The predictive ability of a QCT-FE model of the proximal femoral stiffness under multiple load cases is strongly influenced by experimental uncertainties.

Despite significant improvements in terms of the predictive ability of Quantitative Computed Tomography based Finite Element (QCT-FE) models in estimating femoral strength (fracture load and stiffness), no substantial clinical adoption of this method has taken place to date. Narrowing the wide variability of FE results by standardizing the methodology and validation protocols, as well as reducing the uncertainties in the FEA process have been proposed as routes towards improved reliability. The aim of this study was to: First, validate a QCT-FE model of proximal femoral stiffness in multiple stance load cases, and second, using a parametric approach, determine the influence of select experimental and modeling parameters on the predictive ability of our model. Ten fresh frozen human femoral samples were tested in neutral stance, 15° adducted and 15° abducted load cases. Voxel-based linear-elastic QCT-FE models of the samples were generated to predict the models' stiffness values in all load cases. The base FE models were validated against the experimental results using linear regression. Thirty six deviated models were created using the minimum and maximum values of experiment-based "plausible range" for 18 parameters in 4 categories of embedding, loading, material, and segmentation. The predictive ability of the models were compared in terms of the coefficient of determination (R2) of the linear regression between the measured and predicted stiffness values in all load cases. Our model was capable of capturing 90% of the variation in the experimental stiffness of the samples in neutral stance position (R2 = 0.9, concordance correlation coefficient (CCC) = 0.93, percent root mean squared error (RMSE%) = 8.4%, slope and intercept not significantly different from unity and zero, respectively). Embedding and loading categories strongly affected the predictive ability of the models with an average percent difference in R2 of 4.36% ± 2.77 and 2.96% ± 1.69 for the stance-neutral load case, respectively. The performance of the models were significantly different in adducted and abducted load cases with their R2 dropping to 71% and 70%, respectively. Similarly, off-axes load cases were affected by the parameters differently compared to the neutral load case, with the loading parameter category imposing more than 10% difference on their R2, larger than all other categories. We also showed that automatically selecting the best performing plausible value for each parameter and each sample would result in a perfectly linear correlation (R2> 0.99) between the "tuned" model's predicted stiffness and experimental results. Based on our results, high sensitivity of the model performance to experimental parameters requires extra diligence in modeling the embedding geometry and the loading angles since these sources of uncertainty could dwarf the effects of material modeling and image processing parameters. The results of this study could help in improving the robustness of the QCT-FE models of proximal femur by limiting the uncertainties in the experimental and modeling steps.

GC-CNNnet: Diagnosis of Alzheimer's Disease with PET Images Using Genetic and Convolutional Neural Network.

There is a wide variety of effects of Alzheimer's disease (AD), a neurodegenerative disease that can lead to cognitive decline, deterioration of daily life, and behavioral and psychological changes. A polymorphism of the ApoE gene ε 4 is considered a genetic risk factor for Alzheimer's disease. The purpose of this paper is to demonstrate that single-nucleotide polymorphic markers (SNPs) have a causal relationship with quantitative PET imaging traits. Additionally, the classification of AD is based on the frequency of brain tissue variations in PET images using a combination of k-nearest-neighbor (KNN), support vector machine (SVM), linear discrimination analysis (LDA), and convolutional neural network (CNN) techniques. According to the results, the suggested SNPs appear to be associated with quantitative traits more strongly than the SNPs in the ApoE genes. Regarding the classification result, the highest accuracy is obtained by the CNN with 91.1%. These results indicate that the KNN and CNN methods are beneficial in diagnosing AD. Nevertheless, the LDA and SVM are demonstrated with a lower level of accuracy.

A Review of Methods of Diagnosis and Complexity Analysis of Alzheimer's Disease Using EEG Signals.

This study will concentrate on recent research on EEG signals for Alzheimer's diagnosis, identifying and comparing key steps of EEG-based Alzheimer's disease (AD) detection, such as EEG signal acquisition, preprocessing function extraction, and classification methods. Furthermore, highlighting general approaches, variations, and agreement in the use of EEG identified shortcomings and guidelines for multiple experimental stages ranging from demographic characteristics to outcomes monitoring for future research. Two main targets have been defined based on the article's purpose: (1) discriminative (or detection), i.e., look for differences in EEG-based features across groups, such as MCI, moderate Alzheimer's disease, extreme Alzheimer's disease, other forms of dementia, and stable normal elderly controls; and (2) progression determination, i.e., look for correlations between EEG-based features and clinical markers linked to MCI-to-AD conversion and Alzheimer's disease intensity progression. Limitations mentioned in the reviewed papers were also gathered and explored in this study, with the goal of gaining a better understanding of the problems that need to be addressed in order to advance the use of EEG in Alzheimer's disease science.

