Foldover-free shape deformation for biomedicine.

Shape deformation as a fundamental geometric operation underpins a wide range of applications, from geometric modelling, medical imaging to biomechanics. In medical imaging, for example, to quantify the difference between two corresponding images, 2D or 3D, one needs to find the deformation between both images. However, such deformations, particularly deforming complex volume datasets, are prone to the problem of foldover, i.e. during deformation, the required property of one-to-one mapping no longer holds for some points. Despite numerous research efforts, the construction of a mathematically robust foldover-free solution subject to positional constraints remains open. In this paper, we address this challenge by developing a radial basis function-based deformation method. In particular we formulate an effective iterative mechanism which ensures the foldover-free property is satisfied all the time. The experimental results suggest that the resulting deformations meet the internal positional constraints. In addition to radial basis functions, this iterative mechanism can also be incorporated into other deformation approaches, e.g. B-spline based FFDs, to develop different deformable approaches for various applications.

Face image-sketch synthesis via generative adversarial fusion.

Face image-sketch synthesis is widely applied in law enforcement and digital entertainment fields. Despite the extensive progression in face image-sketch synthesis, there are few methods focusing on generating a color face image from a sketch. The existing methods pay less attention to learning the illumination or highlight distribution on the face region. However, the illumination is the key factor that makes the generated color face image looks vivid and realistic. Moreover, existing methods tend to employ some image preprocessing technologies and facial region patching approaches to generate high-quality face images, which results in the high complexity and memory consumption in practice. In this paper, we propose a novel end-to-end generative adversarial fusion model, called GAF, which fuses two U-Net generators and a discriminator by jointly learning the content and adversarial loss functions. In particular, we propose a parametric tanh activation function to learn and control illumination highlight distribution over faces, which is integrated between the two U-Net generators by an illumination distribution layer. Additionally, we fuse the attention mechanism into the second U-Net generator of GAF to keep the identity consistency and refine the generated facial details. The qualitative and quantitative experiments on the public benchmark datasets show that the proposed GAF has better performance than existing image-sketch synthesis methods in synthesized face image quality (FSIM) and face recognition accuracy (NLDA). Meanwhile, the good generalization ability of GAF has also been verified. To further demonstrate the reliability and authenticity of face images generated using GAF, we use the generated face image to attack the well-known face recognition system. The result shows that the face images generated by GAF can maintain identity consistency and well maintain everyone's unique facial characteristics, which can be further used in the benchmark of facial spoofing. Moreover, the experiments are implemented to verify the effectiveness and rationality of the proposed parametric tanh activation function and attention mechanism in GAF.

Numerical Response Surfaces of Volume of Ablation and Retropulsion Amplitude by Settings of Ho:YAG Laser Lithotripter.

Objectives: Although laser lithotripsy is now the preferred treatment option for urolithiasis due to shorter operation time and a better stone-free rate, the optimal laser settings for URS (ureteroscopic lithotripsy) for less operation time remain unclear. The aim of this study was to look for quantitative responses of calculus ablation and retropulsion by performing operator-independent experiments to determine the best fit versus the pulse energy, pulse width, and the number of pulses. Methods: A lab-built Ho:YAG laser was used as the laser pulse source, with a pulse energy from 0.2 J up to 3.0 J and a pulse width of 150 µs up to 1000 µs. The retropulsion was monitored using a high-speed camera, and the laser-induced craters were evaluated with a 3-D digital microscope. The best fit to the experimental data is done by a design of experiment software. Results: The numerical formulas for the response surfaces of ablation speed and retropulsion amplitude are generated. Conclusions: The longer the pulse, the less the ablation or retropulsion, while the longer pulse makes the ablation decrease faster than the retropulsion. The best quadratic fit of the response surface for the volume of ablation varied nonlinearly with pulse duration and pulse number.

Multiphase Interface Tracking with Fast Semi-Lagrangian Contouring.

We propose a semi-Lagrangian method for multiphase interface tracking. In contrast to previous methods, our method maintains an explicit polygonal mesh, which is reconstructed from an unsigned distance function and an indicator function, to track the interface of arbitrary number of phases. The surface mesh is reconstructed at each step using an efficient multiphase polygonization procedure with precomputed stencils while the distance and indicator function are updated with an accurate semi-Lagrangian path tracing from the meshes of the last step. Furthermore, we provide an adaptive data structure, multiphase distance tree, to accelerate the updating of both the distance function and the indicator function. In addition, the adaptive structure also enables us to contour the distance tree accurately with simple bisection techniques. The major advantage of our method is that it can easily handle topological changes without ambiguities and preserve both the sharp features and the volume well. We will evaluate its efficiency, accuracy and robustness in the results part with several examples.

Mining Spatial-Temporal Patterns and Structural Sparsity for Human Motion Data Denoising.

