Assembling High-Quality Lymph Node Clinical Target Volumes for Cervical Cancer Radiotherapy Using a Deep Learning-Based Approach.

AIM: The study aimed to explore an approach for accurately assembling high-quality lymph node clinical target volumes (CTV) on CT images in cervical cancer radiotherapy with the encoder-decoder 3D network. METHODS: 216 cases of CT images treated at our center between 2017 and 2020 were included as a sample, which were divided into two cohorts, including 152 cases and 64 controls, respectively. Para-aortic lymph node, common iliac, external iliac, internal iliac, obturator, presacral, and groin nodal regions were delineated as sub-CTV manually in the cohort including 152 cases. Then, the 152 cases were randomly divided into training (96 cases), validation (36 cases), and test (20 cases) groups for the training process. Each structure was individually trained and optimized through a deep learning model. An additional 64 cases with 6 different clinical conditions were taken as examples to verify the feasibility of CTV generation based on our model. Dice similarity coefficient (DSC) and Hausdorff distance (HD) metrics were both used for quantitative evaluation. RESULTS: Comparing auto-segmentation results to ground truth, the mean DSC value/HD was 0.838/7.7mm, 0.853/4.7mm, 0.855/4.7mm, 0.844/4.7mm, 0.784/5.2mm, 0.826/4.8mm and 0.874/4.8mm for CTV_PAN, CTV_common iliac, CTV_internal iliac, CTV_external iliac, CTV_obturator, CTV_presacral, and CTV_groin, respectively. The similarity comparison results of six different clinical situations were 0.877/4.4mm, 0.879/4.6mm, 0.881/4.2mm, 0.882/4.3mm, 0.872/6.0mm, and 0.875/4.9mm for DSC value/HD, respectively. CONCLUSION: We have developed a deep learning-based approach to segmenting lymph node sub-regions automatically and assembling high-quality CTVs according to clinical needs in cervical cancer radiotherapy. This work can increase the efficiency of the process of cervical cancer detection and treatment.

Shrub and precipitation interactions shape functional diversity of nematode communities on the Qinghai-Tibet Plateau.

Land use and climate change alter biodiversity patterns and ecosystem functioning worldwide. Land abandonment with consequent shrub encroachment and changes in precipitation gradients are known factors in global change. Yet, the consequences of interactions between these factors on the functional diversity of belowground communities remain insufficiently explored. Here, we investigated the dominant shrub effects on the functional diversity of soil nematode communities along a precipitation gradient on the Qinghai-Tibet Plateau. We collected three functional traits (life-history C-P value, body mass, and diet) and calculated the functional alpha and beta diversity of nematode communities using kernel density n-dimensional hypervolumes. We found that shrubs did not significantly alter the functional richness and dispersion, but significantly decreased the functional beta diversity of nematode communities in a pattern of functional homogenization. Shrubs benefited nematodes with longer life-history, larger body mass, and higher trophic levels. Moreover, the shrub effects on the functional diversity of nematodes depended strongly on precipitation. Increasing precipitation reversed the effects shrubs have on the functional richness and dispersion from negative to positive but amplified the negative effects shrubs have on functional beta diversity of nematodes. Benefactor shrubs had stronger effects on the functional alpha and beta diversity of nematodes than allelopathic shrubs along a precipitation gradient. A piecewise structural equation model showed that shrubs and its interactions with precipitation indirectly increased the functional richness and dispersion through plant biomass and soil total nitrogen, whereas it directly decreased the functional beta diversity. Our study reveals the expected changes in soil nematode functional diversity following shrub encroachment and precipitation, advancing our understanding of global climate change on nematode communities on the Qinghai-Tibet Plateau.

[The Effects of Different Adaptive Statistical Iterative Reconstruction-V and Convolution Kernel Parameters on Auto-Segmentation Stability in CT Images].

