Automatic Delineation of the 3D Left Atrium From LGE-MRI: Actor-Critic Based Detection and Semi-Supervised Segmentation.

Accurate and automatic delineation of the left atrium (LA) is crucial for computer-aided diagnosis of atrial fibrillation-related diseases. However, effective model training typically requires a large amount of labeled data, which is time-consuming and labor-intensive. In this study, we propose a novel LA delineation framework. The region of LA is first detected using an actor-critic based deep reinforcement learning method with a shape-adaptive detection strategy using only box-level annotations, bypassing the need for voxel-level labeling. With the effectively detected LA, the impacts of class-imbalance and interference from surrounding tissues are significantly reduced. Subsequently, a semi-supervised segmentation scheme is coined to precisely delineate the contour of LA in 3D volume. The scheme integrates two independent networks with distinct structures, enabling implicit consistency regularization, capturing more spatial features, and avoiding the error accumulation present in current mainstream semi-supervised frameworks. Specifically, one network is combined with Transformer to capture latent spatial features, while the other network is based on pure CNN to capture local features. The difference prediction between these two sub-networks is exploited to mutually provide high-quality pseudo-labels and correct the cognitive bias. Experimental results on two public datasets demonstrate that our proposed strategy outperforms several state-of-the-art methods in terms of accuracy and clinical convenience.

Dehalogenative Arylation of Unactivated Alkyl Halides via Electroreduction.

Herein, a facile and efficient dehalogenative arylation of unactivated alkyl halides enabled by electrochemical reductive coupling is developed, affording a series of C(sp2)-C(sp3) products in moderate to good yields. This protocol proceeds in the absence of transition metal catalysts and redox mediators. The reaction features mild conditions, broad substrate scope, and high tolerance of functional groups and is demonstrated to be applicable for gram-scale synthesis and late-stage functionalization of natural products.

Improvements of the Tada formula in estimating the intracerebral hemorrhage volume based on computed tomography.

Background: The Tada formula has been used widely for assessing intracerebral hemorrhage (ICH) volume. However, it is only suitable for calculating regular and small volume hematomas. Therefore, we attempted to improve the formula to increase its accuracy and maintain its efficiency. Methods: Computed tomography (CT) data of 15 balls of different shapes filled with predetermined volumes of water were collected to verify the high accuracy of FireVoxel in measuring the volume. CT data from 329 patients with ICH from two different hospitals grouped by hematoma shape and volume were retrospectively reviewed. The distinctly shaped ICH volumes of 245 patients from one of the hospitals were estimated using FireVoxel and the Tada formula grouped by the hematoma shape and volume. Taking the hematoma volumes measured by FireVoxel as the reference standard, the accuracy and reliability of the Tada formula were evaluated. Polynomial fitting was employed to determine the associations of the values calculated between the Tada formula and FireVoxel. Then, a corrected Tada formula (C-Tada formula) was produced, and the limits of agreement between the C-Tada formula and Tada formula were analyzed with Bland-Altman analysis. The C-Tada formula was validated by the CT data of 84 patients from another hospital. Results: The volume measured by FireVoxel can be set as the reference standard. The ICH volume calculated by the Tada formula was significantly greater than that calculated by FireVoxel for different shapes and volumes. The percentage deviation between the volumes calculated by FireVoxel and the Tada formula was also statistically significant and influenced by ICH shape and volume. The limits of agreement between the C-Tada formula and FireVoxel were tighter than those between the Tada formula and FireVoxel. The percentage deviation of the C-Tada formula calculation from the FireVoxel estimate was greatly reduced relative to that for the Tada formula for each group. Conclusions: The C-Tada formula is more clinically valuable than the Tada formula, given its sufficient efficiency and greater accuracy and reliability in ICH volume calculation.

Segmentation of biventricle in cardiac cine MRI via nested capsule dense network.