Which experimental procedures influence the apparent proximal femoral stiffness? A parametric study.

BACKGROUND: Experimental validation is the gold standard for the development of FE predictive models of bone. Employing multiple loading directions could improve this process. To capture the correct directional response of a sample, the effect of all influential parameters should be systematically considered. This study aims to determine the impact of common experimental parameters on the proximal femur's apparent stiffness. METHODS: To that end, a parametric approach was taken to study the effects of: repetition, pre-loading, re-adjustment, re-fixation, storage, and µCT scanning as random sources of uncertainties, and loading direction as the controlled source of variation in both stand and side-fall configurations. Ten fresh-frozen proximal femoral specimens were prepared and tested with a novel setup in three consecutive sets of experiments. The neutral state and 15-degree abduction and adduction angles in both stance and fall configurations were tested for all samples and parameters. The apparent stiffness of the samples was measured using load-displacement data from the testing machine and validated against marker displacement data tracked by DIC cameras. RESULTS: Among the sources of uncertainties, only the storage cycle affected the proximal femoral apparent stiffness significantly. The random effects of setup manipulation and intermittent µCT scanning were negligible. The 15∘ deviation in loading direction had a significant effect comparable in size to that of switching the loading configuration from neutral stance to neutral side-fall. CONCLUSION: According to these results, comparisons between the stiffness of the samples under various loading scenarios can be made if there are no storage intervals between the different load cases on the same samples. These outcomes could be used as guidance in defining a highly repeatable and multi-directional experimental validation study protocol.

Single and Combined Neuroimaging Techniques for Alzheimer's Disease Detection.

Alzheimer's disease (AD) consists of the gradual process of decreasing volume and quality of neuron connection in the brain, which consists of gradual synaptic integrity and loss of cognitive functions. In recent years, there has been significant attention in AD classification and early detection with machine learning algorithms. There are different neuroimaging techniques for capturing data and using it for the classification task. Input data as images will help machine learning models to detect different biomarkers for AD classification. This marker has a more critical role for AD detection than other diseases because beta-amyloid can extract complex structures with some metal ions. Most researchers have focused on using 3D and 4D convolutional neural networks for AD classification due to reasonable amounts of data. Also, combination neuroimaging techniques like functional magnetic resonance imaging and positron emission tomography for AD detection have recently gathered much attention. However, gathering a combination of data can be expensive, complex, and tedious. For time consumption reasons, most patients prefer to throw one of the neuroimaging techniques. So, in this review article, we have surveyed different research studies with various neuroimaging techniques and ML methods to see the effect of using combined data as input. The result has shown that the use of the combination method would increase the accuracy of AD detection. Also, according to the sensitivity metrics from different machine learning methods, MRI and fMRI showed promising results.

Diagnosis of Alzheimer's Disease Severity with fMRI Images Using Robust Multitask Feature Extraction Method and Convolutional Neural Network (CNN).

The automatic diagnosis of Alzheimer's disease plays an important role in human health, especially in its early stage. Because it is a neurodegenerative condition, Alzheimer's disease seems to have a long incubation period. Therefore, it is essential to analyze Alzheimer's symptoms at different stages. In this paper, the classification is done with several methods of machine learning consisting of K-nearest neighbor (KNN), support vector machine (SVM), decision tree (DT), linear discrimination analysis (LDA), and random forest (RF). Moreover, novel convolutional neural network (CNN) architecture is presented to diagnose Alzheimer's severity. The relationship between Alzheimer's patients' functional magnetic resonance imaging (fMRI) images and their scores on the MMSE is investigated to achieve the aim. The feature extraction is performed based on the robust multitask feature learning algorithm. The severity is also calculated based on the Mini-Mental State Examination score, including low, mild, moderate, and severe categories. Results show that the accuracy of the KNN, SVM, DT, LDA, RF, and presented CNN method is 77.5%, 85.8%, 91.7%, 79.5%, 85.1%, and 96.7%, respectively. Moreover, for the presented CNN architecture, the sensitivity of low, mild, moderate, and severe status of Alzheimer patients is 98.1%, 95.2%,89.0%, and 87.5%, respectively. Based on the findings, the presented CNN architecture classifier outperforms other methods and can diagnose the severity and stages of Alzheimer's disease with maximum accuracy.

Diagnosis of Alzheimer's Disease by Time-Dependent Power Spectrum Descriptors and Convolutional Neural Network Using EEG Signal.

