Collagen cross-linking enhances the mechanical properties of bone by improving toughness of the bone matrix. Randomized electrical stimulation contributes positively to the behavior of osteoblastic cell in vitro by promoting osteogenic differentiation and mineral deposition. However, the effects of randomized electrical stimulation on collagen crosslink formation, which are associated with bone toughness, are yet to be elucidated. In this study, we identify the effects of randomized electrical stimulation on collagen cross-linking of osteoblast cells in vitro. MC3T3-E1 osteoblastic cells are stimulated with two electric stimulation (ES) patterns: a periodic pulse train (PrPT) with a constant pulse rate of 500 Hz, and a random pulse train (RdPT) composed of randomized pulse rates up to 500 Hz for 3 min/day for the first 3 days. After the stimulation, lysyl oxidase (LOX) and type I collagen (COL) gene expression, LOX activity, and collagen insolubility are investigated. In addition, Fourier transform infrared spectroscopy (FTIR) is performed to evaluate the degree of collagen maturity, i.e., collagen cross-linking. RdPT ES does not affect LOX and COL mRNA expression levels after a single stimulation of 3 min. However, on day 14 of RdPT ES, the LOX activity is significantly increased compared with the control. The ratio of insoluble collagen in the RdPT-stimulated sample is significantly higher than that of the control on day 21. FTIR exhibits significantly higher collagen maturity (1652/1680 cm⁻¹ area ratio) in the RdPT-stimulated sample than in the control and PrPT groups on day 21. In conclusion, randomized electrical stimulation promotes bone collagen cross-linking in osteoblastic cells through the activation of LOX.

This study aims to establish a monocular camera-based system that uses a generative adversarial network (GAN) to estimate a 3D pose of a human and his/her orientation relative to the camera from images by considering anatomical knowledge such as segment length and joint range of motion. The proposed network is trained by unsupervised learning using only 2D joint positions as training data, i.e., does not require the ground truth data on 3D joint positions and angles in real space for training. Unsupervised learning of the proposed network was performed based on a new loss function consisting of the typical GAN loss function and three new terms, which provide constraints on the quaternion norm, joint range of motion, and the similarity between the real and fake 2D poses, respectively. Numerical validation was performed using the Human3.6M human pose dataset, which includes real-space measured images, 2D pose, and 3D joint angles and positions. The results show that the proposed network is slightly less accurate than the depth estimation method obtained by supervised learning, but is as accurate as the depth estimation method obtained by unsupervised learning using GAN. However, qualitative comparisons in plots of 3D pose suggest that the joint range of motion constraints introduced in this paper are effective in estimating 3D pose without anatomical failures. Particularly in the scenes with large flexion of the upper and lower limbs, our network can avoid anatomical failures in the estimated 3D pose while the depth estimation methods could not. In addition, the proposed network can adjust itself adaptively for cameras with unknown external parameters.

Prediction models for post-embolization recurrence with hemodynamic parameters from computational fluid dynamics (CFD) were widely studied to manage cerebral aneurysms. However, only spatiotemporally averaged or maximal scalars were used to develop the models. The hemodynamic information was suppressed from 3-dimensional (3D) to 0D after averaging or maximizing, and its fidelity was strongly dependent upon the accuracy of geometry, boundary conditions, and model parameters in CFD. We designed a deep learning network, Velocity-PointNet (Vel-PointNet), to predict the recurrence probability by extracting 3D morphological and hemodynamic features in aneurysm. Geometries of 52 internal carotid-posterior communicating (ICPC) aneurysms (eight recanalized and 44 stable) were acquired from our clinical study. The blood flow was simulated using CFD. Vel-PointNet was trained with 3D morphological-hemodynamic data from CFD results. Our Vel-PointNet model was compared to existing machine learning (ML) models using significant morphological-hemodynamic scalars to verify our assumption regarding the advantage of 3D features over 0D features. Furthermore, the performance of ML models trained with morphological scalars and conventional PointNet that extracted 3D morphological features was evaluated to verify importance of hemodynamic parameters in predictive models. The area under receiver-operating characteristic curve (AUC) and precision recall curve (AUPRC) of Vel-PointNet (1.000/1.000) was higher than the other models. As a result of extracting both morphological and hemodynamic features in 3D, Vel-PointNet was found to be more accurate than traditional approaches at predicting post-embolization recurrence of ICPC aneurysms.

Optical coherence tomography (OCT) is a three-dimensional imaging technique based on low coherence interferometry of near-infrared broadband light and has been applied to medical diagnoses, such as structural diagnosis of the retina in ophthalmology. For several decades, functional variations of OCT have been also developed. OCT angiography (OCTA) is one of the functional variations of OCT and can visualize blood flows (microvasculature) in biological tissues by analyzing the time-series of OCT signals. Recently, several studies suggested the possibility that OCTA can not only visualize blood flows but also detect the blood flow velocity quantitatively, where the blood flow velocity is estimated by analyzing OCTA signals computed with multiple time separations from the time-series of OCT signals measured at a short time interval. OCTA signals computed with short time separations, however, reduce its contrast between dynamic tissues (blood flows) and other static tissues, resulting in the decrease of signal-to-noise (SN) ratio of the flow detection. This study proposed a novel method to enhance the SN ratio of the flow detection from such OCTA signals computed with short time separations, by means of mapping the gradient of OCTA signals against the time separations. In the experiments, a flow phantom and human skins were employed to compare SN ratios among OCTA with a long time separation (standard OCTA), OCTA with a short time separation, and the proposed method. The results showed that the proposed methods can detect flow signals with a high SN ratio equivalent to that of the standard OCTA and, therefore, suggested that quantitative detection of the flow velocity with OCTA will be possible with a high SN ratio of flow detection.

Fibrosis involves the abnormal accumulation of components of the extracellular matrix, such as collagen, and can lead to organ failure, which accounts for approximately one-third of all deaths worldwide. Accordingly, there is an urgent need to develop low-invasive treatment strategies for fibrosis as there is currently a lack of anti-fibrotic pharmacological agents available. In the present study, we tested the hypothesis that applying traction to fibrous tissues while using heat to denature proteins can be used to adjust the shape and flexibility of tissues. Bovine Achilles tendon samples as a model of fibrous tissue contracted by approximately 25% compared to the original length in response to heating, in line with previous findings. In contrast, the application of traction during heating and cooling increased tissue length by 25% and decreased Young’s modulus by 16%. Tissue staining and detection of unfolded collagen using collagen-hybridizing peptide demonstrated thermal degradation of collagen fibers and conversion to gelatinous material throughout samples, confirming that traction, and heating damages collagen. These results indicate that this technique can be used to alter the shape and reduce the rigidity of organs affected by fibrosis.