Journal of Mechanics in Medicine and Biology

This study compares the prosthetic gait asymmetry in transfemoral amputees with varied femoral lengths to able-bodied participants in terms of temporal-spatial parameters and joint kinematics. Sixteen young and active transfemoral amputees with short and long residual limbs were divided into two groups and subjected to three-dimensional quantitative motion analysis through high definition optoelectronic cameras and force platform. The prosthetic gait of amputees was compared with those of 15 participants with normal gait to analyze the asymmetry. There was a significant difference in temporal-spatial gait parameters between the amputee groups in terms of intact limb stride length and swing time ($p<0.005$). The gait of amputees with long femoral lengths was with comparatively higher mean step length (0.31 versus 0.35m), velocity (0.48 versus 0.44m/s), and cadence (73.58 versus 66.62 steps/min) compared to short residual limbs, however these results are nonsignificant ($p>0.05$). Amputees and able-bodied participants differed significantly for all measured parameters ($p<0.05$). The amputees with short residual limbs had more variation in sagittal plane hip flexion at initial contact than long amputees and able-bodied subjects. Amputees with longer residual limbs demonstrated peak swing prosthetic hip flexion closer to able-bodied individuals. The frontal pelvic obliquity and sagittal pelvic tilt revealed more variance among amputees with shorter residual limbs than longer ones. This study will aid in identifying the underlying mechanism of gait asymmetry in transfemoral amputees in relation to residual femoral length and hence will pave a path for improving the knee and foot prosthetic components.

Soft tissue is an important operation object in robot-assisted surgery and its mechanical response is of great significance to the precision of surgical operation. Accurate prediction of soft tissue mechanical properties can effectively avoid the potential damage of biological tissue caused by excessive operating force. In this paper, three typical mechanical responses of soft tissue were obtained by no-slip compression experiments of liver tissue. Second, the prediction model of soft tissue mechanical response based on gray prediction was established, and the influence of key parameters of the model on the precision of mechanical response prediction was analyzed. The results show that the gray prediction model can accurately predict the mechanical response of soft tissue, and the prediction accuracy is the highest when the number of historical data is 7. The prediction method of soft tissue mechanical response proposed in this paper will provide important data reference for accurate operation of surgery.

Finite element models of the second-stage labor system have been commonly developed for providing objective and quantitative indicators as well as innovative therapeutic solutions for decision supports. However, the reliability of the simulation outcomes remains a challenging issue due to uncertainties in input data and model complexity as well as the lack of validation. The objective of this study was to perform uncertainty quantification (UQ) on the material properties of the pelvis soft tissue with a focus on the uterus tissue during the second labor simulation leading to explore more plausible outcome space for reliable decision support making. The developed modeling and simulation workflow includes an image-based finite element model of the fetal body and pelvis soft tissues (floor, vagina and uterus), an uncertainty modeling procedure using precise and imprecise probabilities and an uncertainty propagation process based on the Monte Carlo method with and without parameter dependency. Obtained results showed that hyperelastic properties of the uterus tissue are very sensitive during the second stage of labor simulation. Moreover, the use of imprecise probability and parameter dependency lead to a more consistent range of values for uterus tissue stress analysis. This study performed, for the first time, an UQ on the hyperelastic properties of the uterus tissue from the in silico simulation of the second-stage labor. This opens new avenues for providing reliable indicators for clinical decision support. As a perspective, the active uterus behavior will be integrated into a more realistic second-stage labor model and simulation. Then, UQ will be conducted for more reliable decision support.

Stroke is a disease that is caused due to the blockage and burst in the blood vessels of the brain, thus resulting in abrupt brain dysfunction, like sensory or motor disorders, unconsciousness, limb paralysis, and pronunciation disorders. The existing stroke prediction algorithms have some limitations because of the lengthy testing procedures and hefty testing expenses. The main goal of this study is to develop and implement the proposed fusion-based, optimized deep learning model for stroke disease prediction using multimodalities. For that, this research considers the Computed Tomography (CT) and electroencephalogram (EEG) signals as input, and all of these inputs are processed separately to predict the stroke disease. While predicting the stroke disease with a CT image, the bilateral filter performs the pre-processing and the disease prediction is done with the DenseNet model, which is tuned by the proposed Jaya Fractional Reptile Search Algorithm (Jaya FRSA). Similar to how the proposed FRSA does CNN-LSTM training, the Convolutional Neural Network-Long Short Term Memory (CNN-LSTM) predicts the stroke disease using the EEG data as an input after the Gaussian filter removes signal noise. Additionally, the CT image and the EEG signal are processed independently from the image and signal properties. Additionally, the CNN-LSTM model and DenseNet model results are combined using the overlap coefficient to get the final disease prediction. According to the experimental study, the suggested method achieved the maximum image accuracy, sensitivity, and specificity of 0.924, 0.930, and 0.935.

Herein, a method was proposed for testing the dynamic tensile properties of a sports medicine suture and investigating the fatigue properties and elongation of the suture after a fatigue test. A test device was designed to test the dynamic tensile properties of sutures. The endurance limit of Nos. 1 and 4 sutures, which were used for laboratory comparisons, was $23.7\pm 1.4$N and $68.5\pm 2.3$N and the elongation rate was $3.3\%\pm 0.6\%,4.0\%\pm 0.3\%$, respectively. Furthermore, the Z-score, which is the offset divided by the standard deviation, was $<2$. The proposed test method is accurate and the repeatability is satisfied, making it suitable for determining the dynamic tensile properties of sutures. It is used as an industry standard to provide technical support for supervising sports medicine sutures.

