BMC biomedical engineering最新文献

Glaucoma is a leading cause of irreversible blindness, necessitating early and accurate diagnosis to prevent vision loss. Traditional diagnostic methods often suffer from subjectivity and variability, emphasizing the need for more reliable approaches. This study evaluates the application of machine learning (ML) techniques in glaucoma diagnosis, analyzing their effectiveness and identifying the most promising methods and datasets. A systematic review of five major databases was conducted, selecting 35 studies based on predefined criteria. The findings reveal that structured data, including optical coherence tomography (OCT), visual field (VF) tests, and demographic factors, significantly enhance diagnostic accuracy. ML models such as support vector machine (SVM), deep learning (DL), random forest, and ensemble methods demonstrated accuracy ranging from 76 to 98.3%, with AUC values between 52.5% and 99%. Despite these advancements, challenges such as data imbalance and limited sample sizes impact model generalizability. The results highlight the potential of ML to improve glaucoma detection, though further research is needed to enhance data quality and model validation for broader clinical applicability.

Electrocardiography (ECG) is a non-invasive tool used to identify abnormalities in heart rhythm. It is used to evaluate dysfunctions in the electrical system of the heart. It offers a mechanism that does not cause any harm to patients. Being affordable makes it accessible. It provides a comprehensive assessment of the condition of the heart. Although it provides a successful analysis opportunity for arrhythmia detection, it is time-consuming and depends on the clinician's experience. In addition, since the ECG patterns in pediatric patients are different from the ECG patterns in adults, physicians consider it a difficult and complex task. For this reason, a custom dataset of pediatric patients was created in this study. This dataset consists of 1318 abnormal beats and 1403 normal beats. MobileNetv2 transfer learning architecture was used to classify this balanced dataset. However, the stability of the results is a valuable. Therefore, the optimization algorithm that minimizes the loss function and the regularization method that controls the complexity of the model are proposed. In this direction, Proposed Optimization Algorithm V5 and Proposed Regularization Method V5 approaches have been integrated into the MobileNetv2 transfer learning model. The accuracy rates produced in the training and test datasets are 0.9801 and 0.9509, respectively. These results have acceptable improvement and stability compared to the accuracies of 0.9633 and 0.9399 produced by the original MobileNetv2 architecture on the training and test dataset, respectively. However, performance values provide limited information about the generalizability of the model. Therefore, the same processes were repeated on a more complex dataset with 6 categories. As a result of the classification, the accuracy rates for the training and test data sets were obtained as 0.9200% and 0.8975%, respectively. Training was performed under the same conditions as the training performed on 2-category datasets. Therefore, it is normal for the test dataset to experience a decrease of approximately 5%. The results obtained show that generalizations can be made for comprehensive, highly diverse and rich datasets.

Background: Gait kinetics explains dynamics of gait deviations, which inform surgical and non-surgical clinical-decision-making to enhance walking performance of children with cerebral palsy. Kinetic gait profile of children with lesser crouch angle is known; however lower-extremity gait kinetics of ambulatory children at a further continuum of the spectrum with greater crouch angle is unclear. Therefore, present cross-sectional study evaluated influence of varying crouch angle on gait kinetics and walk distance.

Method: Following ethical approval and signed informed consent of parents, 3-D gait of 33 ambulatory children with CP(10.4 year) and 31 age-matched typically-developing children was studied to compute the magnitude and timing of lower-extremity external net joint moments and power during stance phase. An average of 3gait trials walked bare-feet at self-selected pace was considered for analyses. Walk distance was measured with 2-min walk test. Typically developing children were classified as Group I, children with mild crouch-angle (mean knee flexion angle during stance)[Formula: see text]16.8⁰and ≤ 25⁰ were classified as Group II(n = 17), whereas children with severe crouch-angle i.e.[Formula: see text] 25⁰ throughout stance phase were classified as Group III(n = 16). Three groups were compared with one-way-ANOVA(p ≤ 0.05). Bonferroni adjustment was made for post-hoc analyses (p ≤ 0.01).

