Refined Feature-based Multi-frame and Multi-scale Fusing Gate network for accurate segmentation of plaques in ultrasound videos.

The accurate segmentation of carotid plaques in ultrasound videos will provide evidence for clinicians to evaluate the properties of plaques and treat patients effectively. However, the confusing background, blurry boundaries and plaque movement in ultrasound videos make accurate plaque segmentation challenging. To address the above challenges, we propose the Refined Feature-based Multi-frame and Multi-scale Fusing Gate Network (RMFG_Net), which captures spatial and temporal features in consecutive video frames for high-quality segmentation results and no manual annotation of the first frame. A spatial-temporal feature filter is proposed to suppress the noise of low-level CNN features and promote the detailed target area. To obtain a more accurate plaque position, we propose a transformer-based cross-scale spatial location algorithm, which models the relationship between adjacent layers of consecutive video frames to achieve stable positioning. To make full use of more detailed and semantic information, multi-layer gated computing is applied to fuse features of different layers, ensuring sufficient useful feature map aggregation for segmentation. Experiments on two clinical datasets demonstrate that the proposed method outperforms other state-of-the-art methods under different evaluation metrics, and it processes images with a speed of 68 frames per second which is suitable for real-time segmentation. A large number of ablation experiments were conducted to demonstrate the effectiveness of each component and experimental setting, as well as the potential of the proposed method in ultrasound video plaque segmentation tasks. The codes can be publicly available from https://github.com/xifengHuu/RMFG_Net.git.

[Relationship of gross tumor volume with lymph node metastasis and prognosis of esophageal carcinoma].

OBJECTIVE: To explore whether there is a relationship between gross tumor volume (GTV) and pathologic lymph node metastasis or prognosis of esophageal carcinoma, and to provide a new prognosis reference for esophageal carcinoma (EC). METHODS: Six hundred and seven patients received radical resection of thoracic esophageal carcinoma from May 2002 to June 2006 in our hospital, and their pre-operative CT images were transmitted to the three-dimensional conformal radiotherapy planning system by the network in digital format. Esophageal GTV targets were outlined and their GTV volumes were calculated. To analyze whether there is a relationship between GTV volume and pathologic lymph node metastasis or prognosis. RESULTS: In the 607 cases of esophageal carcinoma, the GTV volume was (22.5 ± 16.8) cm(3) in 374 stage N0 EC patients, significantly different from that of (30.4 ± 20.1) cm(3) in 233 stage N1 EC cases (P < 0.001). There is a significant difference between the GTV volumes of the groups with and without lymph node metastasis (P < 0.05). There was a significant difference of the GTV volumes of EC patients with one lymph node metastasis and those with ≥ 4 lymph node metastasis (P < 0.05). There was a positive correlation between GTV volume and the number of lymph node metastasis (r = 0.230, P < 0.001). The 1-, 3-, 5-year survival rates since the surgery date were 83.8%, 53.5%, and 36.4%, respectively. There was a significant difference between the survival rates of stage N0 (48.5%) and stage N1 patients (18.2%, P < 0.001), and there was a significant difference between the survival rats of patients with 0, 1 and ≥ 2 lymph node metastasis (P < 0.01). Cox regression model analysis showed that GTV volume, number of lymph node metastasis, pathological type, and lesion site were independent prognostic factors (all P < 0.05). CONCLUSION: The GTV volume of esophageal carcinoma is positively correlated with the number of pathologic lymph node metastasis, and it is an independent prognostic factor for this cancer.

TRSA-Net: Task Relation Spatial Co-Attention for Joint Segmentation, Quantification and Uncertainty Estimation on Paired 2D Echocardiography.