Cell deformation and changes in mechanical properties, such as adhesion, elasticity modulus, and height, are some of the most significant signs of cancer. This paper investigates the mechanical parameters of normal and cancerous cells in the breast and lungs. The lung’s normal cell studied is MRC-5, and the cancerous ones include A-549, COR-L105, and CALU-6. MCF-10A is the normal breast cell studied, and MDA-MB-468, MDA-MB-231, and ZR-75-1 are cancerous. The mechanical specifications and cellular topography were obtained using nanoindentation by the atomic force microscope (AFM). The elasticity modulus and adhesion amounts for at least 96 indents were computed by mathematical averaging for each cell line in two different modes. The results indicate that cancer decreases the cells’ elasticity modulus by one-tenth in the lung and lower than one-third in breast cells. Cancer cells are up to several hundred times more adhesive than normal cells. In addition, a cancerous cell’s height is greater than normal cells. This study’s elastic modules are compared to the other references. Overall, the results are compatible.

Late detection of oral/laryngeal cancers or squamous cell carcinoma results in high patient mortality. Therefore, the detection of early-stage disease symptoms and timely medical treatment are important for improving long-term survival rates. Here, three deep learning models (single-shot detector, Yolo V4 and Tiny Yolo) were developed for target detection and binary type classification (normal/suspicious) for four representative oral/laryngeal regions (tongue, epiglottis, vocal cords, and tonsils) with a single-inspection process. The model performance was evaluated quantitatively on desktop and embedded platforms. We collected 1,632 endoscopic still-images and 20 diagnostic videos from the hospital database to train, validate, and test the models. Experimental results demonstrated that implemented models showed F1-scores ranging between 0.74–0.86, 0.86–1.00, and 0.74–0.87, and average precision ranging between 0.60–0.82, 0.92–1.00, and 0.72–0.98 for the tongue, epiglottis, and vocal cords, respectively, on the desktop platform. In addition, the Yolo V4 model showed performances of 0.92, 0.82, and 2.00 frames per second for the F1-score, average precision, and inference speed, respectively, on the embedded platform. Based on these results, we conclude that the implemented deep-learning-based at-home self-prescreening technique may be a reliable tool for personal oral/laryngeal healthcare, which will be especially important in endemic situations.

In this study, finite element (FE) models of an anatomical normal and a degenerated disc of L4–L5 motion segment, developed using different methodologies of volume mesh constructions, and ways of defining the spinal components’ material properties, were exercised and analyzed. The geometrics details were obtained using digitizing technique and segmentation of computed tomography (CT) scans; with material and structural properties of the spinal components defined based on the literature and greyscale values of CT images for the normal and degenerated L4–L5 motion segments, respectively. The predicted kinematic responses under five different physiological loadings in three anatomical reference planes for both models together with published data were analyzed. The predicted results correlated well within the range and in similar trends with published experimental results. The overall predicted responses showed different volume mesh constructions and the representation of spinal components’ material properties were accurate enough to predict the kinematic responses, although experimental results exhibit considerable divergence due to different experimental techniques. Both models exhibited different rotational and translation segmental stiffness in each corresponding loading configuration. The normal L4–L5 segment exhibited greatest axial torsional stiffness, and least extensional stiffness. The degenerated L4–L5 segment was least flexible (greatest stiffness) in extension and most flexible (least stiffness) in flexion. For compression, the axial stiffness of segment of normal segment was about 10% larger than that of the degenerated segment. The en-bloc efforts of all spinal components have contributed to the different nonlinear kinematic characteristics and segmental stiffness of these different L4–L5 motion segments.

The pandemic disease Coronavirus 2019 (COVID-19) caused thousands of infections and deaths globally. It is important to introduce new medicines to address the critical situation in the medical system. The determination of approximate pIC value is necessary for designing medicines based on molecular compounds. Generally, the approximation of pIC value is a lengthy process, so it is difficult and time-consuming. Hence it is essential to introduce a new technique for automatic approximation. In this research, a Convolutional Neural Network-based transfer learning (CNN-TL) is designed for approximating the pIC value. Initially, Simplified Molecular Input Line Entry System (SMILES) notation is extracted from SMILES string symbols using an entropy-based one-hot encoding matrix and the molecular formula-based encoding. The molecular features are then extracted from the input data using Lorentzian similarity and Deep Residual Network (DRN). The pIC value approximation is performed using the CNN-TL model, where the Visual Geometry Group Network-16 (VGGNet-16) is used to fetch hyperparameters used to initialize the CNN. The experimental results proved that the designed CNN-TL technique achieved minimum error rates with normalized values of 0.406 for R², 0.516 for Root Mean Square Error (RMSE), 0.267 for Mean Square Error (MSE), and for 0.277 Mean Absolute Percentage Error (MAPE).

In this paper, we propose a new electrocardiogram (ECG) denoising technique based on the application of a 1D double-density complex discrete wavelet transform (DWT) denoising method in the stationary bionic wavelet transform (SBWT) domain. SBWT was introduced in our previous work for speech enhancement. The first step in this proposed ECG denoising technique consists of applying the SBWT to the noisy ECG signal to obtain two noisy wavelet coefficients wtb1 and wtb2. The coefficients wtb1 and wtb2 are, respectively, the details and approximation coefficients. To estimate the level, $\sigma$, of the additive white noise corrupting the original ECG signal, we use wtb1 which is then thresholded using soft thresholding function. The noisy approximation wtb2 is denoised by using 1D double-density complex DWT denoising method. The latter requires the use of this noise level, $\sigma$. The results obtained from SNR computations show the performance of the ECG denoising technique proposed in this work.