Search

Abstract Many patients with diabetes struggle with post-meal high blood glucose due to missed or untimely meal-related insulin doses. To address this challenge, our research aims to: (1) study mealtime patterns in patients with type 1 diabetes using wearable insulin pump data, and (2) develop personalized models for predicting future mealtimes to support timely insulin dose administration. Using two independent datasets with over 45,000 meal logs from 82 patients with diabetes, we find that the majority of people ( $$\sim$$ ∼ 60%) have irregular and inconsistent mealtime patterns that change notably through the course of each day and across months in their own historical data. We also show the feasibility of predicting future mealtimes with personalized LSTM-based models that achieve an average F1 score of > 95% with less than 0.25 false positives per day. Our research lays the groundwork for developing a meal prediction system that can nudge patients with diabetes to administer bolus insulin doses before meal consumption to reduce the occurrence of post-meal high blood glucose.

Many patients with diabetes struggle with post-meal high blood glucose due to missed or untimely meal-related insulin doses. To address this challenge, our research aims to: (1) study mealtime patterns in patients with type 1 diabetes using wearable insulin pump data, and (2) develop personalized models for predicting future mealtimes to support timely insulin dose administration. Using two independent datasets with over 45,000 meal logs from 82 patients with diabetes, we find that the majority of people (∼ 60%) have irregular and inconsistent mealtime patterns that change notably through the course of each day and across months in their own historical data. We also show the feasibility of predicting future mealtimes with personalized LSTM-based models that achieve an average F1 score of > 95% with less than 0.25 false positives per day. Our research lays the groundwork for developing a meal prediction system that can nudge patients with diabetes to administer bolus insulin doses before meal consumption to reduce the occurrence of post-meal high blood glucose. [ABSTRACT FROM AUTHOR]

Abstract Background Personalized modeling is a promising tool to improve abdominal aortic aneurysm (AAA) rupture risk assessment. Computed tomography (CT) and quantitative flow (Q-flow) magnetic resonance imaging (MRI) are widely regarded as the gold standard for acquiring patient-specific geometry and velocity profiles, respectively. However, their frequent utilization is hindered by various drawbacks. Ultrasound is used extensively in current clinical practice and offers a safe, rapid and cost-effective method to acquire patient-specific geometries and velocity profiles. This study aims to extract and validate patient-specific velocity profiles from Doppler ultrasound and to examine the impact of the velocity profiles on computed hemodynamics. Methods Pulsed-wave Doppler (PWD) and color Doppler (CD) data were successfully obtained for six volunteers and seven patients and employed to extract the flow pulse and velocity profile over the cross-section, respectively. The US flow pulses and velocity profiles as well as generic Womersley profiles were compared to the MRI velocities and flows. Additionally, CFD simulations were performed to examine the combined impact of the velocity profile and flow pulse. Results Large discrepancies were found between the US and MRI velocity profiles over the cross-sections, with differences for US in the same range as for the Womersley profile. Differences in flow pulses revealed that US generally performs best in terms of maximum flow, forward flow and ratios between forward and backward flow, whereas it often overestimates the backward flow. Both spatial patterns and magnitude of the computed hemodynamics were considerably affected by the prescribed velocity boundary conditions. Larger errors and smaller differences between the US and generic CFD cases were observed for patients compared to volunteers. Conclusion These results show that it is feasible to acquire the patient-specific flow pulse from PWD data, provided that the PWD acquisition could be performed proximal to the aneurysm region, and resulted in a triphasic flow pattern. However, obtaining the patient-specific velocity profile over the cross-section using CD data is not reliable. For the volunteers, utilizing the US flow profile instead of the generic flow profile generally resulted in improved performance, whereas this was the case in more than half of the cases for the patients.

