Anatomical Prior-Based Automatic Segmentation for Cardiac Substructures from Computed Tomography Images.

Cardiac substructure segmentation is a prerequisite for cardiac diagnosis and treatment, providing a basis for accurate calculation, modeling, and analysis of the entire cardiac structure. CT (computed tomography) imaging can be used for a noninvasive qualitative and quantitative evaluation of the cardiac anatomy and function. Cardiac substructures have diverse grayscales, fuzzy boundaries, irregular shapes, and variable locations. We designed a deep learning-based framework to improve the accuracy of the automatic segmentation of cardiac substructures. This framework integrates cardiac anatomical knowledge; it uses prior knowledge of the location, shape, and scale of cardiac substructures and separately processes the structures of different scales. Through two successive segmentation steps with a coarse-to-fine cascaded network, the more easily segmented substructures were coarsely segmented first; then, the more difficult substructures were finely segmented. The coarse segmentation result was used as prior information and combined with the original image as the input for the model. Anatomical knowledge of the large-scale substructures was embedded into the fine segmentation network to guide and train the small-scale substructures, achieving efficient and accurate segmentation of ten cardiac substructures. Sixty cardiac CT images and ten substructures manually delineated by experienced radiologists were retrospectively collected; the model was evaluated using the DSC (Dice similarity coefficient), Recall, Precision, and the Hausdorff distance. Compared with current mainstream segmentation models, our approach demonstrated significantly higher segmentation accuracy, with accurate segmentation of ten substructures of different shapes and sizes, indicating that the segmentation framework fused with prior anatomical knowledge has superior segmentation performance and can better segment small targets in multi-target segmentation tasks.

Uncovering nasopharyngeal carcinoma from chronic rhinosinusitis and healthy subjects using routine medical tests via machine learning.

Nasopharyngeal carcinoma (NPC) is one of the most common types of cancers in South China and Southeast Asia. Clinical data has shown that early detection is essential for improving treatment effectiveness and survival rate. Unfortunately, because the early symptoms of NPC are rather minor and similar to that of diseases such as Chronic Rhinosinusitis (CRS), early detection is a challenge. This paper proposes using machine learning methods to detect NPC using routine medical test data, namely Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Network (ANN), k-Nearest-Neighbor (KNN) and Logistic Regression (LR). We collected a dataset containing 523 newly diagnosed NPC patients before treatment, 501 newly diagnosed CRS patients before treatment as well as 600 healthy controls. The routine medical test data including age, gender, blood test features, liver function test features, and urine sediment test features. For comparison, we also used data from Epstein-Barr Virus (EBV) antibody tests, which is a specialized test not included among routine medical tests. In our first test, all four methods were tested on classifying NPC vs CRS vs controls; RF gives the best overall performance. Using only routine medical test data, it gives an accuracy of 83.1%, outperforming LR by 12%. In our second test, using only routine medical test data, when classifying NPC vs non-NPC (i.e. CRS or controls), RF achieves an accuracy of 88.2%. In our third test, when classifying NPC vs. controls, RF using only routine test data achieves an accuracy significantly better than RF using only EBV antibody data. Finally, in our last test, RF trained with NPC vs controls, using routine test data only, continued to perform well on an entirely separate dataset. This is a promising result because preliminary NPC detection using routine medical data is easy and inexpensive to implement. We believe this approach will play an important role in the detection and treatment of NPC in the future.

Automatic composition of Guzheng (Chinese Zither) music using long short-term memory network (LSTM) and reinforcement learning (RL).

In recent years, with the advance of Artificial Intelligence, automatic music composition has been demonstrated. However, there are many music genres and music instruments. For a same piece of music, different music instruments would produce different effects. Invented some 2500 years ago, Guzheng is one of the oldest music instruments in China and the world. It has distinct timbres and patterns that cannot be duplicated by other music instruments. Therefore, it is interesting to see whether AI can compose Guzheng music or alike. In this paper we present a method that can automatically compose and play Guzheng music. First, we collect a number of existing Guzheng music pieces and convert them into Music Instrument Digital Interface format. Second, we use these data to train a Long Short-Term Memory (LSTM) network and use the trained network to generate new Guzheng music pieces. Next, we use the Reinforcement Learning to optimize the LSTM network by adding special Guzheng playing techniques. Comparing to the existing AI methods, such as LSTM and Generative Adversary Network, our new method is more effective in capturing the characteristics of Guzheng music. According to the evaluations from skilled Guzheng players and general audiences, our Guzheng music is very close to the real Guzheng music. The presented method can also be used to automatically compose the music of other Chinese music instruments.

A Novel Design Method of Gradient Porous Structure for Stabilized and Lightweight Mandibular Prosthesis.

Compared to conventional prostheses with homogenous structures, a stress-optimized functionally gradient prosthesis will better adapt to the host bone due to its mechanical and biological advantages. Therefore, this study aimed to investigate the damage resistance of four regular lattice scaffolds and proposed a new gradient algorithm for stabilized and lightweight mandibular prostheses. Scaffolds with four configurations (regular hexahedron, regular octahedron, rhombic dodecahedron, and body-centered cubic) having different porosities underwent finite element analysis to select an optimal unit cell. Meanwhile, a homogenization algorithm was used to control the maximum stress and increase the porosity of the scaffold by adjusting the strut diameters, thereby avoiding fatigue failure and material wastage. Additionally, the effectiveness of the algorithm was verified by compression tests. The results showed that the load transmission capacity of the scaffold was strongly correlated with both configuration and porosity. Scaffolds with regular hexahedron unit cells can withstand stronger loads at the same porosity. The optimized gradient scaffold showed higher porosity and lower maximum stress than the target stress value, and the compression tests also confirmed the simulation results. A mandibular prosthesis was established using a regular hexahedron unit cell, and the strut diameters were gradually changed according to the proposed algorithm and the simulation results. Compared with the initial homogeneous prosthesis, the optimized gradient prosthesis reduced the maximum stress by 24.48% and increased the porosity by 6.82%, providing a better solution for mandibular reconstruction.

