Deep random forest with ferroelectric analog content addressable memory.

Deep random forest (DRF), which combines deep learning and random forest, exhibits comparable accuracy, interpretability, low memory and computational overhead to deep neural networks (DNNs) in edge intelligence tasks. However, efficient DRF accelerator is lagging behind its DNN counterparts. The key to DRF acceleration lies in realizing the branch-split operation at decision nodes. In this work, we propose implementing DRF through associative searches realized with ferroelectric analog content addressable memory (ACAM). Utilizing only two ferroelectric field effect transistors (FeFETs), the ultra-compact ACAM cell performs energy-efficient branch-split operations by storing decision boundaries as analog polarization states in FeFETs. The DRF accelerator architecture and its model mapping to ACAM arrays are presented. The functionality, characteristics, and scalability of the FeFET ACAM DRF and its robustness against FeFET device non-idealities are validated in experiments and simulations. Evaluations show that the FeFET ACAM DRF accelerator achieves ∼106×/10× and ∼106×/2.5× improvements in energy and latency, respectively, compared to other DRF hardware implementations on state-of-the-art CPU/ReRAM.

Achieve fairness without demographics for dermatological disease diagnosis.

In medical image diagnosis, fairness has become increasingly crucial. Without bias mitigation, deploying unfair AI would harm the interests of the underprivileged population and potentially tear society apart. Recent research addresses prediction biases in deep learning models concerning demographic groups (e.g., gender, age, and race) by utilizing demographic (sensitive attribute) information during training. However, many sensitive attributes naturally exist in dermatological disease images. If the trained model only targets fairness for a specific attribute, it remains unfair for other attributes. Moreover, training a model that can accommodate multiple sensitive attributes is impractical due to privacy concerns. To overcome this, we propose a method enabling fair predictions for sensitive attributes during the testing phase without using such information during training. Inspired by prior work highlighting the impact of feature entanglement on fairness, we enhance the model features by capturing the features related to the sensitive and target attributes and regularizing the feature entanglement between corresponding classes. This ensures that the model can only classify based on the features related to the target attribute without relying on features associated with sensitive attributes, thereby improving fairness and accuracy. Additionally, we use disease masks from the Segment Anything Model (SAM) to enhance the quality of the learned feature. Experimental results demonstrate that the proposed method can improve fairness in classification compared to state-of-the-art methods in two dermatological disease datasets.

Description and reconstruction of typical structured light beams with vector spherical wave functions.

It is well known that the generalized Lorenz-Mie theory (GLMT) is a rigorous analytical method for dealing with the interaction between light beams and spherical particles, which involves the description and reconstruction of the light beams with vector spherical wave functions (VSWFs). In this paper, a detailed study on the description and reconstruction of the typical structured light beams with VSWFs is reported. We first systematically derive the so-called beam shape coefficients (BSCs) of typical structured light beams, including the fundamental Gaussian beam, Hermite-Gaussian beam, Laguerre-Gaussian beam, Bessel beam, and Airy beam, with the aid of the angular spectrum decomposition method. Then based on the derived BSCs, we reconstruct these structured light beams using VSWFs and compare the results of the reconstructed beams with those of the original beams. Our results will be useful in the study of the interaction of typical structured light beams with spherical particles in the framework of GLMT.

Ferroelectric compute-in-memory annealer for combinatorial optimization problems.

Computationally hard combinatorial optimization problems (COPs) are ubiquitous in many applications. Various digital annealers, dynamical Ising machines, and quantum/photonic systems have been developed for solving COPs, but they still suffer from the memory access issue, scalability, restricted applicability to certain types of COPs, and VLSI-incompatibility, respectively. Here we report a ferroelectric field effect transistor (FeFET) based compute-in-memory (CiM) annealer for solving larger-scale COPs efficiently. Our CiM annealer converts COPs into quadratic unconstrained binary optimization (QUBO) formulations, and uniquely accelerates in-situ the core vector-matrix-vector (VMV) multiplication operations of QUBO formulations in a single step. Specifically, the three-terminal FeFET structure allows for lossless compression of the stored QUBO matrix, achieving a remarkably 75% chip size saving when solving Max-Cut problems. A multi-epoch simulated annealing (MESA) algorithm is proposed for efficient annealing, achieving up to 27% better solution and ~ 2X speedup than conventional simulated annealing. Experimental validation is performed using the first integrated FeFET chip on 28nm HKMG CMOS technology, indicating great promise of FeFET CiM array in solving general COPs.