Clinical workflow of cardiac assessment on 2D echocardiography requires both accurate segmentation and quantification of the Left Ventricle (LV) from paired apical 4-chamber and 2-chamber. Moreover, uncertainty estimation is significant in clinically understanding the performance of a model. However, current research on 2D echocardiography ignores this vital task while joint segmentation with quantification, hence motivating the need for a unified optimization method. In this paper, we propose a multitask model with Task Relation Spatial co-Attention (referred as TRSA-Net) for joint segmentation, quantification, and uncertainty estimation on paired 2D echo. TRSA-Net achieves multitask joint learning by novelly exploring the spatial correlation between tasks. The task relation spatial co-attention learns the spatial mapping among task-specific features by non-local and co-excitation, which forcibly joints embedded spatial information in the segmentation and quantification. The Boundary-aware Structure Consistency (BSC) and Joint Indices Constraint (JIC) are integrated into the multitask learning optimization objective to guide the learning of segmentation and quantification paths. The BSC creatively promotes structural similarity of predictions, and JIC explores the internal relationship between three quantitative indices. We validate the efficacy of our TRSA-Net on the public CAMUS dataset. Extensive comparison and ablation experiments show that our approach can achieve competitive segmentation performance and highly accurate results on quantification.

MVFStain: Multiple virtual functional stain histopathology images generation based on specific domain mapping.

To the best of our knowledge, artificial intelligence stain generation is an urgent requirement for histopathology images. Pathological examinations usually only utilize hematoxylin and eosin (H&E) regular staining to show histomorphological characteristics, but to accurately diagnose the disease, functional staining (such as oil red O and Ki67) are also required to provide important auxiliary information. However, the same tissue section is usually stained with one stain, and additional functional staining is not only time-consuming but also causes inevitable morphological misalignment due to manual manipulation. This brings difficulties to the development of artificial intelligence pathological image analysis tools. In this work, we propose a histopathology staining transfer framework to generate virtual functional staining images from H&E regular staining images. Compared with the framework that emphasizes generation diversity in the natural image field, we use KL loss and histo loss to align and separate style feature spaces in different domains to obtain domain-variant style features. The proposed multiple virtual functional stain (MVFStain) abstracts staining conversion to domain mapping and comprehensively utilizes multiple staining information. We evaluated the proposed method on four datasets (lung lesion, lung lobes, breast, and atherosclerotic lesion). The experiment involves the translation of H&E to nine other functional stains: CC10, Ki67, proSPC, HER2, PR, ER, oil red O, α-SMA, and macrophages. The major quantitative results are divided into image quality and positive signal prediction. MVFStain is close to or even surpasses one-to-one image translation on psnr and HTI image quality metrics. The best psnr reaches 26.1919, and HIT reaches 0.9430. We used mIOD to evaluate the optical density of positive signals, and CNR and gCNR to evaluate the lesion detectability. The results show that the mIOD of positive signals of virtual staining was slightly lower than the ground truth and close the lesion detectability of artificial staining. These results prove that the potential exists to develop a successful clinical alternative to artificial functional stains.

3D Reconstruction Method based on Medical Image Feature Point Matching.

Medical 3D image reconstruction is an important image processing step in medical image analysis. How to speed up the speed while improving the accuracy in 3D reconstruction is an important issue. To solve this problem, this paper proposes a 3D reconstruction method based on image feature point matching. By improving SIFT, the initial matching of feature points is realized by using the neighborhood voting method, and then the initial matching points are optimized by the improved RANSAC algorithm, and a new SFM reconstruction method is obtained. The experimental results show that the feature matching rate of this algorithm on Fountain data is 95.42% and the matching speed is 4.751 s. It can be seen that this algorithm can shorten the reconstruction time and obtain sparse point clouds with more reasonable distribution and better reconstruction effect.

Medical image management and analysis system based on web for fungal keratitis images.

The medical image management and analysis system proposed in this paper is a medical software developed by the Browser/Server (B/S) architecture after investigating the workflow of the relevant departments of the hospital, which realizes the entire process of patients from consultation to printing of reports. The computer-aided diagnosis function is added based on image management. Due to the difficulty in collecting medical image data, in the computer-aided diagnosis module, this paper only uses the common fungal keratitis collected from the hospital in the laboratory. Focused microscope images are used for experiments. First, the images were trained with three convolutional neural networks, AlexNet, ZFNet, and VGG16. These models which classify fungal keratitis were obtained and integrated was performed to obtain better classification results. Finally, the model was integrated with the system designed in this paper, which realized the automatic diagnosis of Confocal Microscopy (CM) images of fungal keratitis online and provided it to medical staff for reference. The system can improve the work efficiency of the image-related departments while reducing the workload of doctors in the department to manually read the films.