Novel Design and Optimization of Porous Titanium Structure for Mandibular Reconstruction.

A porous material is considered to be a potential material that can be used to repair bone defects. However, the methods of designing of a highly porous structure within the allowable stress range remain to be researched. Therefore, this study was aimed at presenting a method for generating a three-dimensional tetrahedral porous structure characterized by low peak stress and high porosity for the reconstruction of mandibular defects. Firstly, the initial tetrahedral porous structure was fabricated with the strut diameters set to 0.4 mm and a mean cell size of 2.4 mm in the design model space. Following this, the simulation analysis was carried out. Further, a homogenization algorithm was used for homogenizing the stress distribution, increasing porosity, and controlling peak stress of the porous structure by adjusting the strut diameters. The results showed that compared with the initial porous structure, the position of the large stress regions remained unchanged, and the peak stress fluctuated slightly in the mandible and fixation system with the optimized porous structure under two occlusions. The optimized porous structure had a higher porosity and more uniform stress distribution, and the maximum stress was lower than the target stress value. The design and optimization technique of the porous structure presented in this paper can be used to control peak stress, improve porosity, and fabricate a lightweight scaffold, which provides a potential solution for mandibular reconstruction.

Symmetry Analysis of Oriental Polygonal Pagodas Using 3D Point Clouds for Cultural Heritage.

Ancient pagodas are usually parts of hot tourist spots in many oriental countries due to their unique historical backgrounds. They are usually polygonal structures comprised by multiple floors, which are separated by eaves. In this paper, we propose a new method to investigate both the rotational and reflectional symmetry of such polygonal pagodas through developing novel geometric models to fit to the 3D point clouds obtained from photogrammetric reconstruction. The geometric model consists of multiple polygonal pyramid/prism models but has a common central axis. The method was verified by four datasets collected by an unmanned aerial vehicle (UAV) and a hand-held digital camera. The results indicate that the models fit accurately to the pagodas' point clouds. The symmetry was realized by rotating and reflecting the pagodas' point clouds after a complete leveling of the point cloud was achieved using the estimated central axes. The results show that there are RMSEs of 5.04 cm and 5.20 cm deviated from the perfect (theoretical) rotational and reflectional symmetries, respectively. This concludes that the examined pagodas are highly symmetric, both rotationally and reflectionally. The concept presented in the paper not only work for polygonal pagodas, but it can also be readily transformed and implemented for other applications for other pagoda-like objects such as transmission towers.

3D computational models explain muscle activation patterns and energetic functions of internal structures in fish swimming.

How muscles are used is a key to understanding the internal driving of fish swimming. However, the underlying mechanisms of some features of the muscle activation patterns and their differential appearance in different species are still obscure. In this study, we explain the muscle activation patterns by using 3D computational fluid dynamics models coupled to the motion of fish with prescribed deformation and examining the torque and power required along the fish body with two primary swimming modes. We find that the torque required by the hydrodynamic forces and body inertia exhibits a wave pattern that travels faster than the curvature wave in both anguilliform and carangiform swimmers, which can explain the traveling wave speeds of the muscle activations. Notably, intermittent negative power (i.e., power delivered by the fluid to the body) on the posterior part, along with a timely transfer of torque and energy by tendons, explains the decrease in the duration of muscle activation towards the tail. The torque contribution from the body elasticity further clarifies the wave speed increase or the reverse of the wave direction of the muscle activation on the posterior part of a carangiform swimmer. For anguilliform swimmers, the absence of the aforementioned changes in the muscle activation on the posterior part is consistent with our torque prediction and the absence of long tendons from experimental observations. These results provide novel insights into the functions of muscles and tendons as an integral part of the internal driving system, especially from an energy perspective, and they highlight the differences in the internal driving systems between the two primary swimming modes.

Three-dimensional printing of patient-specific surgical plates in head and neck reconstruction: A prospective pilot study.

BACKGROUND: Surgical plates have been extensively used in head and neck reconstruction and conventional plates are mass-produced with universal configurations. To overcome disadvantages of conventional surgical plates, we have been exploring patient-specific surgical plates using the three-dimensional (3D) printing technology. We hypothesized that the application of 3D-printed patient-specific surgical plates in head and neck reconstruction is feasible, safe and precise. METHODS: We are conducting a prospective clinical trial to assess the feasibility, safety and accuracy of applying 3D-printed patient-specific surgical plates in head and neck reconstruction. The primary endpoint was the intraoperative success rate. Secondary endpoints included the incidence and severity of postoperative adverse events within six months postoperatively. The accuracy of surgical outcomes was also explored by comparing the planned and final positions of the maxilla, mandible and grafted bone segments. RESULTS: From December 2016 to October 2017, ten patients were enrolled and underwent head and neck reconstruction using 3D-printed patient-specific surgical plates. The patient-specific surgical plates adapted to bone surface precisely and no plate-bending was performed. The intraoperative success rate was 100%. The average follow-up period was 6.5 months. No major adverse events were observed. The mean absolute distance deviation of integral mandible or maxilla was 1.40 ±â€¯0.63 mm, which showed a high accuracy of reconstruction. CONCLUSIONS: The 3D printing of patient-specific surgical plates could be effective in head and neck reconstruction. Surgical procedures were simplified. The precise jaw reconstruction was achieved with high accuracy. Long-term results with a larger sample size are warranted to support a final conclusion. The study protocol has been registered in ClinicalTrials.gov with a No. of NCT03057223.