Predicting 5-year recurrence risk in colorectal cancer: development and validation of a histology-based deep learning approach.

BACKGROUND: Accurate estimation of the long-term risk of recurrence in patients with non-metastatic colorectal cancer (CRC) is crucial for clinical management. Histology-based deep learning is expected to provide more abundant information for risk stratification. METHODS: We developed and validated a weakly supervised deep-learning model for predicting 5-year relapse-free survival (RFS) to stratify patients with different risks based on histological images from three hospitals of 614 cases with non-metastatic CRC. A deep prognostic factor (DL-RRS) was established to stratify patients into high and low-risk group. The areas under the curve (AUCs) were calculated to evaluate the performances of models. RESULTS: Our proposed model achieves the AUCs of 0.833 (95% CI: 0.736-0.905) and 0.715 (95% CI: 0.647-0.776) on validation cohort and external test cohort, respectively. The 5-year RFS rate was 45.7% for high DL-RRS patients, and 82.5% for low DL-RRS patients respectively in the external test cohort (HR: 3.89, 95% CI: 2.51-6.03, P < 0.001). Adjuvant chemotherapy was associated with improved RFS in Stage II patients with high DL-RRS (HR: 0.15, 95% CI: 0.06-0.38, P < 0.001). CONCLUSIONS: DL-RRS has a good predictive performance of 5-year recurrence risk in CRC, and will better serve the clinical decision-making.

A clinically applicable AI system for diagnosis of congenital heart diseases based on computed tomography images.

Congenital heart disease (CHD) is the most common type of birth defect. Without timely detection and treatment, approximately one-third of children with CHD would die in the infant period. However, due to the complicated heart structures, early diagnosis of CHD and its types is quite challenging, even for experienced radiologists. Here, we present an artificial intelligence (AI) system that achieves a comparable performance of human experts in the critical task of classifying 17 categories of CHD types. We collected the first-large CT dataset from three different CT machines, including more than 3750 CHD patients over 14 years. Experimental results demonstrate that it can achieve diagnosis accuracy (86.03%) comparable with junior cardiovascular radiologists (86.27%) in a World Health Organization appointed research and cooperation center in China on most types of CHD, and obtains a higher sensitivity (82.91%) than junior cardiovascular radiologists (76.18%). The accuracy of the combination of our AI system (97.20%) and senior radiologists achieves comparable results to that of junior radiologists and senior radiologists (97.16%) which is the current clinical routine. Our AI system can further provide 3D visualization of hearts to senior radiologists for interpretation and flexible review, surgeons for precise intuition of heart structures, and clinicians for more precise outcome prediction. We demonstrate the potential of our model to be integrated into current clinic practice to improve the diagnosis of CHD globally, especially in regions where experienced radiologists can be scarce.

Reflection and refraction of higher-order Hermite-Gaussian beams: a vector wave analysis.

Hermite-Gaussian beams, as a typical kind of higher-order mode laser beams, have attracted intensive attention because of their interesting properties and potential applications. In this paper, a full vector wave analysis of the higher-order Hermite-Gaussian beams upon reflection and refraction is reported. The explicit analytical expressions for the electric and magnetic field components of the reflected and refracted Hermite-Gaussian beams are derived with the aid of angular spectrum representation and vector potential in the Lorenz gauge. Based on the derived analytical expressions, local field distributions of higher-order Hermite-Gaussian beams reflection and refraction at a plane interface between air and BK7 glass are displayed and analyzed.

ImageCAS: A large-scale dataset and benchmark for coronary artery segmentation based on computed tomography angiography images.

