across the whole CoW in patients with stable and growing aneurysms.

Application of Data-Driven computing to patient-specific prediction of the viscoelastic response of human brain under transcranial ultrasound stimulation.

We present a class of model-free Data-Driven solvers that effectively enable the utilization of in situ and in vivo imaging data directly in full-scale calculations of the mechanical response of the human brain to sonic and ultrasonic stimulation, entirely bypassing the need for analytical modeling or regression of the data. The well-posedness of the approach and its convergence with respect to data are proven analytically. We demonstrate the approach, including its ability to make detailed spatially resolved patient-specific predictions of wave patterns, using public-domain MRI images, MRE data and commercially available solid-mechanics software.

Point of View on Outcome Prediction Models in Post-Stroke Motor Recovery

Stroke is a leading cause of disability worldwide which can cause significant and persistent upper limb (UL) impairment. It is difficult to predict UL motor recovery after stroke and to forecast the expected outcomes of rehabilitation interventions during the acute and subacute phases when using clinical data alone. Accurate prediction of response to treatment could allow for more timely and targeted interventions, thereby improving recovery, resource allocation, and reducing the economic impact of post-stroke disability. Initial motor impairment is currently the strongest predictor of post-stroke motor recovery. Despite significant progress, current prediction models could be refined with additional predictors, and an emphasis on the time dependency of patient-specific predictions of UL recovery profiles. In the current paper a panel of experts provide their opinion on additional predictors and aspects of the literature that can help advance stroke outcome prediction models. Potential strategies include close attention to post-stroke data collection timeframes and adoption of individual-computerized modeling methods connected to a patient’s health record. These models should account for the non-linear and the variable recovery pattern of spontaneous neurological recovery. Additionally, input data should be extended to include cognitive, genomic, sensory, neural injury, and function measures as additional predictors of recovery. The accuracy of prediction models may be further improved by including standardized measures of outcome. Finally, we consider the potential impact of refined prediction models on healthcare costs.

Aleatoric and epistemic uncertainty extraction of patient-specific deep learning-based dose predictions in LDR prostate brachytherapy.

Objective.In brachytherapy, deep learning (DL) algorithms have shown the capability of predicting 3D dose volumes. The reliability and accuracy of such methodologies remain under scrutiny for prospective clinical applications. This study aims to establish fast DL-based predictive dose algorithms for low-dose rate (LDR) prostate brachytherapy and to evaluate their uncertainty and stability.Approach.Data from 200 prostate patients, treated with125I sources, was collected. The Monte Carlo (MC) ground truth dose volumes were calculated with TOPAS considering the interseed effects and an organ-based material assignment. Two 3D convolutional neural networks, UNet and ResUNet TSE, were trained using the patient geometry and the seed positions as the input data. The dataset was randomly split into training (150), validation (25) and test (25) sets. The aleatoric (associated with the input data) and epistemic (associated with the model) uncertainties of the DL models were assessed.Main results.For the full test set, with respect to the MC reference, the predicted prostateD90metric had mean differences of -0.64% and 0.08% for the UNet and ResUNet TSE models, respectively. In voxel-by-voxel comparisons, the average global dose difference ratio in the [-1%, 1%] range included 91.0% and 93.0% of voxels for the UNet and the ResUNet TSE, respectively. One forward pass or prediction took 4 ms for a 3D dose volume of 2.56 M voxels (128 × 160 × 128). The ResUNet TSE model closely encoded the well-known physics of the problem as seen in a set of uncertainty maps. The ResUNet TSE rectum D2cchad the largest uncertainty metric of 0.0042.Significance.The proposed DL models serve as rapid dose predictors that consider the patient anatomy and interseed attenuation effects. The derived uncertainty is interpretable, highlighting areas where DL models may struggle to provide accurate estimations. The uncertainty analysis offers a comprehensive evaluation tool for dose predictor model assessment.

Predicting Immunotherapy Outcomes in Glioblastoma Patients through Machine Learning.