Results: Gait speed, cadence and 2-minute walk distance decreased from Group I to II to III(p ≤ 0.01). Hip flexion, extension and adduction; knee flexion and ankle dorsiflexion moments were significantly different between three groups(p ≤ 0.01)). Rise in crouch-angle was associated with an increase in peak hip flexion moment and increase in power generated at hip and decrease in power generated at knee and ankle (p ≤ 0.01). The timing of peak hip and knee moments during stance phase also differed across the 3 groups (p ≤ 0.01) indicating a delay in the occurrence of peak hip flexion-extension; abduction-adduction and knee flexion moment with a rise in crouch angle.

Conclusion: Present findings inform lower-extremity joint kinetics during gait across the spectrum of mild to severe crouch angle with reference to typically-developing children. Precise knowledge of magnitude and pattern of net joint moments and power along with the timing of moments and decline in walking distance in children with severe crouch, can guide therapeutic interventions to restore the optimum dynamic lever arm function for improved walking performance.

Trial registration: CTRI registration no. CTRI/22/12/048524/27/12/2022.

Trial registry: CTRI/22/12.

Trial registration number: 048524. Trial registration date: 27th December 2022.

Transmission of electrical impulses along axons is commonly modelled with the cable equation, which neglects the inductive effects that have been measured in nerves. By using the telegrapher's equations, it is possible to incorporate inductive effects and compare with the non-inductive case. Although both of these approaches have been extensively studied, the question remains as to which of these provides a more accurate model of human physiology. Many of the electrical properties of nerves are frequency-dependent, a fact which is not very relevant in a low-frequency domain, but which becomes salient when higher frequencies are considered, and necessitates the exploration of the magnitude of their effects. We compare the effects of both inductance and other variable parameters across a wide frequency range using both the cable equation and the telegrapher's equations, demonstrating that it is possible for axons to transmit high-frequency signals much more effectively than might be expected, especially in the absence of an action potential. This implies that the high-frequency domain necessitates use of the more complete model.

Background: The aim of this study was to produce ultrasmall superparamagnetic iron oxide (USPIO) nanoparticles (NPs) conjugated to the FROP-1 peptide for targeted magnetic resonance imaging (MRI) of breast cancer cell lines and to evaluate its application as a specific and targeted T1-weighted MR imaging contrast agent in vitro. Sodium citrate-stabilized Fe₃O₄ NPs were conjugated with the FROP-1 peptide by 1-ethyl-3-(3-dimethylaminopropyl) carbide diamide hydrochloride (EDC) to form a novel Fe₃O₄@FROP-1 specific target contrast agent. The specificity and targeting of Fe₃O₄@FROP-1 to bind FROP-1 receptors were investigated in vitro by cellular uptake and cellular MR imaging.

Results: In this study, the synthesis of water-soluble ultrasmall Fe₃O₄ NPs was performed by the co-precipitation method. XRD, TEM, and VSM analyses showed the formation of the Fe₃O₄ NPs with an average size of about 3.78 ± 0.2 nm. FT-IR spectroscopy approved the conjugation of the FROP-1 peptide with the Fe₃O₄ NPs. The synthesized Fe₃O₄@FROP-1 NPs showed good biocompatibility, and the high r1 relaxivity and r2/r1, respectively, were 2.608 mM^- 1S^- 1 and 1.18. The biocompatibility of the Fe₃O₄ and Fe₃O₄@FROP-1 NPs on the MCF-7, SKBR-3, MDA-MB-231, and MCF-10 cell lines was determined using cytotoxicity analysis. The specific targeting effect on the cells was verified by in vitro cellular uptake and cell MR imaging.

Conclusion: It was found that the contrast intensity of the Fe₃O₄@FROP-1 nanoprobe increases as Fe concentration increases. Cellular uptake of the Fe₃O₄ and Fe₃O₄@FROP-1 NPs was quantified using ICP-MS. The synthesized NPs had better imaging performance than Dotarem (gadoterate meglumine). The findings showed that Fe₃O₄@FROP-1 NPs have potential utility as a specific and targeted T1-weighted contrast agent in breast cancer MR imaging.