Cardiovascular disease (CVD) accounts for about half of non-communicable diseases. Vessel stenosis in the coronary artery is considered to be the major risk of CVD. Computed tomography angiography (CTA) is one of the widely used noninvasive imaging modalities in coronary artery diagnosis due to its superior image resolution. Clinically, segmentation of coronary arteries is essential for the diagnosis and quantification of coronary artery disease. Recently, a variety of works have been proposed to address this problem. However, on one hand, most works rely on in-house datasets, and only a few works published their datasets to the public which only contain tens of images. On the other hand, their source code have not been published, and most follow-up works have not made comparison with existing works, which makes it difficult to judge the effectiveness of the methods and hinders the further exploration of this challenging yet critical problem in the community. In this paper, we propose a large-scale dataset for coronary artery segmentation on CTA images. In addition, we have implemented a benchmark in which we have tried our best to implement several typical existing methods. Furthermore, we propose a strong baseline method which combines multi-scale patch fusion and two-stage processing to extract the details of vessels. Comprehensive experiments show that the proposed method achieves better performance than existing works on the proposed large-scale dataset. The benchmark and the dataset are published at https://github.com/XiaoweiXu/ImageCAS-A-Large-Scale-Dataset-and-Benchmark-for-Coronary-Artery-Segmentation-based-on-CT.

A novel risk stratification model for STEMI after primary PCI: global longitudinal strain and deep neural network assisted myocardial contrast echocardiography quantitative analysis.

Background: In ST-segment elevation myocardial infarction (STEMI) with the restoration of TIMI 3 flow by percutaneous coronary intervention (PCI), visually defined microvascular obstruction (MVO) was shown to be the predictor of poor prognosis, but not an ideal risk stratification method. We intend to introduce deep neural network (DNN) assisted myocardial contrast echocardiography (MCE) quantitative analysis and propose a better risk stratification model. Methods: 194 STEMI patients with successful primary PCI with at least 6 months follow-up were included. MCE was performed within 48 h after PCI. The major adverse cardiovascular events (MACE) were defined as cardiac death, congestive heart failure, reinfarction, stroke, and recurrent angina. The perfusion parameters were derived from a DNN-based myocardial segmentation framework. Three patterns of visual microvascular perfusion (MVP) qualitative analysis: normal, delay, and MVO. Clinical markers and imaging features, including global longitudinal strain (GLS) were analyzed. A calculator for risk was constructed and validated with bootstrap resampling. Results: The time-cost for processing 7,403 MCE frames is 773 s. The correlation coefficients of microvascular blood flow (MBF) were 0.99 to 0.97 for intra-observer and inter-observer variability. 38 patients met MACE in 6-month follow-up. We proposed A risk prediction model based on MBF [HR: 0.93 (0.91-0.95)] in culprit lesion areas and GLS [HR: 0.80 (0.73-0.88)]. At the best risk threshold of 40%, the AUC was 0.95 (sensitivity: 0.84, specificity: 0.94), better than visual MVP method (AUC: 0.70, Sensitivity: 0.89, Specificity: 0.40, IDI: -0.49). The Kaplan-Meier curves showed that the proposed risk prediction model allowed for better risk stratification. Conclusion: The MBF + GLS model allowed more accurate risk stratification of STEMI after PCI than visual qualitative analysis. The DNN-assisted MCE quantitative analysis is an objective, efficient and reproducible method to evaluate microvascular perfusion.

Graph matching and deep neural networks based whole heart and great vessel segmentation in congenital heart disease.

Congenital heart disease (CHD) is one of the leading causes of mortality among birth defects, and due to significant variations in the whole heart and great vessel, automatic CHD segmentation using CT images has been always under-researched. Even though some segmentation algorithms have been developed in the literature, none perform very well under the complex structure of CHD. To deal with the challenges, we take advantage of deep learning in processing regular structures and graph algorithms in dealing with large variations and propose a framework combining both the whole heart and great vessel segmentation in complex CHD. We benefit from deep learning in segmenting the four chambers and myocardium based on the blood pool, and then we extract the connection information and apply graph matching to determine the categories of all the vessels. Experimental results on 68 3D CT images covering 14 types of CHD illustrate our framework can increase the Dice score by 12% on average compared with the state-of-the-art whole heart and great vessel segmentation method in normal anatomy. We further introduce two cardiovascular imaging specialists to evaluate our results in the standard of the Van Praagh classification system, and achieves well performance in clinical evaluation. All these results may pave the way for the clinical use of our method in the incoming future.

Partial Unbalanced Feature Transport for Cross-Modality Cardiac Image Segmentation.