Glioblastoma is a highly aggressive cancer associated with a dismal prognosis, with a mere 5% of patients surviving beyond five years post diagnosis. Current therapeutic modalities encompass surgical intervention, radiotherapy, chemotherapy, and immune checkpoint inhibitors (ICBs). However, the efficacy of ICBs remains limited in glioblastoma patients, necessitating a proactive approach to anticipate treatment response and resistance. In this comprehensive study, we conducted a rigorous analysis involving two distinct glioblastoma patient cohorts subjected to PD-1 blockade treatments. Our investigation revealed that a significant portion (60%) of patients exhibit persistent disease progression despite ICB intervention. To elucidate the underpinnings of resistance, we characterized the immune profiles of glioblastoma patients with continued cancer progression following anti-PD1 therapy. These profiles revealed multifaceted defects, encompassing compromised macrophage, monocyte, and T follicular helper responses, impaired antigen presentation, aberrant regulatory T cell (Tregs) responses, and heightened expression of immunosuppressive molecules (TGFB, IL2RA, and CD276). Building upon these resistance profiles, we leveraged cutting-edge machine learning algorithms to develop predictive models and accompanying software. This innovative computational tool achieved remarkable success, accurately forecasting the progression status of 82.82% of the glioblastoma patients in our study following ICBs, based on their unique immune characteristics. In conclusion, our pioneering approach advocates for the personalization of immunotherapy in glioblastoma patients. By harnessing patient-specific attributes and computational predictions, we offer a promising avenue for the enhancement of clinical outcomes in the realm of immunotherapy. This paradigm shift towards tailored therapies underscores the potential to revolutionize the management of glioblastoma, opening new horizons for improved patient care.

PERSONALIZED STATISTICAL MODELLING OF MENISCAL KINEMATICS

Intra-articular cartilage pressure distribution in the knee joint is critical in the understanding of osteoarthritis. Combining personalized statistical modeling of the morphological characteristics with discrete element modeling enables patient-specific predictions of the pressure on the tibial plateau. However, modeling of the meniscus during gait is complicated by the dynamic nature of the structure. Nevertheless, the position of the meniscus has a substantial impact on intra-articular stress distribution. Therefore, the focus of this presentation will be on how modeling of meniscal movement during knee flexion improves insight in general meniscal kinematics for the use in tibiofemoral stress distribution calculations.

Predicting recurrent interventions after radiocephalic arteriovenous fistula creation with machine learning and the PREDICT-AVF web app.

Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines discourage ongoing access salvage attempts after two interventions prior to successful use or more than three interventions per year overall. The goal was to develop a tool for prediction of radiocephalic arteriovenous fistula (AVF) intervention requirements to help guide shared decision-making about access appropriateness. Prospective cohort study of 914 adult patients in the United States and Canada undergoing radiocephalic AVF creation at one of the 39 centers participating in the PATENCY-1 or -2 trials. Clinical data, including demographics, comorbidities, access history, anatomic features, and post-operative ultrasound measurements at 4-6 and 12 weeks were used to predict recurrent interventions required at 1 year postoperatively. Cox proportional hazards, random survival forest, pooled logistic, and elastic net recurrent event survival prediction models were built using a combination of baseline characteristics and post-operative ultrasound measurements. A web application was created, which generates patient-specific predictions contextualized with the KDOQI guidelines. Patients underwent an estimated 1.04 (95% CI 0.94-1.13) interventions in the first year. Mean (SD) age was 57 (13) years; 22% were female. Radiocephalic AVFs were created at the snuffbox (2%), wrist (74%), or proximal forearm (24%). Using baseline characteristics, the random survival forest model performed best, with an area under the receiver operating characteristic curve (AUROC) of 0.75 (95% CI 0.67-0.82) at 1 year. The addition of ultrasound information to baseline characteristics did not substantially improve performance; however, Cox models using either 4-6- or 12-week post-operative ultrasound information alone had the best discrimination performance, with AUROCs of 0.77 (0.70-0.85) and 0.76 (0.70-0.83) at 1 year. The interactive web application is deployed at https://predict-avf.com. The PREDICT-AVF web application can guide patient counseling and guideline-concordant shared decision-making as part of a patient-centered end-stage kidney disease life plan.

Outcome Prediction Model Following Proximal Femoral Osteotomy in Legg-Calvé-Perthes Disease Using Machine Learning Algorithms.