COVID-19 has claimed the lives of thousands over the past years. Although pathogenic laboratory testing is the established standard, it carries a significant drawback with a notable rate of false negatives. Consequently, there is an urgent need for alternative diagnostic approaches to combat this threat. In response to this pressing need for accurate and parameter-free methods for COVID-19 identification, particularly within lung images, we introduce a novel approach that combines the principles of topological data analysis with the capabilities of machine learning. Our proposed methodology entails the extraction of persistent homology features from lung images, effectively capturing the intrinsic topological properties inherent in the data. These extracted persistent homology features then serve as inputs for various machine learning methods employed for classification purposes. Our primary objective is to achieve exceptional accuracy in the detection of COVID-19 all while showcasing the effectiveness of these topological features. The experimental results demonstrate that the Random Forest Classifier and the Support Vector Machine models outperform the rest, showcasing their effectiveness in classifying CT scan lung images with remarkable precision-an accuracy rate of 97.5% for the Random Forest model and an AUC score that surpasses 0.99 for the SVM. Results of the model on the same data after exclusion of the topological features and on other data with application of the same model with topological features showed the efficiency of these features in the classification task.

Background: Exposure to electric fields affects cell membranes impacting their potential and altering cellular excitability, nerve transmission, or muscle contraction. Furthermore, electric stimulation influences cell communication, migration, proliferation, and differentiation, with potential therapeutic applications. In vitro platforms for electrical stimulation are valuable tools for studying these effects and advancing medical research. In this study, we developed and tested a novel multi-channel in vitro electrical stimulator designed for cellular applications. The device aims to facilitate research on the effects of electrical stimulation (ES) on cellular processes, providing a versatile platform that is easy to reproduce and implement in various laboratory settings.

Methods: The stimulator was designed to be simple, cost-effective, and versatile, fitting on standard 12-well plates for parallel experimentation. Extensive testing was conducted to evaluate the performance of the stimulator, including 3D finite element modelling to analyse electric field distribution. Moreover, the stimulator was evaluated in vitro using neuronal and stem cell cultures.

Results: Finite element modelling confirmed that the electric field was sufficiently homogeneous within the stimulation zone, though liquid volume affected field strength. A custom controller was developed to program stimulation protocols, ensuring precise and adjustable current delivery up to 160 V/m. ES promoted neurite outgrowth when applied to SH-SY5Y neural cells or to primary spinal cord-derived cells. In human neuronal progenitor cells (hNPCs), ES enhanced neurite growth as well as differentiation into neurons. In adipose stem cells (ASCs), ES altered the secretome, enriching it in molecules that promoted hNPC differentiation into neurons without enhancing neurite growth.

Conclusions: Our results highlight the potential of this multi-channel electrical stimulator as a valuable tool for advancing the understanding of ES mechanisms and its therapeutic applications. The simplicity and adaptability of this novel platform make it a promising addition to the toolkit of researchers studying electrical stimulation in cellular models.

Aim: The aim of this study is to apply a novel hybrid framework incorporating a Vision Transformer (ViT) and bidirectional long short-term memory (Bi-LSTM) model for classifying physical activity intensity (PAI) in adults using gravity-based acceleration. Additionally, it further investigates how PAI and temporal window (TW) impacts the model' s accuracy.

Method: This research used the Capture-24 dataset, consisting of raw accelerometer data from 151 participants aged 18 to 91. Gravity-based acceleration was utilised to generate images encoding various PAIs. These images were subsequently analysed using the ViT-BiLSTM model, with results presented in confusion matrices and compared with baseline models. The model's robustness was evaluated through temporal stability testing and examination of accuracy and loss curves.