Deep learning based approaches have achieved great success on the automatic cardiac image segmentation task. However, the achieved segmentation performance remains limited due to the significant difference across image domains, which is referred to as domain shift. Unsupervised domain adaptation (UDA), as a promising method to mitigate this effect, trains a model to reduce the domain discrepancy between the source (with labels) and the target (without labels) domains in a common latent feature space. In this work, we propose a novel framework, named Partial Unbalanced Feature Transport (PUFT), for cross-modality cardiac image segmentation. Our model facilities UDA leveraging two Continuous Normalizing Flow-based Variational Auto-Encoders (CNF-VAE) and a Partial Unbalanced Optimal Transport (PUOT) strategy. Instead of directly using VAE for UDA in previous works where the latent features from both domains are approximated by a parameterized variational form, we introduce continuous normalizing flows (CNF) into the extended VAE to estimate the probabilistic posterior and alleviate the inference bias. To remove the remaining domain shift, PUOT exploits the label information in the source domain to constrain the OT plan and extracts structural information of both domains, which are often neglected in classical OT for UDA. We evaluate our proposed model on two cardiac datasets and an abdominal dataset. The experimental results demonstrate that PUFT achieves superior performance compared with state-of-the-art segmentation methods for most structural segmentation.

AIGAN: Attention-encoding Integrated Generative Adversarial Network for the reconstruction of low-dose CT and low-dose PET images.

X-ray computed tomography (CT) and positron emission tomography (PET) are two of the most commonly used medical imaging technologies for the evaluation of many diseases. Full-dose imaging for CT and PET ensures the image quality but usually raises concerns about the potential health risks of radiation exposure. The contradiction between reducing the radiation exposure and remaining diagnostic performance can be addressed effectively by reconstructing the low-dose CT (L-CT) and low-dose PET (L-PET) images to the same high-quality ones as full-dose (F-CT and F-PET). In this paper, we propose an Attention-encoding Integrated Generative Adversarial Network (AIGAN) to achieve efficient and universal full-dose reconstruction for L-CT and L-PET images. AIGAN consists of three modules: the cascade generator, the dual-scale discriminator and the multi-scale spatial fusion module (MSFM). A sequence of consecutive L-CT (L-PET) slices is first fed into the cascade generator that integrates with a generation-encoding-generation pipeline. The generator plays the zero-sum game with the dual-scale discriminator for two stages: the coarse and fine stages. In both stages, the generator generates the estimated F-CT (F-PET) images as like the original F-CT (F-PET) images as possible. After the fine stage, the estimated fine full-dose images are then fed into the MSFM, which fully explores the inter- and intra-slice structural information, to output the final generated full-dose images. Experimental results show that the proposed AIGAN achieves the state-of-the-art performances on commonly used metrics and satisfies the reconstruction needs for clinical standards.

Muffin: A Framework Toward Multi-Dimension AI Fairness by Uniting Off-the-Shelf Models.

Model fairness (a.k.a., bias) has become one of the most critical problems in a wide range of AI applications. An unfair model in autonomous driving may cause a traffic accident if corner cases (e.g., extreme weather) cannot be fairly regarded; or it will incur healthcare disparities if the AI model misdiagnoses a certain group of people (e.g., brown and black skin). In recent years, there are emerging research works on addressing unfairness, and they mainly focus on a single unfair attribute, like skin tone; however, real-world data commonly have multiple attributes, among which unfairness can exist in more than one attribute, called "multi-dimensional fairness". In this paper, we first reveal a strong correlation between the different unfair attributes, i.e., optimizing fairness on one attribute will lead to the collapse of others. Then, we propose a novel Multi-Dimension Fairness framework, namely Muffin, which includes an automatic tool to unite off-the-shelf models to improve the fairness on multiple attributes simultaneously. Case studies on dermatology datasets with two unfair attributes show that the existing approach can achieve 21.05% fairness improvement on the first attribute while it makes the second attribute unfair by 1.85%. On the other hand, the proposed Muffin can unite multiple models to achieve simultaneously 26.32% and 20.37% fairness improvement on both attributes; meanwhile, it obtains 5.58% accuracy gain.

Clinical Efficacy of Topical Tacrolimus on Conjunctival Hyperemia Caused by Prostaglandin Analogues.