The purpose of the current study was (1) to analyze various factors that may be associated with the outcomes of Legg-Calvé-Perthes disease (LCPD), and (2) to develop and internally validate machine learning algorithms capable of providing patient-specific predictions of which patients with LCPD will achieve relevant improvement in radiologic outcomes after proximal femoral varus osteotomy (PFVO). We examined several variables, previously identified as factors, that may influence the outcome of LCPD and developed a machine learning algorithm based on them. In this retrospective study, we analyzed patients aged older than 6 years at the time of LCPD diagnosis who underwent PFVO at our institution between 1979 and 2015. Univariate and multivariate logistic regression analyses were used to examine the effects of variables on the sphericity of the femoral head at skeletal maturity, including age at onset, sex, stage at operation, extent of epiphyseal involvement and collapse, presence of specific epiphyseal, metaphyseal, and acetabular changes, and postoperative neck shaft angle (NSA). Recursive feature selection was used to identify the combination of variables from an initial pool of 13 features that optimized the model performance. Five machine learning algorithms [extreme gradient boosting (XGBoost), multilayer perception, support vector machine, elastic-net penalized logistic regression, and random forest) were trained using 5-fold cross-validation 3 times and applied to an independent testing set of patients. Ninety patients with LCPD who underwent PFVO were included in this study. The mean age at diagnosis was 7.93 (range, 6.0 to 12.33) years. The average follow-up period was 10.11 (range, 5.25 to 22.92) years. A combination of 8 variables, optimized algorithm performance, and specific cutoffs were found to decrease the likelihood of achieving the 1 or 2 Stulberg classification: age at onset ≤ 8.06, lateral classification ≤ B, 12.40 < preoperative migration percentage (MP) ≤ 22.85, Catterall classification ≤ 2, 117.4 < postoperative NSA ≤ 122.90, -10.8 < postoperative MP ≤ 6.5, 139.65 < preoperative NSA ≤ 144.67, and operation at stage 1. The XGBoost model demonstrated the best performance (F1 score: 0.78; area under the curve: 0.84). The XGBoost machine learning algorithm achieved the best performance in predicting the postoperative radiologic outcomes in patients with LCPD who underwent PFVO. In our population, age at onset ≤ 8.06, lateral classification ≤ B, 12.40 < preoperative MP ≤ 22.85, Catterall classification ≤ 2, 117.4 < postoperative NSA ≤ 122.90, -10.8 < postoperative MP ≤ 6.5, 139.65 < preoperative NSA ≤ 144.67, and operation at an early stage had the likelihood of achieving the spherical femoral head for the patients with LCPD that underwent PFVO. After external validation, the online application of this model may enhance shared decision-making. Level III-retrospective cohort study.

Predicting return to work after traumatic brain injury using machine learning and administrative data

BackgroundAccurate patient-specific predictions on return-to-work after traumatic brain injury (TBI) can support both clinical practice and policymaking. The use of machine learning on large administrative data provides interesting opportunities to create such prognostic models. AimThe current study assesses whether return-to-work one year after TBI can be predicted accurately from administrative data. Additionally, this study explores how model performance and feature importance change depending on whether a distinction is made between mild and moderate-to-severe TBI. MethodsThis study used a population-based dataset that combined discharge, claims and social security data of patients hospitalized with a TBI in Belgium during the year 2016. The prediction of TBI was attempted with three algorithms, elastic net logistic regression, random forest and gradient boosting and compared in their performance by their accuracy, sensitivity, specificity and area under the receiver operator curve (ROC AUC). ResultsThe distinct modelling algorithms resulted in similar results, with 83% accuracy (ROC AUC 85%) for a binary classification of employed vs. not employed and up to 76% (ROC AUC 82%) for a multiclass operationalization of employment outcome. Modelling mild and moderate-to-severe TBI separately did not result in considerable differences in model performance and feature importance. The features of main importance for return-to-work prediction were related to pre-injury employment. DiscussionWhile clearly offering some information beneficial for predicting return-to-work, administrative data needs to be supplemented with additional information to allow further improvement of patient-specific prognose.