Background: Personalized modeling is a promising tool to improve abdominal aortic aneurysm (AAA) rupture risk assessment. Computed tomography (CT) and quantitative flow (Q-flow) magnetic resonance imaging (MRI) are widely regarded as the gold standard for acquiring patient-specific geometry and velocity profiles, respectively. However, their frequent utilization is hindered by various drawbacks. Ultrasound is used extensively in current clinical practice and offers a safe, rapid and cost-effective method to acquire patient-specific geometries and velocity profiles. This study aims to extract and validate patient-specific velocity profiles from Doppler ultrasound and to examine the impact of the velocity profiles on computed hemodynamics. Methods: Pulsed-wave Doppler (PWD) and color Doppler (CD) data were successfully obtained for six volunteers and seven patients and employed to extract the flow pulse and velocity profile over the cross-section, respectively. The US flow pulses and velocity profiles as well as generic Womersley profiles were compared to the MRI velocities and flows. Additionally, CFD simulations were performed to examine the combined impact of the velocity profile and flow pulse. Results: Large discrepancies were found between the US and MRI velocity profiles over the cross-sections, with differences for US in the same range as for the Womersley profile. Differences in flow pulses revealed that US generally performs best in terms of maximum flow, forward flow and ratios between forward and backward flow, whereas it often overestimates the backward flow. Both spatial patterns and magnitude of the computed hemodynamics were considerably affected by the prescribed velocity boundary conditions. Larger errors and smaller differences between the US and generic CFD cases were observed for patients compared to volunteers. Conclusion: These results show that it is feasible to acquire the patient-specific flow pulse from PWD data, provided that the PWD acquisition could be performed proximal to the aneurysm region, and resulted in a triphasic flow pattern. However, obtaining the patient-specific velocity profile over the cross-section using CD data is not reliable. For the volunteers, utilizing the US flow profile instead of the generic flow profile generally resulted in improved performance, whereas this was the case in more than half of the cases for the patients. [ABSTRACT FROM AUTHOR]

Virtual brain twins are personalized, generative and adaptive brain models based on data from an individual's brain for scientific and clinical use. After a description of the key elements of virtual brain twins, we present the standard model for personalized whole-brain network models. The personalization is accomplished using a subject's brain imaging data by three means: (1) assemble cortical and subcortical areas in the subject-specific brain space; (2) directly map connectivity into the brain models, which can be generalized to other parameters; and (3) estimate relevant parameters through model inversion, typically using probabilistic machine learning. We present the use of personalized whole-brain network models in healthy ageing and five clinical diseases: epilepsy, Alzheimer's disease, multiple sclerosis, Parkinson's disease and psychiatric disorders. Specifically, we introduce spatial masks for relevant parameters and demonstrate their use based on the physiological and pathophysiological hypotheses. Finally, we pinpoint the key challenges and future directions. [ABSTRACT FROM AUTHOR]

Despite the large impact chronic obstructive pulmonary disease (COPD) that has on the population, the implementation of new technologies for diagnosis and treatment remains limited. Current practices in ambulatory oxygen therapy used in COPD rely on fixed doses overlooking the diverse activities which patients engage in. To address this challenge, we propose a software architecture aimed at delivering patient-personalized edge-based artificial intelligence (AI)-assisted models that are built upon data collected from patients' previous experiences along with an evaluation function. The main objectives reside in proactively administering precise oxygen dosages in real time to the patient (the edge), leveraging individual patient data, previous experiences, and actual activity levels, thereby representing a substantial advancement over conventional oxygen dosing. Through a pilot test using vital sign data from a cohort of five patients, the limitations of a one-size-fits-all approach are demonstrated, thus highlighting the need for personalized treatment strategies. This study underscores the importance of adopting advanced technological approaches for ambulatory oxygen therapy. [ABSTRACT FROM AUTHOR]

To improve abdominal aortic aneurysm (AAA) rupture risk assessment, a large, longitudinal study on AAA hemodynamics and biomechanics is necessary, using personalized fluid-structure interaction (FSI) modeling. 3-dimensional, time-resolved ultrasound (3D+t US) is the preferred image modality to obtain the patient-specific AAA geometry for such a study, since it is safe, affordable and provides temporal information. However, the 3D+t US field-of-view (FOV) is limited and therefore often fails to capture the inlet and aorto-iliac bifurcation geometry. In this study, a framework was developed to add parametric inlet and bifurcation geometries to the abdominal aortic aneurysm geometry by employing dataset statistics and parameters of the AAA geometry. The impact of replacing the patient-specific inlet and bifurcation geometries, acquired using computed tomography (CT) scans, by parametric geometries was evaluated by examining the differences in hemodynamics (systolic and time-averaged wall shear stress and oscillatory shear index) in the aneurysm region. The results show that the inlet geometry has a larger effect on the AAA hemodynamics (median differences of 7.5 to 18.8%) than the bifurcation geometry (median differences all below 1%). Therefore, it is not feasible to replace the patient-specific inlet geometry by a generic one. Future studies should investigate the possibilities of extending the proximal FOV of 3D+t US. However, this study did show the feasibility of adding a parametric bifurcation geometry to the aneurysm geometry. After extending the proximal FOV, the obtained framework can be used to extract AAA geometries from 3D+t US for FSI simulations, despite the absence of the bifurcation geometry. [ABSTRACT FROM AUTHOR]

