Deep learning-driven diagnosis of multi-type vertebra diseases based on computed tomography images.

Background: Osteoporotic vertebral compression fractures (OVCFs) are the most common type of fragility fracture. Distinguishing between OVCFs and other types of vertebra diseases, such as old fractures (OFs), Schmorl's node (SN), Kummell's disease (KD), and previous surgery (PS), is critical for subsequent surgery and treatment. Combining with advanced deep learning (DL) technologies, this study plans to develop a DL-driven diagnostic system for diagnosing multi-type vertebra diseases. Methods: We established a large-scale dataset based on the computed tomography (CT) images of 1,051 patients with OVCFs from Luhe Hospital and used data of 46 patients from Xuanwu Hospital as alternative hospital validation dataset. Each patient underwent one examination. The dataset contained 11,417 CT slices and 19,718 manually annotated vertebrae with diseases. A two-stage DL-based system was developed to diagnose five vertebra diseases. The proposed system consisted of a vertebra detection module (VDModule) and a vertebra classification module (VCModule). Results: The training and testing dataset for the VDModule consisted of 9,135 and 3,212 vertebrae, respectively. The VDModule using the ResNet18-based Faster region-based convolutional neural network (R-CNN) model achieved an area under the curve (AUC), false-positive (FP) rate, and false-negative (FN) rate of 0.982, 1.52%, and 1.33%, respectively, in the testing dataset. The training dataset for VCModule consisted of 14,584 and 47,604 diseased and normal vertebrae, respectively. The testing dataset consisted of 4,489 and 15,122 diseased and normal vertebrae, respectively. The ResNet50-based VCModule achieved an average sensitivity and specificity of 0.919 and 0.995, respectively, in diagnosing four kinds of vertebra diseases except for SN in the testing dataset. In the alternative hospital validation dataset, the ResNet50-based VCModule achieved an average sensitivity and specificity of 0.891 and 0.989, respectively, in diagnosing four kinds of vertebra diseases except for SN. Conclusions: Our proposed DL system can accurately diagnose four vertebra diseases and has strong potential to facilitate the accurate and rapid diagnosis of vertebral diseases.

Substitution for In Vitro and In Vivo Tests: Computational Models from Cell Attachment to Tissue Regeneration.

To get an optimal product of orthopaedic implant or regenerative medicine needs to follow trial-and-error analyses to investigate suitable product's material, structure, mechanical properites etc. The whole process from in vivo tests to clinical trials is expensive and time-consuming. Computational model is seen as a useful analysis tool to make the product development. A series of models for simulating tissue engineering process from cell attachment to tissue regeneration are reviewed. The challenging is that models for simulating tissue engineering processes are developed separately. From cell to tissue regeneration, it would go through blood injection after moving out the defect; to cell disperse and attach on the scaffold; to proliferation, migration and differentiation; and to the final part-becoming mature tissues. This paper reviewed models that related to tissue engineering process, aiming to provide an opportunity for researchers to develop a mature model for whole tissue engineering process. This article focuses on the model analysis methods of cell adhesion, nutrient transport and cell proliferation, differentiation and migration in tissue engineering. In cell adhesion model, one of the most accurate method is to use discrete phase model to govern cell movement and use Stanton-Rutland model for simulating cell attachment. As for nutrient transport model, numerical model coupling with volume of fluid model and species transport model together is suitable for predicting nutrient transport process. For cell proliferation, differentiation and migration, finite element method with random-walk algorithm is one the most advanced way to simulate these processes. Most of the model analysis methods require further experiments to verify the accuracy and effectiveness. Due to the lack of technology to detect the rate of nutrient diffusion, there are especially few researches on model analysis methods in the area of blood coagulation. Therefore, there is still a lot of work to be done in the research of the whole process model method of tissue engineering. In the future, the numerical model would be seen as an optimal way to investigate tissue engineering products bioperformance and also enable to optimize the parameters and material types of the tissue engineering products.

[Design of minimally invasive surgery wrist institution actuated by shape memory alloy].

The rapid development of minimally invasive surgery technology requires higher flexibility of surgical treatment and small volume of medical instrument. This paper proposed a new type of minimally invasive surgery wrist institution actuated by TiNi shape memory alloy (SMA) wire. The wrist institution has some advantages such as compact structure, flexible function, light weight, big movement space, and high output position precision. The paper briefly introduces the properties of TiNi SMA and describes the configuration of wrist institution. We also carried out mechanism simulation analysis to the mechanics model and set up kinematics equations, and finally presented the workspace of the institution.

An accurately controlled antagonistic shape memory alloy actuator with self-sensing.

With the progress of miniaturization, shape memory alloy (SMA) actuators exhibit high energy density, self-sensing ability and ease of fabrication, which make them well suited for practical applications. This paper presents a self-sensing controlled actuator drive that was designed using antagonistic pairs of SMA wires. Under a certain pre-strain and duty cycle, the stress between two wires becomes constant. Meanwhile, the strain to resistance curve can minimize the hysteresis gap between the heating and the cooling paths. The curves of both wires are then modeled by fitting polynomials such that the measured resistance can be used directly to determine the difference between the testing values and the target strain. The hysteresis model of strains to duty cycle difference has been used as compensation. Accurate control is demonstrated through step response and sinusoidal tracking. The experimental results show that, under a combination control program, the root-mean-square error can be reduced to 1.093%. The limited bandwidth of the frequency is estimated to be 0.15 Hz. Two sets of instruments with three degrees of freedom are illustrated to show how this type actuator could be potentially implemented.