Background: Cardiac magnetic resonance image (MRI) has been widely used in diagnosis of cardiovascular diseases because of its noninvasive nature and high image quality. The evaluation standard of physiological indexes in cardiac diagnosis is essentially the accuracy of segmentation of left ventricle (LV) and right ventricle (RV) in cardiac MRI. The traditional symmetric single codec network structure such as U-Net tends to expand the number of channels to make up for lost information that results in the network looking cumbersome. Methods: Instead of a single codec, we propose a multiple codecs structure based on the FC-DenseNet (FCD) model and capsule convolution-capsule deconvolution, named Nested Capsule Dense Network (NCDN). NCDN uses multiple codecs to achieve multi-resolution, which makes it possible to save more spatial information and improve the robustness of the model. Results: The proposed model is tested on three datasets that include the York University Cardiac MRI dataset, Automated Cardiac Diagnosis Challenge (ACDC-2017), and the local dataset. The results show that the proposed NCDN outperforms most methods. In particular, we achieved nearly the most advanced accuracy performance in the ACDC-2017 segmentation challenge. This means that our method is a reliable segmentation method, which is conducive to the application of deep learning-based segmentation methods in the field of medical image segmentation.

Image Segmentation Using Active Contours with Hessian-Based Gradient Vector Flow External Force.

The gradient vector flow (GVF) model has been widely used in the field of computer image segmentation. In order to achieve better results in image processing, there are many research papers based on the GVF model. However, few models include image structure. In this paper, the smoothness constraint formula of the GVF model is re-expressed in matrix form, and the image knot represented by the Hessian matrix is included in the GVF model. Through the processing of this process, the relevant diffusion partial differential equation has anisotropy. The GVF model based on the Hessian matrix (HBGVF) has many advantages over other relevant GVF methods, such as accurate convergence to various concave surfaces, excellent weak edge retention ability, and so on. The following will prove the advantages of our proposed model through theoretical analysis and various comparative experiments.

Sequential shape similarity for active contour based left ventricle segmentation in cardiac cine MR image.

Delineation of the boundaries of the Left Ventricle (LV) in cardiac Magnetic Resonance Images (MRI) is a hot topic due to its important diagnostic power. In this paper, an approach is proposed to extract the LV in a sequence of MR images. In the proposed paper, all images in the sequence are segmented simultaneously and the shape of the LV in each image is supposed to be similar to that of the LV in nearby images in the sequence. We coined the novel shape similarity constraint, and it is called sequential shape similarity (SSS in short). The proposed segmentation method takes the Active Contour Model as the base model and our previously proposed Gradient Vector Convolution (GVC) external force is also adopted. With the SSS constraint, the snake contour can accurately delineate the LV boundaries. We evaluate our method on two cardiac MRI datasets and the Mean Absolute Distance (MAD) metric and the Hausdorff Distance (HD) metric demonstrate that the proposed approach has good performance on segmenting the boundaries of the LV.

Image segmentation using active contours with modified convolutional virtual electric field external force with an edge-stopping function.

The gradient vector flow (GVF) is an effective external force to deform the active contours. However, it suffers from high computational cost. For efficiency, the virtual electric field (VEF) model has been proposed, which can be implemented in real time thanks to fast Fourier transform (FFT). The VEF model has large capture range and low computation cost, but it has the limitations of sensitivity to noise and leakage on weak edge. The recently proposed CONVEF (Convolutional Virtual Electric Field) model takes the VEF model as a convolutional operation and employed another convolution kernel to overcome the drawbacks of the VEF model. In this paper, we employ an edge stopping function similar to that in the anisotropic diffusion to further improve the CONVEF model, and the proposed model is referred to as MCONVEF (Modified CONVEF) model. In addition, a piecewise constant approximation algorithm is borrowed to accelerate the calculation of the MCONVEF model. The proposed MCONVEF model is compared with the GVF and VEF models, and the experimental results are presented to demonstrate its superiority in terms of noise robustness, weak edge preserving and deep concavity convergence.

Photonic RF vector signal generation with enhanced spectral efficiency using precoded double single-sideband modulation.