Introduction: Anterior cervical discectomy and fusion (ACDF) has been considered as the gold standard surgical treatment for cervical degenerative pathologies. Some surgeons tend to use larger-sized interbody cages during ACDF to restore the index intervertebral disc height, hence, this study evaluated the effect of larger-sized interbody cages on the cervical spine with ACDF under both static and cyclic loading.Method: Twenty pre-operative personalized poro-hyperelastic finite element (FE) models were developed. ACDF post-operative models were then constructed and four clinical scenarios (i.e., 1) No-distraction; 2) 1 mm distraction; 3) 2 mm distraction; and 4) 3 mm distraction) were predicted for each patient. The biomechanical responses at adjacent spinal levels were studied subject to static and cyclic loading. Non-parametric Friedman statistical comparative tests were performed and the p values less than 0.05 were reflected as significant.Results: The calculated intersegmental range of motion (ROM) and intradiscal pressure (IDP) from 20 pre-operative FE models were within the overall ranges compared to the available data from literature. Under static loading, greater ROM, IDP, facet joint force (FJF) values were detected post ACDF, as compared with pre-op. Over-distraction induced significantly higher IDP and FJF in both upper and lower adjacent levels in extension. Higher annulus fibrosus stress and strain values, and increased disc height and fluid loss at the adjacent levels were observed in ACDF group which significantly increased for over-distraction groups.Discussion: it was concluded that using larger-sized interbody cages (the height of ≥2 mm of the index disc height) can result in remarkable variations in biomechanical responses of adjacent levels, which may indicate as risk factor for adjacent segment disease. The results of this comprehensive FE investigation using personalized modeling technique highlight the importance of selecting the appropriate height of interbody cage in ACDF surgery.

Introduction. The use of mathematical modeling methods in clinical practice will make it possible to identify the pathogenetic forms of dorsopathies and thus reasonably use the concept of targeted treatment in the management of patients of this category.Aim. To evaluate the possibilities of finite element segmental analysis of the cervical spine for personalized treatment and prediction of the course of dorsopathies.Material and methods. Based on the combined data of computed and magnetic resonance imaging of the patient (female, born in 1951), a model of the C5 – C7 segment was generated, including: vertebrae C5, C6, C7, IVD, anterior and posterior longitudinal ligaments, two pairs of facet joints, spinal cord, nuchal ligament. Computer modeling and finite element method were used to analyze the stress-strain state of the cervical spine of a patient with degenerative-dystrophic changes in the C2 – C7 segments. In the Abaqus/CAE 6.14 software, finite element analysis of the C5 – C7 stress-strain state was carried out in the state of flexion, rotation and compression. The data obtained during compression were compared with previous experiments in silico and in vitro for the norm.Results. For each state, stress and displacement diagrams, load-displacement curves, stress profiles in the MPD were obtained. The axial mobility of the segment under compressive load is two times lower compared to the norm under the same boundary conditions and material models. The degree of involvement of the spinal cord in conflicts with the surrounding anatomical structures was studied. When the model was rotated to the right, conflicts were observed between the spinal cord roots and the bone structures of the vertebrae in the foraminal zones, as well as at the level of the C5 – C6 and C6 – C7 discs with the left posterolateral surfaces of the fibrous rings. When the model was turned to the left, conflicts of the spinal cord were observed in all foraminal zones, as well as at the level of the C6 – C7 disc with the left posterolateral surface of the fibrous rings. Based on the data on stresses in the studied segment, further development of dorsopathies and degenerative changes in the cervical spine was predicted.Conclusions. The use of finite element segmental analysis of the cervical spine creates objective prerequisites for the formation of a combined personalized treatment and prediction of the course of dorsopathies.