Objective: To evaluate the clinical efficacy of tacrolimus ophthalmic solution on conjunctival hyperemia caused by prostaglandin analogues. Methods: A retrospective analysis was performed on 120 patients diagnosed with bilateral primary open-angle glaucoma (POAG). The enrolled patients developed symptoms of conjunctival hyperemia during the administration of travoprost ophthalmic solution. The patients were divided into two groups: 0.004% travoprost solution was administered in the control group. A combination of 0.004% travoprost solution with tacrolimus was administered in the experimental group. Clinopathological parameters including intraocular pressure (IOP), subjective dry eye symptom score (SDES), hyperemia score, and noninvasive tear break-up time (NIBUT) were recorded at week 0, 1, 2, and 4. Two-way ANOVA for repeated measurement was employed for statistical analysis using SPSS 22.0 software. Results: At week 1, 2, and 4, the IOP and SDES of both the control and experimental groups were significantly lower when compared the values at week 0 (before treatment). No significant differences in the IOP values between the two groups were observed at all time points. At week 2, the SDES and hyperemia score were lower in the experimental group than those in the control group, and the NIBUT was significantly longer in the experimental group. The above parameters showed no significant difference at week 4 between the two groups, although the average SDES and hyperemia score were slightly lower in the experimental group. Conclusion: Tacrolimus ophthalmic solution can relieve conjunctival hyperemia, improve ocular surface conditions, and reduce discomfort caused by prostaglandin analogues.

Designing Highly Efficient Cu2O-CuO Heterojunction CO Oxidation Catalysts: The Roles of the Support Type and Cu2O-CuO Interface Effect.

In this work, a series of Cu2O/S (S = α-MnO2, CeO2, ZSM-5, and Fe2O3) supported catalysts with a Cu2O loading amount of 15% were prepared by the facile liquid-phase reduction deposition-precipitation strategy and investigated as CO oxidation catalysts. It was found that the Cu2O/α-MnO2 catalyst exhibits the best catalytic activity for CO oxidation. Additionally, a series of Cu2O-CuO/α-MnO2 heterojunctions with varied proportion of Cu+/Cu2+ were synthesized by further calcining the pristine Cu2O/α-MnO2 catalyst. The ratio of the Cu+/Cu2+ could be facilely regulated by controlling the calcination temperature. It is worth noting that the Cu2O-CuO/α-MnO2-260 catalyst displays the best catalytic performance. Moreover, the kinetic studies manifest that the apparent activation energy could be greatly reduced owing to the excellent redox property and the Cu2O-CuO interface effect. Therefore, the Cu2O-CuO heterojunction catalysts supported on α-MnO2 nanotubes are believed to be the potential catalyst candidates for CO oxidation with advanced performance.

Distributed contrastive learning for medical image segmentation.

Supervised deep learning needs a large amount of labeled data to achieve high performance. However, in medical imaging analysis, each site may only have a limited amount of data and labels, which makes learning ineffective. Federated learning (FL) can learn a shared model from decentralized data. But traditional FL requires fully-labeled data for training, which is very expensive to obtain. Self-supervised contrastive learning (CL) can learn from unlabeled data for pre-training, followed by fine-tuning with limited annotations. However, when adopting CL in FL, the limited data diversity on each site makes federated contrastive learning (FCL) ineffective. In this work, we propose two federated self-supervised learning frameworks for volumetric medical image segmentation with limited annotations. The first one features high accuracy and fits high-performance servers with high-speed connections. The second one features lower communication costs, suitable for mobile devices. In the first framework, features are exchanged during FCL to provide diverse contrastive data to each site for effective local CL while keeping raw data private. Global structural matching aligns local and remote features for a unified feature space among different sites. In the second framework, to reduce the communication cost for feature exchanging, we propose an optimized method FCLOpt that does not rely on negative samples. To reduce the communications of model download, we propose the predictive target network update (PTNU) that predicts the parameters of the target network. Based on PTNU, we propose the distance prediction (DP) to remove most of the uploads of the target network. Experiments on a cardiac MRI dataset show the proposed two frameworks substantially improve the segmentation and generalization performance compared with state-of-the-art techniques.

Enhancing the Low-Temperature CO Oxidation over CuO-Based α-MnO2 Nanowire Catalysts.