Motion capture is an important technique with a wide range of applications in areas such as computer vision, computer animation, film production, and medical rehabilitation. Even with the professional motion capture systems, the acquired raw data mostly contain inevitable noises and outliers. To denoise the data, numerous methods have been developed, while this problem still remains a challenge due to the high complexity of human motion and the diversity of real-life situations. In this paper, we propose a data-driven-based robust human motion denoising approach by mining the spatial-temporal patterns and the structural sparsity embedded in motion data. We first replace the regularly used entire pose model with a much fine-grained partlet model as feature representation to exploit the abundant local body part posture and movement similarities. Then, a robust dictionary learning algorithm is proposed to learn multiple compact and representative motion dictionaries from the training data in parallel. Finally, we reformulate the human motion denoising problem as a robust structured sparse coding problem in which both the noise distribution information and the temporal smoothness property of human motion have been jointly taken into account. Compared with several state-of-the-art motion denoising methods on both the synthetic and real noisy motion data, our method consistently yields better performance than its counterparts. The outputs of our approach are much more stable than that of the others. In addition, it is much easier to setup the training dataset of our method than that of the other data-driven-based methods.

Water content contribution in calculus phantom ablation during Q-switched Tm:YAG laser lithotripsy.

Q-switched (QS) Tm:YAG laser ablation mechanisms on urinary calculi are still unclear to researchers. Here, dependence of water content in calculus phantom on calculus ablation performance was investigated. White gypsum cement was used as a calculus phantom model. The calculus phantoms were ablated by a total 3-J laser pulse exposure (20 mJ, 100 Hz, 1.5 s) and contact mode with N=15 sample size. Ablation volume was obtained on average 0.079, 0.122, and 0.391 mm3 in dry calculus in air, wet calculus in air, and wet calculus in-water groups, respectively. There were three proposed ablation mechanisms that could explain the effect of water content in calculus phantom on calculus ablation performance, including shock wave due to laser pulse injection and bubble collapse, spallation, and microexplosion. Increased absorption coefficient of wet calculus can cause stronger spallation process compared with that caused by dry calculus; as a result, higher calculus ablation was observed in both wet calculus in air and wet calculus in water. The test result also indicates that the shock waves generated by short laser pulse under the in-water condition have great impact on the ablation volume by Tm:YAG QS laser.

Virtual reality training and assessment in laparoscopic rectum surgery.

BACKGROUND: Virtual-reality (VR) based simulation techniques offer an efficient and low cost alternative to conventional surgery training. This article describes a VR training and assessment system in laparoscopic rectum surgery. METHODS: To give a realistic visual performance of interaction between membrane tissue and surgery tools, a generalized cylinder based collision detection and a multi-layer mass-spring model are presented. A dynamic assessment model is also designed for hierarchy training evaluation. RESULTS: With this simulator, trainees can operate on the virtual rectum with both visual and haptic sensation feedback simultaneously. The system also offers surgeons instructions in real time when improper manipulation happens. The simulator has been tested and evaluated by ten subjects. CONCLUSIONS: This prototype system has been verified by colorectal surgeons through a pilot study. They believe the visual performance and the tactile feedback are realistic. It exhibits the potential to effectively improve the surgical skills of trainee surgeons and significantly shorten their learning curve.

Motion adaptation with motor invariant theory.

Bipedal walking is not fully understood. Motion generated from methods employed in robotics literature is stiff and is not nearly as energy efficient as what we observe in nature. In this paper, we propose validity conditions for motion adaptation from biological principles in terms of the topology of the dynamic system. This allows us to provide a closed-form solution to the problem of motion adaptation to environmental perturbations. We define both global and local controllers that improve structural and state stability, respectively. Global control is achieved by coupling the dynamic system with a neural oscillator, which preserves the periodic structure of the motion primitive and ensures stability by entrainment. A group action derived from Lie group symmetry is introduced as a local control that transforms the underlying state space while preserving certain motor invariants. We verify our method by evaluating the stability and energy consumption of a synthetic passive dynamic walker and compare this with motion data of a real walker. We also demonstrate that our method can be applied to a variety of systems.

Bystander responses to a violent incident in an immersive virtual environment.

Under what conditions will a bystander intervene to try to stop a violent attack by one person on another? It is generally believed that the greater the size of the crowd of bystanders, the less the chance that any of them will intervene. A complementary model is that social identity is critical as an explanatory variable. For example, when the bystander shares common social identity with the victim the probability of intervention is enhanced, other things being equal. However, it is generally not possible to study such hypotheses experimentally for practical and ethical reasons. Here we show that an experiment that depicts a violent incident at life-size in immersive virtual reality lends support to the social identity explanation. 40 male supporters of Arsenal Football Club in England were recruited for a two-factor between-groups experiment: the victim was either an Arsenal supporter or not (in-group/out-group), and looked towards the participant for help or not during the confrontation. The response variables were the numbers of verbal and physical interventions by the participant during the violent argument. The number of physical interventions had a significantly greater mean in the in-group condition compared to the out-group. The more that participants perceived that the Victim was looking to them for help the greater the number of interventions in the in-group but not in the out-group. These results are supported by standard statistical analysis of variance, with more detailed findings obtained by a symbolic regression procedure based on genetic programming. Verbal interventions made during their experience, and analysis of post-experiment interview data suggest that in-group members were more prone to confrontational intervention compared to the out-group who were more prone to make statements to try to diffuse the situation.