In this study, a novel photonic vector signal at frequency (RF) bands generation scheme based on the beating of double single sidebands (SSBs) is proposed and experimentally demonstrated. The double SSBs carry separate constant- or multi-amplitude quadrature-amplitude-modulation vector signals are generated from a single I/Q modulator. By adopting phase and amplitude precoding, different constellations can be generated, such as 3-ary phase-shift keying (PSK), 4-PSK, 7-PSK, 8-PSK, and so on. In this work, 10-Gbaud 7-PSK vector signal generation at 20 GHz enabled by two precoded 4-PSK SSB signals via a single I/Q modulator is theoretically and experimentally investigated. Compared to a single-drive Mach-Zehnder modulator or conventional I/Q modulator-based photonic vector signal generation scheme, the spectrum efficiency can be doubled. Differential coding is also implemented at the transmitter side for accurate demodulation of 7-PSK into two 4-PSK signals. The bit-error ratio for 10-Gbaud 7-PSK vector signals can be under hard-decision forward-error-correction threshold of 3.8×10-3 after 10 km standard single-mode fiber transmission.

Transmission of single-carrier 400G signals (515.2-Gb/s) based on 128.8-GBaud PDM QPSK over 10,130- and 6,078 km terrestrial fiber links.

We experimentally demonstrate the coherent transmission system with the highest ETDM-based symbol rate of 128.8-GBaud over record breaking distances. We successfully transmitted single-carrier 515.2-Gb/s PDM-QPSK/9-QAM signals over 10,130km/6,078-km, respectively, over 100km spans of TeraWave SLA + fiber. To the best of our knowledge, it is the highest ETDM-based symbol rate reported so far, and the longest WDM transmission distance with single-carrier 400G signals. For the first time, the 515.2-Gb/s single-carrier PDM-QPSK signals in 200-GHz-grid are successfully transmitted over distance above 10,000km in terrestrial transmission environment. We have also demonstrated the transmission of single carrier 128.8-GBaud filtered QPSK signals in 100-GHz-grid over 6,078-km, which has the line spectral efficiency (SE) of 5.152 (b/s/Hz).

Convolutional virtual electric field for image segmentation using active contours.

Gradient vector flow (GVF) is an effective external force for active contours; however, it suffers from heavy computation load. The virtual electric field (VEF) model, which can be implemented in real time using fast Fourier transform (FFT), has been proposed later as a remedy for the GVF model. In this work, we present an extension of the VEF model, which is referred to as CONvolutional Virtual Electric Field, CONVEF for short. This proposed CONVEF model takes the VEF model as a convolution operation and employs a modified distance in the convolution kernel. The CONVEF model is also closely related to the vector field convolution (VFC) model. Compared with the GVF, VEF and VFC models, the CONVEF model possesses not only some desirable properties of these models, such as enlarged capture range, u-shape concavity convergence, subject contour convergence and initialization insensitivity, but also some other interesting properties such as G-shape concavity convergence, neighboring objects separation, and noise suppression and simultaneously weak edge preserving. Meanwhile, the CONVEF model can also be implemented in real-time by using FFT. Experimental results illustrate these advantages of the CONVEF model on both synthetic and natural images.

Enhanced performance of visible light communication employing 512-QAM N-SC-FDE and DD-LMS.

In this paper, a novel hybrid time-frequency adaptive equalization algorithm based on a combination of frequency domain equalization (FDE) and decision-directed least mean square (DD-LMS) is proposed and experimentally demonstrated in a Nyquist single carrier visible light communication (VLC) system. Adopting this scheme, as well with 512-ary quadrature amplitude modulation (512-QAM) and wavelength multiplexing division (WDM), an aggregate data rate of 4.22-Gb/s is successfully achieved employing a single commercially available red-green-blue (RGB) light emitting diode (LED) with low bandwidth. The measured Q-factors for 3 wavelength channels are all above the Q-limit. To the best of our knowledge, this is the highest data rate ever achieved by employing a commercially available RGB-LED.