BackgroundFood craving relates to unhealthy eating behaviors such as overeating or binge eating and is thus a promising target for digital interventions. Yet, craving varies strongly across the day and is more likely in some contexts (external, internal) than in others. Prediction of food cravings ahead of time would enable preventive interventions.ObjectiveThe objective of this study was to investigate whether upcoming food cravings could be detected and predicted from passive smartphone sensor data (excluding geolocation information) without the need for repeated questionnaires.MethodsMomentary food craving ratings, given six times a day for 14 days by 56 participants, served as the dependent variable. Predictor variables were environmental noise, light, device movement, screen activity, notifications, and time of the day recorded from 150 to 30 min prior to these ratings.ResultsIndividual high vs. low craving ratings could be predicted on the test set with a mean area under the curve (AUC) of 0.78. This outperformed a baseline model trained on past craving values in 85% of participants by 14%. Yet, this AUC value is likely the upper bound and needs to be independently validated with longer data sets that allow a split into training, validation, and test sets.ConclusionsCraving states can be forecast from external and internal circumstances as these can be measured through smartphone sensors or usage patterns in most participants. This would allow for just-in-time adaptive interventions based on passive data collection and hence with minimal participant burden.

Introduction: While the short-term post-operative outcome of lumbar fusion is satisfying for most patients, adjacent segment disease (ASD) can be prevalent in long-term clinical observations. It might be valuable to investigate if inherent geometrical differences among patients can significantly alter the biomechanics of adjacent levels post-surgery. This study aimed to utilize a validated geometrically personalized poroelastic finite element (FE) modeling technique to evaluate the alteration of biomechanical response in adjacent segments post-fusion.Methods: Thirty patients were categorized for evaluation in this study into two distinct groups [i.e., 1) non-ASD and 2) ASD patients] based on other long-term clinical follow-up investigations. To evaluate the time-dependent responses of the models subjected to cyclic loading, a daily cyclic loading scenario was applied to the FE models. Different rotational movements in different planes were superimposed using a 10 Nm moment after daily loading to compare the rotational motions with those at the beginning of cyclic loading. The biomechanical responses of the lumbosacral FE spine models in both groups were analyzed and compared before and after daily loading.Results: The achieved comparative errors between the FE results and clinical images were on average below 20% and 25% for pre-op and post-op models, respectively, which confirms the applicability of this predictive algorithm for rough pre-planning estimations. The results showed that the disc height loss and fluid loss were increased for the adjacent discs in post-op models after 16 h of cyclic loading. In addition, significant differences in disc height loss and fluid loss were observed between the patients who were in the non-ASD and ASD groups. Similarly, the increased stress and fiber strain in the annulus fibrosus (AF) was higher in the adjacent level of post-op models. However, the calculated stress and fiber strain values were significantly higher for patients with ASD.Discussion: Evaluating the biomechanical response of pre-op and post-op modeling in the non-ASD and ASD groups showed that the inherent geometric differences among patients cause significant variations in the estimated mechanical response. In conclusion, the results of the current study highlighted the effect of geometrical parameters (which may refer to the anatomical conditions or the induced modifications regarding surgical techniques) on time-dependent responses of lumbar spine biomechanics.

Personalized decision-making can be implemented in a Federated learning (FL) framework that can collaboratively train a decision model by extracting knowledge across intelligent clients, e.g. smartphones or enterprises. FL can mitigate the data privacy risk of collaborative training since it merely collects local gradients from users without access to their data. However, FL is fragile in the presence of statistical heterogeneity that is commonly encountered in personalized decision making, e.g., non-IID data over different clients. Existing FL approaches usually update a single global model to capture the shared knowledge of all users by aggregating their gradients, regardless of the discrepancy between their data distributions. By comparison, a mixture of multiple global models could capture the heterogeneity across various clients if assigning the client to different global models (i.e., centers) in FL. To this end, we propose a novel multi-center aggregation mechanism to cluster clients using their models' parameters. It learns multiple global models from data as the cluster centers, and simultaneously derives the optimal matching between users and centers. We then formulate it as an optimization problem that can be efficiently solved by a stochastic expectation maximization (EM) algorithm. Experiments on multiple benchmark datasets of FL show that our method outperforms several popular baseline methods. The experimental source codes are publicly available on the Github repository (GitHub repository: https://github.com/mingxuts/multi-center-fed-learning). [ABSTRACT FROM AUTHOR]