Towards Prediction of Blood Flow in Coiled Aneurysms Before Treatment: A Porous Media Approach.

Modeling blood flow in aneurysms treated with coils could be used to understand the complete embolization of the aneurysm, through thrombus formation that fills the entire sac. Modeling of the endovascular coil mass as a porous medium is a technique that allows for study of aneurysm hemodynamics, efficiently for patient-specific treatment outcome predictions. Models in the literature use mean porosity of coils in the aneurysmal volume, proving inadequate for outcome prediction. However, models that consider heterogeneous porosity distribution have shown more accurate hemodynamics. We recently published the porous crown model, considering the heterogeneous coil mass distribution, validated on two patients. This study aims (i) to validate the porous crown model for a larger cohort (eight patients), and (ii) to propose a porous medium model translatable to clinical practice in treatment planning. We analyzed the porosity distribution of the endovascular coils deployed inside the cerebral aneurysm phantoms of eight patients using 3D x-ray synchrotron images. The permeability and inertial factor of the porous crown model are calculated using previously published methodology. We propose a new "bilinear" porous model, that uses the same hypothesis, but the permeability and inertial factor can be defined from just basic information available in the neuro-suite, i.e., the aneurysmal sac volume and the coil volume fraction targeted by the neurosurgeon. These two models are compared to the coil-resolved simulations, considered as the gold standard. The results show that both the porous crown model and the bilinear model produce similarly accurate hemodynamics in the aneurysm. The error in the standard (mean porosity) porous model is 66%, whereas the error of the bilinear model is 26%, compared to the coil-resolved. The bilinear model is promising as a means of treatment outcome prediction at time of intervention.

A fluid–solid-growth solver for cardiovascular modeling

We implement full, three-dimensional constrained mixture theory for vascular growth and remodeling into a finite element fluid–structure interaction (FSI) solver. The resulting “fluid–solid-growth” (FSG) solver allows long term, patient-specific predictions of changing hemodynamics, vessel wall morphology, tissue composition, and material properties. This extension from short term (FSI) to long term (FSG) simulations increases clinical relevance by enabling mechanobioloigcally-dependent studies of disease progression in complex domains.

Improved Electrical Impedance Tomography Reconstruction via a Bayesian Approach With an Anatomical Statistical Shape Model.

electrical impedance tomography (EIT) is a promising technique for rapid and continuous bedside monitoring of lung function. Accurate and reliable EIT reconstruction of ventilation requires patient-specific shape information. However, this shape information is often not available and current EIT reconstruction methods typically have limited spatial fidelity. This study sought to develop a statistical shape model (SSM) of the torso and lungs and evaluate whether patient-specific predictions of torso and lung shape could enhance EIT reconstructions in a Bayesian framework. torso and lung finite element surface meshes were fitted to computed tomography data from 81 participants, and a SSM was generated using principal component analysis and regression analyses. Predicted shapes were implemented in a Bayesian EIT framework and were quantitatively compared to generic reconstruction methods. Five principal shape modes explained 38% of the cohort variance in lung and torso geometry, and regression analysis yielded nine total anthropometrics and pulmonary function metrics that significantly predicted these shape modes. Incorporation of SSM-derived structural information enhanced the accuracy and reliability of the EIT reconstruction as compared to generic reconstructions, demonstrated by reduced relative error, total variation, and Mahalanobis distance. As compared to deterministic approaches, Bayesian EIT afforded more reliable quantitative and visual interpretation of the reconstructed ventilation distribution. However, no conclusive improvement of reconstruction performance using patient specific structural information was observed as compared to the mean shape of the SSM. The presented Bayesian framework builds towards a more accurate and reliable method for ventilation monitoring via EIT.

Intentional deep overfit learning for patient-specific dose predictions in adaptive radiotherapy.