A critical challenge: dosage-related efficacy and acute complication intracoronary injection of autologous bone marrow mesenchymal stem cells in acute myocardial infarction.

BACKGROUND: Previous studies showed improvement in heart function by injecting bone marrow mesenchymal stem cells (BMSCs) after AMI. Emerging evidence suggested that both the number and function of BMSCs decline with ageing. We designed a randomized, controlled trial to further investigate the safety and efficacy of this treatment. METHODS: Patients with ST-elevation AMI undergoing successful reperfusion treatment within 12 hours were randomly assigned to receive an intracoronary infusion of BMSCs (n=21) or standard medical treatment (n=22) (the numbers of patients were limited because of the complication of coronary artery obstruction). RESULTS: There is a closely positive correlation of the number and function of BMSCs vs. the cardiac function reflected by LVEF at baseline (r=0.679, P=0.001) and at 12-month follow-up (r=0.477, P=0.039). Six months after cell administration, myocardial viability within the infarct area by 18-FDG SPECT was improved in both groups compared with baseline, but no significant difference in the BMSCs compared with control groups (4.0±0.4% 95%CI 3.1-4.9 vs. 3.2±0.5% 95%CI 2.1-4.3, P=0.237). 99mTc-sestamibi SPECT demonstrated that myocardial perfusion within the infarct area in the BMSCs did not differ from the control group (4.4±0.5% 95%CI 3.2-5.5 vs. 3.9±0.6% 95%CI 2.6-5.2, P=0.594). Similarly, LVEF after 12 and 24 months follow-up did not show any difference between the two groups. In the BMSCs group, one patient suffered a serious complication of coronary artery occlusion during the BMSCs injection procedure. CONCLUSIONS: The clinical benefits of intracoronary injection of autologous BMSCs in acute STEMI patients need further investigation and reevaluation.

An RBF-based reparameterization method for constrained texture mapping.

Texture mapping has long been used in computer graphics to enhance the realism of virtual scenes. However, to match the 3D model feature points with the corresponding pixels in a texture image, surface parameterization must satisfy specific positional constraints. However, despite numerous research efforts, the construction of a mathematically robust, foldover-free parameterization that is subject to positional constraints continues to be a challenge. In the present paper, this foldover problem is addressed by developing radial basis function (RBF)-based reparameterization. Given initial 2D embedding of a 3D surface, the proposed method can reparameterize 2D embedding into a foldover-free 2D mesh, satisfying a set of user-specified constraint points. In addition, this approach is mesh free. Therefore, generating smooth texture mapping results is possible without extra smoothing optimization.

Motion-Sensitive Anchor Identification of Least-Squares Meshes from Examples.

A least-squares mesh is a surface representation consisting of a small set of anchor points and the differential and topological properties of the surface. In this paper, we present a novel method to identify motion-sensitive anchor points for least-squares meshes from a set of examples. We present a new method, called clustered teleconnection analysis, to identify the maximally excited points in a subset of basis vectors deduced using principal component analysis. We demonstrate by means of examples that our approach has a smaller reconstruction error and equivalent performance to the current best approaches.

Graphic and haptic simulation system for virtual laparoscopic rectum surgery.

BACKGROUND: Medical simulators with vision and haptic feedback techniques offer a cost-effective and efficient alternative to the traditional medical trainings. They have been used to train doctors in many specialties of medicine, allowing tasks to be practised in a safe and repetitive manner. This paper describes a virtual-reality (VR) system which will help to influence surgeons' learning curves in the technically challenging field of laparoscopic surgery of the rectum. METHODS: Data from MRI of the rectum and real operation videos are used to construct the virtual models. A haptic force filter based on radial basis functions is designed to offer realistic and smooth force feedback. To handle collision detection efficiently, a hybrid model is presented to compute the deformation of intestines. Finally, a real-time cutting technique based on mesh is employed to represent the incision operation. RESULTS: Despite numerous research efforts, fast and realistic solutions of soft tissues with large deformation, such as intestines, prove extremely challenging. This paper introduces our latest contribution to this endeavour. With this system, the user can haptically operate with the virtual rectum and simultaneously watch the soft tissue deformation. CONCLUSIONS: Our system has been tested by colorectal surgeons who believe that the simulated tactile and visual feedbacks are realistic. It could replace the traditional training process and effectively transfer surgical skills to novices.