Using strategies that obtain biomarkers where early symptoms coincide, the early detection of Alzheimer's disease and its complications is essential. Electroencephalogram is a technology that allows thousands of neurons with equal spatial orientation of the duration of cerebral cortex electrical activity to be registered by postsynaptic potential. Therefore, in this paper, the time-dependent power spectrum descriptors are used to diagnose the electroencephalogram signal function from three groups: mild cognitive impairment, Alzheimer's disease, and healthy control test samples. The final feature used in three modes of traditional classification methods is recorded: k-nearest neighbors, support vector machine, linear discriminant analysis approaches, and documented results. Finally, for Alzheimer's disease patient classification, the convolutional neural network architecture is presented. The results are indicated using output assessment. For the convolutional neural network approach, the accurate meaning of accuracy is 82.3%. 85% of mild cognitive impairment cases are accurately detected in-depth, but 89.1% of the Alzheimer's disease and 75% of the healthy population are correctly diagnosed. The presented convolutional neural network outperforms other approaches because performance and the k-nearest neighbors' approach is the next target. The linear discriminant analysis and support vector machine were at the low area under the curve values.

Effect of CT imaging on the accuracy of the finite element modelling in bone.

The finite element (FE) analysis is a highly promising tool to simulate the behaviour of bone. Skeletal FE models in clinical routine rely on the information about the geometry and bone mineral density distribution from quantitative computed tomography (CT) imaging systems. Several parameters in CT imaging have been reported to affect the accuracy of FE models. FE models of bone are exclusively developed in vitro under scanning conditions deviating from the clinical setting, resulting in variability of FE results (< 10%). Slice thickness and field of view had little effect on FE predicted bone behaviour (≤ 4%), while the reconstruction kernels showed to have a larger effect (≤ 20%). Due to large interscanner variations (≤ 20%), the translation from an experimental model into clinical reality is a critical step. Those variations are assumed to be mostly caused by different "black box" reconstruction kernels and the varying frequency of higher density voxels, representing cortical bone. Considering the low number of studies together with the significant effect of CT imaging on the finite element model outcome leading to high variability in the predicted behaviour, we propose further systematic research and validation studies, ideally preceding multicentre and longitudinal studies.

Influence of processing parameters on mechanical properties of a 3D-printed trabecular bone microstructure.

Natural bone microstructure has shown to be the most efficient choice for the bone scaffold design. However, there are several process parameters involved in the generation of a microCT-based 3D-printed (3DP) bone. In this study, the effect of selected parameters on the reproducibility of mechanical properties of a 3DP trabecular bone structure is investigated. MicroCT images of a distal radial sample were used to reconstruct a 3D ROI of trabecular bone. Nine tensile tests on bulk material and 54 compression tests on 8.2 mm cubic samples were performed (9 cases × 6 specimens/case). The effect of input-image resolution, STL mesh decimation, boundary condition, support material, and repetition parameters on the weight, elastic modulus, and strength were studied. The elastic modulus and the strength of bulk material showed consistent results (CV% = 9 and 6%, respectively). The weight, elastic modulus, and strength of the cubic samples showed small intragroup variation (average CV% = 1.2, 9, and 5.5%, respectively). All studied parameters had a significant effect on the outcome variables with less effect on the weight. Utmost care to every step of the 3DP process and involved parameters is required to be able to reach the desired mechanical properties in the final printed specimen. © 2019 The Authors. Journal of Biomedical Materials Research Part B: Applied Biomaterials published by Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater 108B:38-47, 2020.

QCT-based finite element prediction of pathologic fractures in proximal femora with metastatic lesions.

Predicting pathologic fractures in femora with metastatic lesions remains a clinical challenge. Currently used guidelines are inaccurate, especially to predict non-impeding fractures. This study evaluated the ability of a nonlinear quantitative computed tomography (QCT)-based homogenized voxel finite element (hvFE) model to predict patient-specific pathologic fractures. The hvFE model was generated highly automated from QCT images of human femora. The femora were previously loaded in a one-legged stance setup in order to assess stiffness, failure load, and fracture location. One femur of each pair was tested in its intact state, while the contralateral femur included a simulated lesion on either the superolateral- or the inferomedial femoral neck. The hvFE model predictions of the stiffness (0.47 < R2 < 0.94), failure load (0.77 < R2 < 0.98), and exact fracture location (68%) were in good agreement with the experimental data. However, the model underestimated the failure load by a factor of two. The hvFE models predicted significant differences in stiffness and failure load for femora with superolateral- and inferomedial lesions. In contrast, standard clinical guidelines predicted identical fracture risk for both lesion sites. This study showed that the subject-specific QCT-based hvFE model could predict the effect of metastatic lesions on the biomechanical behaviour of the proximal femur with moderate computational time and high level of automation and could support treatment strategy in patients with metastatic bone disease.

Optimizing finite element predictions of local subchondral bone structural stiffness using neural network-derived density-modulus relationships for proximal tibial subchondral cortical and trabecular bone.