The framework of adaptive radiation therapy (ART) was crafted to address the underlying sources of intra-patient variation that were observed throughout numerous patients' radiation sessions. ART seeks to minimize the consequential dosimetric uncertainty resulting from this daily variation, commonly through treatment planning re-optimization. Re-optimization typically consists of manual evaluation and modification of previously utilized optimization criteria. Ideally, frequent treatment plan adaptation through re-optimization on each day's computed tomography (CT) scan may improve dosimetric accuracy and minimize dose delivered to organs at risk (OARs) as the planning target volume (PTV) changes throughout the course of treatment. Re-optimization in its current form is time-consuming and inefficient. In response to this ART bottleneck, we propose a deep learning based adaptive dose prediction model that utilizes a head and neck (H&N) patient's initial planning data to fine-tune a previously trained population model towards a patient-specific model. Our fine-tuned, patient-specific (FT-PS) model, which is trained using the intentional deep overfit learning (IDOL) method, may enable clinicians and treatment planners to rapidly evaluate relevant dosimetric changes daily and re-optimize accordingly. An adaptive population (AP) model was trained using adaptive data from 33 patients. Separately, 10 patients were selected for training FT-PS models. The previously trained AP model was utilized as the base model weights prior to re-initializing model training for each FT-PS model. Ten FT-PS models were separately trained by fine-tuning the previous model weights based on each respective patient's initial treatment plan. From these 10 patients, 26 ART treatment plans were withheld from training as the test dataset for retrospective evaluation of dose prediction performance between the AP and FT-PS models. Each AP and FT-PS dose prediction was compared against the ground truth dose distribution as originally generated during the patient's course of treatment. Mean absolute percent error (MAPE) evaluated the dose differences between a model's prediction and the ground truth. MAPE was calculated within the 10% isodose volume region of interest for each of the AP and FT-PS models dose predictions and averaged across all test adaptive sessions, yielding 5.759% and 3.747% respectively. MAPE differences were compared between AP and FT-PS models across each test session in a test of statistical significance. The differences were statistically significant in a paired t-test with two-tailed p-value equal to and 95% confidence interval (CI) equal to [-2.483, -1.542]. Furthermore, MAPE was calculated using each individually segmented structure as an ROI. Nineteen of 24 structures demonstrated statistically significant differences between the AP and FT-PS models. We utilized the IDOL method to fine-tune a population-based dose prediction model into an adaptive, patient-specific model. The averaged MAPE across the test dataset was 5.759% for the population-based model versus 3.747% for the fine-tuned, patient-specific model, and the difference in MAPE between models was found to be statistically significant. Our work demonstrates the feasibility of patient-specific models in adaptive radiotherapy, and offers unique clinical benefit by utilizing initial planning data that contains the physician's treatment intent.

MSR27 Predicting Return to Work after Traumatic Brain Injury: A Center-TBI Study

This study aims to make patient-specific predictions on one-year return to work after traumatic brain injury using machine learning algorithms. Accurate predictions can support both clinical practice and policymaking, while extending the very limited knowledge on determinants of employment outcomes after traumatic brain injury (TBI).

Predicting Recurrent Interventions After Radiocephalic Arteriovenous Fistula Creation with Machine Learning and the PREDICT-AVF Web App