A series of CuO-based catalysts supported on the α-MnO2 nanowire were facilely synthesized and employed as the CO oxidation catalysts. The achieved catalysts were systematically characterized by XRD, SEM, EDS-mapping, XPS and H2-TPR. The catalytic performances toward CO oxidation had been carefully evaluated over these CuO-based catalysts. The effects of different loading methods, calcination temperatures and CuO loading on the low temperature catalytic activity of the catalyst were investigated and compared with the traditional commercial MnO2 catalyst with a block structure. It was found that the slenderness ratio of a CuO/α-MnO2 nanowire catalyst decreases with the increase in CuO loading capacity. The results showed that when CuO loading was 3 wt%, calcination temperature was 200 °C and the catalyst that was supported by the deposition precipitation method had the highest catalytic activity. Besides, the α-MnO2 nanowire-supported catalysts with excellent redox properties displayed much better catalytic performances than the commercial MnO2-supported catalyst. In conclusion, the CuO-based catalysts that are supported by α-MnO2 nanowires are considered as a series of promising CO oxidation catalysts.

Design and Preparation of Imidazole Ionic Liquid-Based Magnetic Polymers and Its Adsorption on Sunset Yellow Dye.

Magnetic polymers are often used as loading materials for ionic liquids because of their excellent magnetic separation properties. In this study, a novel imidazolium-based ionic liquid-modified magnetic polymer was synthesized by suspension polymerization and grafting, denoted as γ-Fe2O3@GMA@IM, and this magnetic polymer was used for the adsorption of the acid dye FCF. The magnetic polymer was characterized by SEM, FTIR, XRD, VSM and TGA. These techniques were used to reveal the overall physical properties of magnetic polymers, including the presence of morphology, functional groups, crystalline properties, magnetism and thermal stability. Studies have shown that γ-Fe2O3@GMA@IM can adsorb FCF in a wide pH range (2-10), with a maximum adsorption capacity of 445 mg/g. The adsorption data were more in line with the pseudo-second-order kinetic model and the Freundlich isotherm. In order to investigate its reusability, this study used 10% NaCl as the desorption solution, and carried out five batches of adsorption-desorption cycles. After five cycles, the adsorption effect was maintained at 98.3%, which showed a good recycling performance.

Myocardial Segmentation of Cardiac MRI Sequences With Temporal Consistency for Coronary Artery Disease Diagnosis.

Coronary artery disease (CAD) is the most common cause of death globally, and its diagnosis is usually based on manual myocardial (MYO) segmentation of MRI sequences. As manual segmentation is tedious, time-consuming, and with low replicability, automatic MYO segmentation using machine learning techniques has been widely explored recently. However, almost all the existing methods treat the input MRI sequences independently, which fails to capture the temporal information between sequences, e.g., the shape and location information of the myocardium in sequences along time. In this article, we propose a MYO segmentation framework for sequence of cardiac MRI (CMR) scanning images of the left ventricular (LV) cavity, right ventricular (RV) cavity, and myocardium. Specifically, we propose to combine conventional neural networks and recurrent neural networks to incorporate temporal information between sequences to ensure temporal consistency. We evaluated our framework on the automated cardiac diagnosis challenge (ACDC) dataset. The experiment results demonstrate that our framework can improve the segmentation accuracy by up to 2% in the Dice coefficient.

DeU-Net 2.0: Enhanced deformable U-Net for 3D cardiac cine MRI segmentation.

Automatic segmentation of cardiac magnetic resonance imaging (MRI) facilitates efficient and accurate volume measurement in clinical applications. However, due to anisotropic resolution, ambiguous borders and complicated shapes, existing methods suffer from the degradation of accuracy and robustness in cardiac MRI segmentation. In this paper, we propose an enhanced Deformable U-Net (DeU-Net) for 3D cardiac cine MRI segmentation, composed of three modules, namely Temporal Deformable Aggregation Module (TDAM), Enhanced Deformable Attention Network (EDAN), and Probabilistic Noise Correction Module (PNCM). TDAM first takes consecutive cardiac MR slices (including a target slice and its neighboring reference slices) as input, and extracts spatio-temporal information by an offset prediction network to generate fused features of the target slice. Then the fused features are also fed into EDAN that exploits several flexible deformable convolutional layers and generates clear borders of every segmentation map. A Multi-Scale Attention Module (MSAM) in EDAN is proposed to capture long range dependencies between features of different scales. Meanwhile, PNCM treats the fused features as a distribution to quantify uncertainty. Experimental results show that our DeU-Net achieves the state-of-the-art performance in terms of the commonly used evaluation metrics on the Extended ACDC dataset and competitive performance on other two datasets, validating the robustness and generalization of DeU-Net.