Depth-based removal of thermal reflection with the light-field theory.

Thermal imaging is a useful imaging technique in many scenarios. It can capture the temperature distribution of scenes in the dark and see through sparse smoke and dust. However, some surfaces such as steel and glass with high reflectivity lead to a reflection problem in thermal imaging, while heavy mist and gases lead to the occlusion problem. We proposed an efficient algorithm to solve the occlusion problem in our earlier work. The reflection in thermal images causes errors in detection and temperature measurement. Therefore, the precise model and efficient algorithms to solve this problem are in high demand. In this paper, we mainly model the reflection problem in thermal imaging and propose an algorithm to deal with it. In our experiments, a thermal camera array is built to capture the thermal light-field images. We first separate a part of the reflection pixels from thermal images based on the depth information. After that, the thermal reflection is removed by optimizing a designed cost function. The experiment results show that our reflection removal method can separate the thermal reflection with high precision, retain the objects in the scene, and get better performance than existing methods.

Global disassortative rewiring strategy for enhancing the robustness of scale-free networks against localized attack.

The robustness of complex networks against attack has been an important issue for decades. Most of the previous studies focused on targeted attack (TA) and random attack (RA), while recently localized attack (LA) has drawn the attention of researchers. However, the existing studies related to LA mainly aim to reveal the properties on various network topologies so that the strategy to enhance network robustness against LA is still not well studied. In this paper, we propose a global disassortative rewiring strategy to enhance the robustness of scale-free networks against LA without changing the degree distribution. The validations are conducted on simulated scale-free networks and two real-life networks. As global disassortative rewiring strategy outperforms the other strategies, it can be proved effective in enhancing network robustness against LA and may contribute to future network risk reduction. In addition, by avoiding calculating and comparing the localized-robustness measurement within each rewire operation, our strategy offers a significant advantage in computational efficiency.

Biomedical Photoacoustic Imaging With Unknown Spatially Distributed Ultrasound Sensor Array.

OBJECTIVE: With the growth of interest in different medical study on biological function, non-invasive photoacoustic imaging of biological tissue attracts the interests for researchers. To eliminate the limited angle effect of photoacoustic imaging based on ultrasound linear array, spatially distributed ultrasound sensor array is applied. The accurate sensor array position determines the quality of the imaging results. In this study, we proposed three methods based on photoacoustic and ultrasound signals to enhance the imaging quality using a 256-element full-ring array. METHODS: Groups of photoacoustic and ultrasound signals are used to regress the position of each element sensor. RESULT: In phantom study and mouse brain study, photoacoustic imaging results can both yield details clearly with average error rate of less than 1% (50 [Formula: see text]). CONCLUSION: The performance of our three methods have proved that they can be potentially applied to other ultrasound-based medical imaging studies with unknown distributed positions of sensor array to enhance the imaging quality. SIGNIFICANCE: The proposed methods can contribute to precise biomedical imaging with unknown distributed positions of sensor array in different application scenarios.

Reliable Thyroid Carcinoma Detection with Real-Time Intelligent Analysis of Ultrasound Images.

Thyroid carcinoma is one of the most common endocrine diseases globally, and the incidence has been on the rise in recent years. Ultrasound imaging is the primary clinical method for early thyroid nodule diagnosis. Regions of interest (ROIs) of nodules in ultrasound images are difficult to detect because of their irregular shape nand vague margins. Accurate real-time thyroid nodule detection can provide ROIs for subsequent nodule diagnosis automatically, avoid variabilities between the subjective interpretations and inter-observer effectively and alleviate the workloads of medical practitioners. The aim of this study was to present a reliable, real-time detection method based on the Faster R-CNN (region-based convolutional network) framework for accurate and fast detection of thyroid nodules in ultrasound images. Our study proposed a faster and more accurate thyroid nodule detection method based on the Faster R-CNN framework by adding three strategies: feature pyramid, spatial remapping and anchor-box redesign. Specifically, the network takes raw ultrasound images as inputs and generates boxes with positions and the possibilities that these boxes contain thyroid nodules. The proposed method could locate and detect target nodules accurately with a mean average precision of 92.79% with more than 9000 patient images. In addition, the detection rate has accelerated to >16 frames per second, four times faster than that of the initial network. Therefore, it can meet the requirements of clinical application. The performance of the fivefold cross-validation was also accurate and robust. The proposed automatic thyroid nodule detection method yields better performance in accuracy and detection speed, which indicates the potential value of our method in assisting clinical ultrasound image interpretation.

Multiple Delay and Sum With Enveloping Beamforming Algorithm for Photoacoustic Imaging.

Delay and Sum (DAS) is one of the most common beamforming algorithms for photoacoustic imaging (PAI) reconstruction. Based on calculating beamformed signal with simple delaying and summing, DAS can function in a quick response and is quite suitable for real-time PAI. However, high sidelobes and intense artifacts may appear when using DAS due to summing with unnecessary data. In this paper, a beamforming algorithm called Multiple Delay and Sum with Enveloping (multi-DASE) is introduced to solve this problem. Compared to DAS, the multi-DASE algorithm calculates not only the initial value of the beamformed signal but also the complete N-shaped photoacoustic signal for each pixel. Through computer simulation, a phantom experiment and experiment on human finger joint, the multi-DASE algorithm is compared with other beamforming methods in removing artifacts by evaluating the quality of the reconstructed images. Furthermore, by rearranging the calculation sequences, the multi-DASE algorithm can be computing in parallel using GPU acceleration to meet the needs of real-time clinical application.

Modeling the occlusion problem in thermal imaging to allow seeing through mist and foliage.

Thermal imaging can easily see through smoke and dust. It is a useful technique in the military and industrial fields. However, thermal imaging can also be blocked by heavy mist or gases with high emissivity such as CO2. Allowing a thermal camera to see through these obstacles is in high demand. In this paper, we modeled the occlusion problem in thermal imaging and proposed an algorithm to image the objects through mist and foliage. We built a system to capture the thermal light field camera. We took thermal reflection and absorption of the obstacles into consideration. We removed the obstacle part in thermal images by estimating the intensity of infrared radiation. Then, we refocused the thermal images on the specific depth of the object for reconstruction. The experiment's results show that a proposed algorithm can reconstruct the occluded objects in a clear shape while blurring the obstacles. Based on the thermal occlusion model and refocusing, the thermal camera can image a human through mist and foliage.