Integrated 10 Gb/s multilevel multiband passive optical network and 500 Mb/s indoor visible light communication system based on Nyquist single carrier frequency domain equalization modulation.

We propose and experimentally demonstrate a novel integrated passive optical network (PON) and indoor visible light communication (VLC) system based on Nyquist single carrier frequency domain equalization (N-SC-FDE) modulation with direct detection. In this system, a directly modulated laser and a commercially available red light emitting diode are served as the transmitters of the PON and VLC, respectively. To enable high spectral efficiency, high-speed transmission, and flexible multiple access with simplified optical network unit-side digital signal processing, multilevel, multiband quadrature amplitude modulations 128/64/16 are implemented here. VLC N-SC-FDE signals are successfully delivered a further 30 cm indoor distance after transmitting over a span of 40 km single mode fiber (SMF) together with 3 sub-band PON signals. As a proof of concept, a 10 Gb/s PON and 500 Mb/s VLC integrated system for three wired users and one wireless user is successfully achieved, which shows the promising potential and feasibility of this proposal to extend multiple services from metropolitan to suburban areas.

Gigabit polarization division multiplexing in visible light communication.

In this Letter, polarization division multiplexing is proposed and experimentally demonstrated for the first time that we know of, in visible light communication systems based on incoherent light emitting diodes and two orthogonal groups of linear polarizers. Spectrally efficient 16-ary quadrature amplitude modulation Nyquist single carrier frequency domain equalization is employed to obtain a maximum spectral efficiency. We achieve an aggregate data rate of 1 Gb/s, with bit error rate results for two polarization directions both below the 7% pre-forward-error-correction threshold of 3.8×10(-3) after 80 cm free-space transmission. Moreover, the cross talk between x and y polarization is also discussed and analyzed.

Demonstration of 575-Mb/s downlink and 225-Mb/s uplink bi-directional SCM-WDM visible light communication using RGB LED and phosphor-based LED.

We propose and experimentally demonstrate a novel full-duplex bi-directional subcarrier multiplexing (SCM)-wavelength division multiplexing (WDM) visible light communication (VLC) system based on commercially available red-green-blue (RGB) light emitting diode (LED) and phosphor-based LED (P-LED) with 575-Mb/s downstream and 225-Mb/s upstream transmission, employing various modulation orders of quadrature amplitude modulation (QAM) orthogonal frequency division multiplexing (OFDM). For the downlink, red and green colors/wavelengths are assigned to carry useful information, while blue chip is just kept lighting to maintain the white color illumination, and for the uplink, the low-cost P-LED is implemented. In this demonstration, pre-equalization and post-equalization are also adopted to compensate the severe frequency response of LEDs. Using this scheme, 4-user downlink and 1-user uplink transmission can be achieved. Furthermore, it can support more users by adjusting the bandwidth of each sub-channel. Bit error rates (BERs) of all links are below pre-forward-error-correction (pre-FEC) threshold of 3.8x 10(-3) after 66-cm free-space delivery. The results show that this scheme has great potential in the practical VLC system.

High-speed quasi-balanced detection OFDM in visible light communication.

In this paper, we proposed and experimentally demonstrated a novel quasi-balanced detection (QBD) technique in orthogonal frequency division multiplexing (OFDM) visible light communication (VLC) system. By employing opposite signals to odd and even consecutive symbols, the nonlinearity distortion, and direct current can be eliminated efficiently. Additionally, the sensitivity of receiver can also be improved by 3dB, thus a longer transmission distance and high-order modulation formats can be received. We achieved physical data rate of 2.1-Gb/s enabled by wavelength-division-multiplexing (WDM), pre- and post-equalization, and the resulting bit error ratios (BERs) were below the FEC limit of 3.8x10(-3). The distance was above 2.5 meters that was long enough for indoor communication. Compared with conventional direct-detection optical (DDO-OFDM) and asymmetrically-clipped optical (ACO-OFDM), the BER can be enhanced by 22.2dB and 20.8dB, respectively, which shows great potential in short range and low cost access network.