[Exploration of the classification of gross tumor volume and pathological staging of esophageal carcinoma].

OBJECTIVE: Using the volume calculating function of treatment planning system of 3DCRT to work out the value of GTV standard classifications and to provide the reference for clinical staging of esophageal carcinoma. METHODS: Six hundred and seven patients underwent radical resection of thoracic esophageal carcinoma in our hospital, and their pre-operative CT images were transmitted in digital format to the three-dimensional conformal radiotherapy planning system by the network. Esophageal lesion GTV targets were outlined, and their volumes were automatically computed by the planning system. Compared the differences of the GTV volumes in different pathological T stages, and analyzed the relationship between GTV volumes and pathological T stages. According to the median volume of GTV at different pathological T stages, divided the values of GTV volume corresponding to different T stages and selected the suitable classification standard of GTV volume. RESULTS: The esophageal carcinoma GTV length, maximum diameter and volume were related to pathological T staging and with a positive correlation (all P < 0.001). The Spearman correlation coefficient (r) was 0.376, 0.466 and 0.464, respectively, P < 0.001. Except that the length, maximum diameter and volume of GTV in pathological T3 and T4 had no significant difference, other indicators of the pathological T stages showed significant differences between the groups (P < 0.001). According to the median volume of GTV at different pathological T stages, the GTV volumes were divided into three grades: 13.0 cm(3). When compared them with pathological T1, T2, and T3-T4 stages, the coincidence rate was 73.8%. The consistency was good between the GTV volume grades and pathological T stages (Kappa = 0.40, P < 0.001). The overall 5-year survival rates of GTV grades 1, 2, 3 were 78.1%, 31.5% and 33.5%, respectively (P < 0.0001). If the GTVs were divided into four grades: 39.0 cm(3), the coincidence rate of GTV volume grades and pathology T staging was only 54.7%, and the consistency was poor, Kappa = 0.24, P < 0.001. The overall 5-year survival rate of GTV 1, 2, 3, 4 were 78.1%, 31.5%, 36.2% and 27.5%, respectively (P < 0.0001). CONCLUSION: The length, maximum diameter and volume of esophageal carcinoma GTV are related to pathological T staging with a positive correlation. The classification that esophageal carcinoma GTVs divided into three grades has a good coincidence with the pathological T staging.

Influence of hydropower stations on the water microbiota in the downstream of Jinsha River, China.

Construction of hydropower stations has been an important approach to meet China's increasing power demand, but the impact of construction of hydropower stations on river microbiota is not fully understood. To evaluate this, the microbial composition from 18 sampling sites in the downstream of Jinsha River of China, upstream and downstream of two completed and two under-construction hydropower stations, were analyzed using high-throughput 16S rRNA gene sequencing. Three independent samples from each site were analyzed. A total of 18,683 OTUs from 1,350 genera were identified at 97% sequence similarity. Our results showed that the completion of hydropower stations would significantly increase the relative abundances of Acidobacteria, Chlorobi, Chloroflexi, Cyanobacteria, Nitrospirae, and Planctomycetes, especially the relative abundance of Synechococcus dOTUs and thus increase the risk of algal blooms. PCA based on all KEGG pathways and the significantly different KEGG pathways showed the predicted metabolic characteristics of the water microbiota by PICRUSt in the activated hydropower station group were significant difference to the other groups. Results from canonical correspondence analysis showed that water temperature and dissolved oxygen had significant effects on microbiota composition. These results are important for assessing the impact of hydropower stations on river microbiota and their potential environmental risks.

Detecting vulnerable plaque with vulnerability index based on convolutional neural networks.