Radiocephalic arteriovenous fistulas (AVFs) are often criticized for requiring numerous recurrent interventions to achieve and maintain hemodialysis access function. Recent Kidney Disease Outcomes Quality Initiative guidelines discourage ongoing access salvage attempts after two interventions prior to successful use or more than three interventions per year overall. This study's goal was to develop a useful prediction tool for preoperative and postoperative assessment of intervention requirements to help guide decision-making about access appropriateness and futility. The data source was patient-level data from the international multicenter PATENCY-1 and PATENCY-2 randomized controlled trials, which enrolled patients undergoing new radiocephalic AVF creation. Routine ultrasounds were performed at 4 to 6 weeks and 12 weeks from AVF creation and interpreted by a core laboratory. The outcome for the prediction models was the number of recurrent surgical or endovascular interventions required at 1 year. Cox, random survival forest, pooled logistic, and penalized pooled logistic (elastic net) recurrent event survival prediction models were built using a combination of baseline characteristics and postoperative ultrasound measurements. Models were trained with 10-fold cross-validation hyperparameter selection within a randomly split training dataset. Discrimination and calibration performance was assessed on a holdout testing dataset. An interactive web application was built, which generates patient-specific predictions at 1 year, draws cumulative hazard curves, and contextualizes the predictions with the Kidney Disease Outcomes Quality Initiative guidelines for recurrent interventions after access creation. The cohort included 914 patients; mean age was 57 ± 13 years and 22% were female. Radiocephalic AVFs were created at the snuffbox (2%), wrist (74%), or proximal forearm (24%). Patients underwent a median of 1.04 (95% confidence interval [CI], 0.95-1.13) interventions in the first year. Using only baseline characteristics, the random survival forest model performed best, with a C-statistic of 0.72 (95% CI, 0.65-0.78) at 1 year (Table). The addition of ultrasound information to baseline characteristics did not substantially improve performance; however, Cox models using either 4- to 6-week or 12-week postoperative ultrasound information alone had the best discrimination performance, with C-statistics of 0.74 (95% CI, 0.67-0.81) and 0.78 (95% CI, 0.72-0.84) at 1 year (Table). The interactive web application is deployed at PREDICT-AVF.com (Figure). When assessing candidates for radiocephalic AVF in the preoperative setting, baseline patient characteristics can predict future interventions with moderate performance. In the postoperative setting, ultrasound findings alone are sufficient to predict future interventions. A free point-of-care web application can guide patient counseling and guideline-concordant surgeon decision-making as part of a patient-centered end-stage kidney disease life plan.TableDiscrimination performance of each model and feature set, measured by the area under the receiver operating characteristic curve (C-statistic) at 1 year postoperativelyFeaturesCoxRandom survival ForestPooled logistic regressionElastic netBaseline0.704 (0.639-0.769)0.715 (0.647-0.783)0.696 (0.631-0.762)0.703 (0.638-0.769)Baseline + 4- to 6-week US0.704 (0.629-0.779)0.737 (0.666-0.808)0.702 (0.627-0.778)0.703 (0.627-0.778)Baseline + 12 week US0.737 (0.670-0.803)0.766 (0.701-0.832)0.730 (0.663-0.797)0.746 (0.680-0.812)4- to 6-Week US0.743 (0.674-0.813)0.696 (0.620-0.773)0.740 (0.670-0.810)0.742 (0.672-0.812)12-Week US0.782 (0.722-0.842)0.721 (0.652-0.791)0.774 (0.712-0.835)0.775 (0.714-0.837)US, Ultrasound.Perfect discrimination corresponds to an area under the receiver operating characteristic curve of 1. Feature sets included baseline (age, sex, race, smoking status, diabetes, heart failure, renal transplant history, anticoagulation use, antiplatelet use, statin use, hemodialysis status at the time of arteriovenous fistula creation, central venous catheter history, number of prior failed accesses, vein and artery diameter, anastomotic suture technique, access location, and access side), baseline with the addition of postoperative ultrasound measurements (flow volume, access diameter, and luminal stenosis), and ultrasound measurements alone. Open table in a new tab

A Microstructure-Based Mechanistic Approach to Detect Degeneration Effects on Potential Damage Zones and Morphology of Young and Old Human Intervertebral Discs.

There is an increasing demand to develop predictive medicine through the creation of predictive models and digital twins of the different body organs. To obtain accurate predictions, real local microstructure, morphology changes and their accompanying physiological degenerative effects must be taken into account. In this article, we present a numerical model to estimate the long-term aging effect on the human intervertebral disc response by means of a microstructure-based mechanistic approach. It allows to monitor in-silico the variations in disc geometry and local mechanical fields induced by age-dependent long-term microstructure changes. Both lamellar and interlamellar zones of the disc annulus fibrosus are constitutively represented by considering the main underlying microstructure features in terms of proteoglycans network viscoelasticity, collagen network elasticity (along with content and orientation) and chemical-induced fluid transfer. With age,anoticeable increase in shear strain is especially observed in the posterior and lateral posterior regions of the annulus which is in correlation with the high vulnerability of elderly people to back problems and posterior disc hernia. Important insights about the relation between age-dependent microstructure features, disc mechanics and disc damage are revealed using the present approach. These numerical observations are hardly obtainable using current experimental technologies whichmakes our numerical tool useful for patient-specific long-term predictions.