工业工程最新文献

Prosthesis loosening is the leading cause of postoperative revision in unicompartmental knee arthroplasty (UKA). The deviation of medial and lateral translational installation of the prosthesis during surgery is a common clinical phenomenon and an important factor in increasing the risk of prosthesis loosening. This study established a UKA finite element model and a bone-prosthesis fixation interface micromotion prediction model. The predicted medial contact force and joint motion of the knee joint from a patient-specific lower extremity musculoskeletal multibody dynamics model of UKA were used as boundary conditions. The effects of 9 femoral component medial and lateral translational installation deviations on the Von Mises stress of the proximal tibia, the contact stress, and the micro-motion of the bone prosthesis fixation interface were quantitatively studied. It was found that compared with the neutral position (a/A of 0.492), the lateral translational deviation of the femoral component significantly increased the tibial Von Mises stress and the bone-prosthesis fixation interface contact stress. The maximum Von Mises stress and the maximum contact stress of the fixation interface increased by 14.08% and 143.15%, respectively, when a/A was 0.361. The medial translational deviation of the femoral component significantly increased the bone-prosthesis fixation interface micro-motion. The maximum value of micromotion under the conditions of femoral neutral and medial translation deviation was in the range of 20-50 μm, which is suitable for osseointegration. Therefore, based on considerations such as the micromotion range suitable for osseointegration reported in the literature, the risk of reducing prosthesis loosening, and factors that may induce pain, it is recommended that clinicians control the mounting position of the femoral component during surgery within the safe range of 0-4 mm medial translation deviation.

Traditional manual testing of ventilator performance is labor-intensive, time-consuming, and prone to errors in data recording, making it difficult to meet the current demands for testing efficiency in the development and manufacturing of ventilators. Therefore, in this study we designed an automated testing system for essential performance parameters of ventilators. The system mainly comprises a ventilator airflow analyzer, an automated switch module for simulated lungs, and a test control platform. Under the control of testing software, this system can perform automated tests of critical performance parameters of ventilators and generate a final test report. To validate the effectiveness of the designed system, tests were conducted on two different brands of ventilators under four different operating conditions, comparing tidal volume, oxygen concentration, and positive end expiratory pressure accuracy using both the automated testing system and traditional manual methods. Bland-Altman statistical analysis indicated good consistency between the accuracy of automated tests and manual tests for all respiratory parameters. In terms of testing efficiency, the automated testing system required approximately one-third of the time needed for manual testing. These results demonstrate that the designed automated testing system provides a novel approach and means for quality inspection and measurement calibration of ventilators, showing broad application prospects.

During transfer tasks, the dual-arm nursing-care robot require a human-robot mechanics model to determine the balance region to support the patient safely and stably. Previous studies utilized human-robot two-dimensional static equilibrium models, ignoring the human body volume and muscle torques, which decreased model accuracy and confined the robot ability to adjust the patient's posture in three-dimensional spatial. Therefore, this study proposes a three-dimensional spatial mechanics modeling method based on individualized human musculoskeletal multibody dynamics. Firstly, based on the mechanical features of dual-arm support, this study constructed a foundational three-dimensional human-robot mechanics model including body posture, contact position and body force. With the computed tomography data from subjects, a three-dimensional femur-pelvis-sacrum model was reconstructed, and the individualized musculoskeletal dynamics was analyzed using the ergonomics software, which derived the human joint forces and completed the mechanic model. Then, this study established a dual-arm robot transfer platform to conduct subject transfer experiments, showing that the constructed mechanics model possessed higher accuracy than previous methods. In summary, this study provides a three-dimensional human-robot mechanics model adapting to individual transfers, which has potential application in various scenarios such as nursing-care and rehabilitating robots.

Heart rate is a crucial indicator of human health with significant physiological importance. Traditional contact methods for measuring heart rate, such as electrocardiograph or wristbands, may not always meet the need for convenient health monitoring. Remote photoplethysmography (rPPG) provides a non-contact method for measuring heart rate and other physiological indicators by analyzing blood volume pulse signals. This approach is non-invasive, does not require direct contact, and allows for long-term healthcare monitoring. Deep learning has emerged as a powerful tool for processing complex image and video data, and has been increasingly employed to extract heart rate signals remotely. This article reviewed the latest research advancements in rPPG-based heart rate measurement using deep learning, summarized available public datasets, and explored future research directions and potential advancements in non-contact heart rate measurement.

The accurate segmentation of breast ultrasound images is an important precondition for the lesion determination. The existing segmentation approaches embrace massive parameters, sluggish inference speed, and huge memory consumption. To tackle this problem, we propose T ²KD Attention U-Net (dual-Teacher Knowledge Distillation Attention U-Net), a lightweight semantic segmentation method combined double-path joint distillation in breast ultrasound images. Primarily, we designed two teacher models to learn the fine-grained features from each class of images according to different feature representation and semantic information of benign and malignant breast lesions. Then we leveraged the joint distillation to train a lightweight student model. Finally, we constructed a novel weight balance loss to focus on the semantic feature of small objection, solving the unbalance problem of tumor and background. Specifically, the extensive experiments conducted on Dataset BUSI and Dataset B demonstrated that the T ²KD Attention U-Net outperformed various knowledge distillation counterparts. Concretely, the accuracy, recall, precision, Dice, and mIoU of proposed method were 95.26%, 86.23%, 85.09%, 83.59%and 77.78% on Dataset BUSI, respectively. And these performance indexes were 97.95%, 92.80%, 88.33%, 88.40% and 82.42% on Dataset B, respectively. Compared with other models, the performance of this model was significantly improved. Meanwhile, compared with the teacher model, the number, size, and complexity of student model were significantly reduced (2.2×10 ⁶ vs. 106.1×10 ⁶, 8.4 MB vs. 414 MB, 16.59 GFLOPs vs. 205.98 GFLOPs, respectively). Indeedy, the proposed model guarantees the performances while greatly decreasing the amount of computation, which provides a new method for the deployment of clinical medical scenarios.