Plaque rupture and subsequent thrombosis are major processes of acute cardiovascular events. The Vulnerability Index is a very important indicator of whether a plaque is ruptured, and these easily ruptured or fragile plaques can be detected early. The higher the general vulnerability index, the higher the instability of the plaque. Therefore, determining a clear vulnerability index classification point can effectively reduce unnecessary interventional therapy. However, the current critical value of the vulnerability index has not been well defined. In this study, we proposed a neural network-based method to determine the critical point of vulnerability index that distinguishes vulnerable plaques from stable ones. Firstly, based on MatConvNet, the intravascular ultrasound images under different vulnerability index labels are classified. Different vulnerability indexes can obtain different accuracy rates for the demarcation points. The corresponding data points are fitted to find the existing relationship to judge the highest classification. In this way, the vulnerability index corresponding to the highest classification accuracy rate is judged. Then the article is based on the same experiment of different components of the aortic artery in the artificial neural network, and finally the vulnerability index corresponding to the highest classification accuracy can be obtained. The results show that the best vulnerability index point is 1.716 when the experiment is based on the intravascular ultrasound image, and the best vulnerability index point is 1.607 when the experiment is based on the aortic artery component data. Moreover, the vulnerability index and classification accuracy rate has a periodic relationship within a certain range, and finally the highest AUC is 0.7143 based on the obtained vulnerability index point on the verification set. In this paper, the convolution neural network is used to find the best vulnerability index classification points. The experimental results show that this method has the guiding significance for the classification and diagnosis of vulnerable plaques, further reduce interventional treatment of cardiovascular disease.

Automatic diagnosis of fungal keratitis using data augmentation and image fusion with deep convolutional neural network.

BACKGROUND AND OBJECTIVES: Fungal keratitis is caused by inflammation of the cornea that results from infection by fungal organisms. The lack of an early effective diagnosis often results in serious complications even blindness. Confocal microscopy is one of the most effective methods in the diagnosis of fungal keratitis, but the diagnosis depends on the subjective judgment of medical experts. METHODS: To address this problem, this paper proposes a novel convolutional neural network framework for the automatic diagnosis of fungal keratitis using data augmentation and image fusion. Firstly, a normal image is augmented by flipping to solve the problem of having a limited and imbalanced database. Secondly, a sub-area contrast stretching algorithm is proposed for image preprocessing to highlight the key structures in the images and to filter out irrelevant information. Thirdly, the histogram matching fusion algorithm is implemented, then the preprocessed image is fused with the original image to form a new algorithm framework and a new database. Finally, the traditional convolutional neural network is integrated into the novel algorithm framework to perform the experiments. RESULTS: Experiments show that the accuracy of traditional AlexNet and VGGNet is 99.35% and 99.14%, that of AlexNet and VGGNet based on MF fusion is 99.80% and 99.83%, and that of AlexNet and VGGNet based on histogram matching fusion (HMF) is 99.95% and 99.89%. The experimental results show that the AlexNet framework using data augmentation and image fusion achieves a perfect trade-off between the diagnostic performance and the computational complexity, with a diagnostic accuracy of 99.95%. CONCLUSIONS: These experimental results demonstrate the novel convolutional neural network framework perfectly balances the diagnostic performance and computational complexity, and can improve the effect and real-time performance in the diagnosis of fungal keratitis.

Degradation enhancement of rice straw by co-culture of Phanerochaete chrysosporium and Trichoderma viride.

Straw is one of the most abundant stock of renewable biomass from crop production. However, its utilization efficiency is still very low. Although co-cultivation of fungi increases the degrading rate, the co-cultivation condition needs to be optimized. To optimize the co-culture condition of Phanerochaete chrysosporium and Trichoderma viride degrading rice straw, we first tested the antagonistic characteristic between the fungi. The results showed that the best co-culture pattern was to first inoculate P. chrysosporium and culture for 4 days, then inoculate T. viride, and co-culture the two fungi for 4 days. The optimum fermentation condition was 14% (w/v) of inoculum concentration, the equivalent inoculation of the fungi, culture temperature at 30 °C, and 1:1.4 for solid-liquid ratio. Under the optimum condition, the degradation ratios of lignin and cellulose were 26.38% and 33.29%, respectively; the soluble carbon content in the culture product was 23.07% (w/v). The results would provide important reference information for the efficient utilization of rice straw to produce more accessible energy resources, such as ethanol and glucose.