Result: The ViT-BiLSTM model excelled in PAI classification task, achieving an overall accuracy of 98.5% ± 1.48% across five TWs-98.7% for 1s, 98.1% for 5s, 98.2% for 10s, 99% for 15s, and 98.65% for 30s of TW. The model consistently exhibited superior accuracy in predicting sedentary (98.9% ± 1%) compared to light physical activity (98.2% ± 2%) and moderate-to-vigorous physical activity (98.2% ± 3%). ANOVA showed no significant accuracy variation across PAIs (F = 2.18, p = 0.13) and TW (F = 0.52, p = 0.72). Accuracy and loss curves show the model consistently improves its performance across epochs, demonstrating its excellent robustness.

Conclusion: This study demonstrates the ViT-BiLSTM model's efficacy in classifying PAI using gravity-based acceleration, with performance remaining consistent across diverse TWs and intensities. However, PAI and TW could result in slight variations in the model's performance. Future research should concern and investigate the impact of gravity-based acceleration on PAI thresholds, which may influence model's robustness and reliability.

Background: The ST response to high frequency EM heating may give an indication of rate of BF in underlying tissue. This novel method, which we have termed REFLO (Rapid Electromagnetic Flow) has potential for applications such as detection of PAD. The method utilizes the relationship between blood flow rate and tissue temperature increase during exposure to radio frequency (RF) energy. We are developing an REFLO device to screen for peripheral artery disease (PAD). PAD is characterized by impaired blood flow to the legs, as reflected in the skin microcirculation. The REFLO system incorporates a radio frequency transmitter and a compact transducer housing a micropatch antenna and an infrared (IR) temperature sensor. At high RF frequencies (> 6 GHz) tissue heating is confined to the skin, such that an indication of blood flow may be inferred from the temperature response to controlled heating. The objective of this study is to determine the extent to which the magnitude and depth of heating as well as device sensitivity are functions of (i) RF frequency and (ii) thickness of the dermal tissue layer.

Results: Results show that it is feasible to measure blood flow rate with REFLO technology. Surface temperature increases were found to be more dependent upon the magnitude of power absorption than location of absorption within the skin. While surface temperature response does depend upon radio wave frequency and thickness of the dermis layer, such dependencies are mild. Sensitivity to blood flow rate was found to be proportional to the magnitude of absorbed power.

Conclusion: Results show that it is feasible to discriminate between blood flow rates using REFLO technology at frequencies within the 10-94 GHz range. All frequencies analyzed produced similar levels of sensitivity to blood flow rate despite significant differences in penetration depth. These results are being used in the development of a preclinical prototype for quick and easy detection of asymptomatic PAD in humans.

Despite significant technological progress in prosthetic hands, a device with functionality akin to a biological extremity is far from realization. To better support the development of next-generation technologies, we investigated the grasping capabilities of clinically prescribable and commercially available (CPCA) prosthetic hands against those that are 3D-printed, which offer cost-effective and customizable solutions. Our investigation utilized the Anthropomorphic Hand Assessment Protocol (AHAP) as a benchtop evaluation of the multi-grasp performance of 3D-printed devices against CPCA prosthetic hands. Our comparison sample included three open-source 3D-printed prosthetic hands (HACKberry Hand, HANDi Hand, and BEAR PAW) and three CPCA prosthetic hands (Össur i-Limb Quantum, RSL Steeper BeBionic Hand V3, and Psyonic Ability Hand), along with including previously published AHAP data for four additional 3D-printed hands (Dextrus v2.0, IMMA, InMoov, and Limbitless). Our findings revealed a notable grasping performance disparity, with 3D-printed prostheses generally underperforming compared to their CPCA counterparts, specifically in cylindrical, diagonal volar, extension, and spherical grips. We propose that the observed performance shortfalls are likely attributed to the design or build quality of the 3D-printed prostheses, owing to the fact that 3D-printed hands often have a lower technology readiness level for widespread use. Addressing the limitations highlighted in this work and subsequent research will play a crucial role in refining the design and functionality of both 3D-printed and CPCA prosthetic devices.