BACKGROUND: Quantitative computed tomography based subject-specific finite element modeling has potential to clarify the role of subchondral bone alterations in knee osteoarthritis initiation, progression, and pain. However, it is unclear what density-modulus equation(s) should be applied with subchondral cortical and subchondral trabecular bone when constructing finite element models of the tibia. Using a novel approach applying neural networks, optimization, and back-calculation against in situ experimental testing results, the objective of this study was to identify subchondral-specific equations that optimized finite element predictions of local structural stiffness at the proximal tibial subchondral surface. METHODS: Thirteen proximal tibial compartments were imaged via quantitative computed tomography. Imaged bone mineral density was converted to elastic moduli using multiple density-modulus equations (93 total variations) then mapped to corresponding finite element models. For each variation, root mean squared error was calculated between finite element prediction and in situ measured stiffness at 47 indentation sites. Resulting errors were used to train an artificial neural network, which provided an unlimited number of model variations, with corresponding error, for predicting stiffness at the subchondral bone surface. Nelder-Mead optimization was used to identify optimum density-modulus equations for predicting stiffness. FINDINGS: Finite element modeling predicted 81% of experimental stiffness variance (with 10.5% error) using optimized equations for subchondral cortical and trabecular bone differentiated with a 0.5g/cm3 density. INTERPRETATION: In comparison with published density-modulus relationships, optimized equations offered improved predictions of local subchondral structural stiffness. Further research is needed with anisotropy inclusion, a smaller voxel size and de-blurring algorithms to improve predictions.

Individual and combined effects of OA-related subchondral bone alterations on proximal tibial surface stiffness: a parametric finite element modeling study.

The role of subchondral bone in OA pathogenesis is unclear. While some OA-related changes to morphology and material properties in different bone regions have been described, the effect of these alterations on subchondral bone surface stiffness has not been investigated. The objectives of this study were to characterize the individual (Objective 1) and combined (Objective 2) effects of OA-related morphological and mechanical alterations to subchondral and epiphyseal bone on surface stiffness of the proximal tibia. We developed and validated a parametric FE model of the proximal tibia using quantitative CT images of 10 fresh-frozen cadaveric specimens and in situ macro-indentation testing. Using this validated FE model, we estimated the individual and combined roles of OA-related alterations in subchondral cortical thickness and elastic modulus, and subchondral trabecular and epiphyseal trabecular elastic moduli on local surface stiffness. A 20% increase in subchondral cortical or subchondral trabecular elastic moduli resulted in little change in stiffness (1% increase). A 20% reduction in epiphyseal trabecular elastic modulus, however, resulted in an 11% reduction in stiffness. Our parametric analysis suggests that subchondral bone stiffness is affected primarily by epiphyseal trabecular bone elastic modulus rather than subchondral cortical and trabecular morphology or mechanical properties. Our results suggest that observed OA-related alterations to epiphyseal trabecular bone (e.g., lower mineralization, bone volume fraction, density and elastic modulus) may contribute to OA proximal tibiae being less stiff than normal.

Prediction of local proximal tibial subchondral bone structural stiffness using subject-specific finite element modeling: Effect of selected density-modulus relationship.

BACKGROUND: Quantitative computed tomography based subject-specific finite element modeling has potential to clarify the role of subchondral bone alterations in knee osteoarthritis initiation, progression, and pain initiation. Calculation of bone elastic moduli from image data is a basic step when constructing finite element models. However, different relationships between elastic moduli and imaged density (known as density-modulus relationships) have been reported in the literature. The objective of this study was to apply seven different trabecular-specific and two cortical-specific density-modulus relationships from the literature to finite element models of proximal tibia subchondral bone, and identify the relationship(s) that best predicted experimentally measured local subchondral structural stiffness with highest explained variance and least error. METHODS: Thirteen proximal tibial compartments were imaged via quantitative computed tomography. Imaged bone mineral density was converted to elastic moduli using published density-modulus relationships and mapped to corresponding finite element models. Proximal tibial structural stiffness values were compared to experimentally measured stiffness values from in-situ macro-indentation testing directly on the subchondral bone surface (47 indentation points). FINDINGS: Regression lines between experimentally measured and finite element calculated stiffness had R(2) values ranging from 0.56 to 0.77. Normalized root mean squared error varied from 16.6% to 337.6%. INTERPRETATION: Of the 21 evaluated density-modulus relationships in this study, Goulet combined with Snyder and Schneider or Rho appeared most appropriate for finite element modeling of local subchondral bone structural stiffness. Though, further studies are needed to optimize density-modulus relationships and improve finite element estimates of local subchondral bone structural stiffness.