The high incidence of stroke occurrence necessitates the understanding of its causes and possible ways for early prediction and prevention. In this respect, statistical methods offer the "big picture," but they have a weak predictive ability at an individual level. This research proposes a new personalized modeling method based on computational spiking neural networks (SNN) for the identification of causal associations between clinical and environmental time series data that can be used to predict individual stroke events. The method is tested on 804 stroke patients. Given a clinical data set of patients who experienced a stroke in the past and the corresponding environmental time-series data for a selected time-window before the stroke event, the method identifies the clusters of individuals with a high risk for stroke under similar conditions. The methodology involves a pipeline of processes when creating a personalized model for an individual x : (1) selecting a group of individuals Gx with similar personal records to x ; (2) training a personalized SNN x model of several days of environmental data related to the Gx group to predict the risk of stroke for x at least one day earlier; (3) model interpretability through 3D visualization; (4) discovery of personalized predictive markers. The results are twofold, first proposing a new computational methodology and second presenting new findings. It is found that certain environmental factors, such as SO2, PM10, CO, and PM2.5, increase the risk of stroke if an individual x belongs to a certain cluster of people, characterized by a combination of family history of stroke and diabetes, overweight, vascular/heart disease, age, and other. For the used population data, the proposed method can predict accurately individual risk of stroke before the day of the stroke. The paper presents a new methodology for personalized machine learning methods to define subgroups of the population with a high risk of stroke and to predict early individual risk of the stroke event. This makes the proposed cognitive computation method useful to reduce morbidity and mortality in society. The method is broadly applicable for predicting individual risk of other diseases and mental health conditions. [ABSTRACT FROM AUTHOR]

Purpose: This study proposes a novel approach to obtain personalized estimates of cardiovascular parameters by combining (i) electrocardiography and ballistocardiography for noninvasive cardiovascular monitoring, (ii) a physiology-based mathematical model for predicting personalized cardiovascular variables, and (iii) an evolutionary algorithm (EA) for searching optimal model parameters.Methods: Electrocardiogram (ECG), ballistocardiogram (BCG), and a total of six blood pressure measurements are recorded on three healthy subjects. The R peaks in the ECG are used to segment the BCG signal into single BCG curves for each heart beat. The time distance between R peaks is used as an input for a validated physiology-based mathematical model that predicts distributions of pressures and volumes in the cardiovascular system, along with the associated BCG curve. An EA is designed to search the generation of parameter values of the cardiovascular model that optimizes the match between model-predicted and experimentally-measured BCG curves. The physiological relevance of the optimal EA solution is evaluated a posteriori by comparing the model-predicted blood pressure with a cuff placed on the arm of the subjects to measure the blood pressure.Results: The proposed approach successfully captures amplitudes and timings of the most prominent peak and valley in the BCG curve, also known as the J peak and K valley. The values of cardiovascular parameters pertaining to ventricular function can be estimated by the EA in a consistent manner when the search is performed over five different BCG curves corresponding to five different heart-beats of the same subject. Notably, the blood pressure predicted by the physiology-based model with the personalized parameter values provided by the EA search exhibits a very good agreement with the cuff-based blood pressure measurement.Conclusion: The combination of EA with physiology-based modeling proved capable of providing personalized estimates of cardiovascular parameters and physiological variables of great interest, such as blood pressure. This novel approach opens the possibility for developing quantitative devices for noninvasive cardiovascular monitoring based on BCG sensing.