A SEM study of the antenna and mouthparts of Omosita colon (Linnaeus) (Coleoptera: Nitidulidae).

There are direct relationships between the behavioral mechanisms and sensilla. To obtain a better understanding of the behavioral mechanisms in Omosita colon (Linnaeus) (Coleoptera: Nitidulidae), we investigated the types, quantities, and distribution of sensilla on the antenna and mouthparts of O. colon by scanning electron microscopy. The clavate antenna comprised the scape, pedicel, and nine segment flagellomeres and had six types of sensilla, including two subtypes of sensilla chaetica (SC), three subtypes of sensilla basiconica (SB) and sensilla trichodea (ST), and one type of sensilla cavity, sensilla styloconica, and Böhm bristles (BB). The chewing mouthparts of O. colon consist of the labrum, mandible, maxillae, labium, and hypopharynx and had seven types sensilla, including two subtypes of SC and sensilla placodea, seven subtypes of SB, and one type of BB, ST, sensilla coeloconica, and sensilla campaniformia. In this research, we also deduced the relationships between the sensilla on the antenna and mouthparts and their functions.

Implementation of Constant Dose Rate and Constant Angular Spacing Intensity-modulated Arc Therapy for Cervical Cancer by Using a Conventional Linear Accelerator.

BACKGROUND: Volumetric-modulated arc therapy (VMAT) can only be implemented on the new generation linacs such as the Varian Trilogy® and Elekta Synergy®. This prevents most existing linacs from delivering VMAT. The purpose of this study was to investigate the feasibility of using a conventional linear accelerator delivering constant dose rate and constant angular spacing intensity-modulated arc therapy (CDR-CAS-IMAT) for treating cervical cancer. METHODS: Twenty patients with cervical cancer previously treated with intensity-modulated radiation therapy (IMRT) using Varian Clinical 23EX were retreated using CDR-CAS-IMAT. The planning target volume (PTV) was set as 50.4 Gy in 28 fractions. Plans were evaluated based on the ability to meet the dose volume histogram. The homogeneity index (HI), target volume conformity index (CI), the dose to organs at risk, radiation delivery time, and monitor units (MUs) were also compared. The paired t-test was used to analyze the two data sets. All statistical analyses were performed using SPSS 19.0 software. RESULTS: Compared to the IMRT group, the CDR-CAS-IMAT group showed better PTV CI (0.85 ± 0.03 vs. 0.81 ± 0.03, P = 0.001), clinical target volume CI (0.46 ± 0.05 vs. 0.43 ± 0.05, P = 0.001), HI (0.09 ± 0.02 vs. 0.11 ± 0.02, P = 0.005) and D95 (5196.33 ± 28.24 cGy vs. 5162.63 ± 31.12 cGy, P = 0.000), and cord D2 (3743.8 ± 118.7 cGy vs. 3806.2 ± 98.7 cGy, P = 0.017) and rectum V40 (41.9 ± 6.1% vs. 44.2 ± 4.8%, P = 0.026). Treatment time (422.7 ± 46.7 s vs. 84.6 ± 7.8 s, P = 0.000) and the total plan Mus (927.4 ± 79.1 vs. 787.5 ± 78.5, P = 0.000) decreased by a factor of 0.8 and 0.15, respectively. The IMRT group plans were superior to the CDR-CAS-IMAT group plans considering decreasing bladder V50 (17.4 ± 4.5% vs. 16.6 ± 4.2%, P = 0.049), bowel V30 (39.6 ± 6.5% vs. 36.6 ± 7.5%, P = 0.008), and low-dose irradiation volume; there were no significant differences in other statistical indexes. CONCLUSIONS: Patients with cervical cancer treated with CDR-CAS-IMAT using Varian Clinical 23EX can get equivalent or superior dose distribution compared to those treated with IMRT. CDR-CAS-IMAT has a less treatment time and MU, which can reduce the uncertainty factor and patient discomfort in treatment.