Purpose: This study proposes a novel approach to obtain personalized estimates of cardiovascular parameters by combining (i) electrocardiography and ballistocardiography for noninvasive cardiovascular monitoring, (ii) a physiology-based mathematical model for predicting personalized cardiovascular variables, and (iii) an evolutionary algorithm (EA) for searching optimal model parameters. Methods: Electrocardiogram (ECG), ballistocardiogram (BCG), and a total of six blood pressure measurements are recorded on three healthy subjects. The R peaks in the ECG are used to segment the BCG signal into single BCG curves for each heart beat. The time distance between R peaks is used as an input for a validated physiology-based mathematical model that predicts distributions of pressures and volumes in the cardiovascular system, along with the associated BCG curve. An EA is designed to search the generation of parameter values of the cardiovascular model that optimizes the match between model-predicted and experimentally-measured BCG curves. The physiological relevance of the optimal EA solution is evaluated a posteriori by comparing the model-predicted blood pressure with a cuff placed on the arm of the subjects to measure the blood pressure. Results: The proposed approach successfully captures amplitudes and timings of the most prominent peak and valley in the BCG curve, also known as the J peak and K valley. The values of cardiovascular parameters pertaining to ventricular function can be estimated by the EA in a consistent manner when the search is performed over five different BCG curves corresponding to five different heart-beats of the same subject. Notably, the blood pressure predicted by the physiology-based model with the personalized parameter values provided by the EA search exhibits a very good agreement with the cuff-based blood pressure measurement. Conclusion: The combination of EA with physiology-based modeling proved capable of providing personalized estimates of cardiovascular parameters and physiological variables of great interest, such as blood pressure. This novel approach opens the possibility for developing quantitative devices for noninvasive cardiovascular monitoring based on BCG sensing. [ABSTRACT FROM AUTHOR]

Precise model personalization is a key step towards the application of cardiovascular physical models. In this manuscript, we propose to use deep learning (DL) to solve the parameter estimation problem in cardiovascular hemodynamics. Based on the convolutional neural network (CNN) and fully connected neural network (FCNN), a multi‐input deep neural network (DNN) model is developed to map the nonlinear relationship between measurements and the parameters to be estimated. In this model, two separate network structures are designed to extract the features of two types of measurement data, including pressure waveforms and a vector composed of heart rate (HR) and pulse transit time (PTT), and a shared structure is used to extract their combined dependencies on the parameters. Besides, we try to use the transfer learning (TL) technology to further strengthen the personalized characteristics of a trained‐well network. For assessing the proposed method, we conducted the parameter estimation using synthetic data and in vitro data respectively, and in the test with synthetic data, we evaluated the performance of the TL algorithm through two individuals with different characteristics. A series of estimation results show that the estimated parameters are in good agreement with the true values. Furthermore, it is also found that the estimation accuracy can be significantly improved by a multicycle combination strategy. Therefore, we think that the proposed method has the potential to be used for parameter estimation in cardiovascular hemodynamics, which can provide an immediate, accurate, and sustainable personalization process, and deserves more attention in the future. [ABSTRACT FROM AUTHOR]

• A 0D-1D closed-loop model is developed for the hemodynamics simulation. • The optimization method for an individual modelling has wide applicability. • 22 samples at both rest and EECP states are collected to verify the model. • The model effectively simulates the personalized hemodynamics of rest and EECP. • This method can help choose the best strategy on the patient-specific treatment. Enhanced external counterpulsation (EECP) is a non-invasive treatment modality capable of treating a variety of ischemic diseases. Currently, no effective methods of predicting the patient-specific hemodynamic effects of EECP are available. In this study, a personalized 0D-1D model of the cardiovascular system was developed for hemodynamic simulation to simulate the changes in blood flow in the EECP state and develop the best treatment protocol for each individual. A 0D-1D closed-loop model of the cardiovascular system was developed for hemodynamic simulation, consisting of a 1D wave propagation model for arteries, a 0D model for veins and capillaries, and a one-fiber model for the heart. Additionally, a simulation model coupling EECP with a 1D model was established. Physiological data, including the blood flow in different arteries, were clinically collected from 22 volunteers at rest and in the EECP state. Sensitivity analysis and a simulated annealing algorithm were used to build personalized 0D-1D models using the clinical data in the rest state as optimization objectives. Then, the clinical data on EECP were used to verify the applicability and accuracy of the personalized models. The simulation results and clinical data were found to be in agreement for all 22 subjects, with waveform similarity coefficients (r) exceeding 90% for most arteries at rest and 80% for most arteries during EECP. The 0D-1D closed-loop model and the optimized method can facilitate personalized modeling of the cardiovascular system using the data in the rest state and effectively predict the hemodynamic changes in the EECP state, which is significant for the numerical simulation of personalized hemodynamics. The model can also potentially be used to make decisions regarding patient-specific treatment. [ABSTRACT FROM AUTHOR]