Objective The study aims to investigate the effects of different adaptive statistical iterative reconstruction-V（ ASiR-V） and convolution kernel parameters on stability of CT auto-segmentation which is based on deep learning. Method Twenty patients who have received pelvic radiotherapy were selected and different reconstruction parameters were used to establish CT images dataset. Then structures including three soft tissue organs (bladder, bowelbag, small intestine) and five bone organs (left and right femoral head, left and right femur, pelvic) were segmented automatically by deep learning neural network. Performance was evaluated by dice similarity coefficient（ DSC） and Hausdorff distance, using filter back projection(FBP) as the reference. Results Auto-segmentation of deep learning is greatly affected by ASIR-V, but less affected by convolution kernel, especially in soft tissues. Conclusion The stability of auto-segmentation is affected by parameter selection of reconstruction algorithm. In practical application, it is necessary to find a balance between image quality and segmentation quality, or improve segmentation network to enhance the stability of auto-segmentation.

Cd and pb Co-Pollution Increased Ecological Risk and Changed Rhizosphere Characteristics of Arabidopsis Thaliana During Phytoremediation.

Previous studies have reported that co-contamination can result in more complex effects on the phytoremediation efficiency of plants relative to those of a single pollutant. However, the effect of co-contamination on plant rhizosphere characteristics has rarely been revealed. This study was carried out to assess the changes in soil pH, the content and fractionation of dissolved organic matter (DOM), and the metal solubility in the rhizosphere of Arabidopsis thaliana when treated with Cd and Pb simultaneously. The results showed that co-contamination increased the concentrations of DOM by 24.8% and 30.9% in the rhizosphere soil of A. thaliana relative to individual Cd or Pb pollution, respectively. At the end of the experiment, co-contamination significantly decreased the initial soil pH from 6.6 ± 0.3 to 5.5 ± 0.4, whereas a decrease was not observed under Pb pollution alone. Variations in soil pH and DOM can change the fractions of the two metals in the rhizosphere soil of A. thaliana. DOM in co-contaminated soil showed a higher Cd (1.05 mg L-1) and Pb (0.75 mg L-1) extraction ability relative to that in the Cd-polluted (0.89 mg Cd L-1 and 0.59 mg Pb L-1) or Pb-polluted (0.68 mg Cd L-1 and 0.63 mg Pb L-1) soils. The soluble Cd content in the co-contaminated (0.44 mg L-1) soil was significantly lower than that in the Cd-polluted (0.71 mg L-1) soil because A. thaliana is a Cd accumulator, whereas the soluble Pb content showed the opposite trend (47.0 mg L-1 vs. 37.4 mg L-1) because the species is a Pb excluder. Therefore, A. thaliana in co-contaminated soil would pose a leaching risk for the non-hyperaccumulated metals, thereby increasing the potential ecological risk during the phytoremediation process.

Quantitative Comparisons of Deep-learning-based and Atlas-based Auto- segmentation of the Intermediate Risk Clinical Target Volume for Nasopharyngeal Carcinoma.

BACKGROUND: Manual segment target volumes were time-consuming and inter-observer variability couldn't be avoided. With the development of computer science, auto-segmentation had the potential to solve this problem. OBJECTIVE: To evaluate the accuracy and stability of Atlas-based and deep-learning-based auto-segmentation of the intermediate risk clinical target volume, composed of CTV2 and CTVnd, for nasopharyngeal carcinoma quantitatively. METHODS AND MATERIALS: A cascade-deep-residual neural network was constructed to automatically segment CTV2 and CTVnd by deep learning method. Meanwhile, a commercially available software was used to automatically segment the same regions by Atlas-based method. The datasets included contrast computed tomography scans from 102 patients. For each patient, the two regions were manually delineated by one experienced physician. The similarity between the two auto-segmentation methods was quantitatively evaluated by Dice similarity coefficient, the 95th Hausdorff distance, volume overlap error and relative volume difference, respectively. Statistical analyses were performed using the ranked Wilcoxon test. RESULTS: The average Dice similarity coefficient (±standard deviation) given by the deep-learning- based and Atlas-based auto-segmentation were 0.84 (±0.03) and 0.74 (±0.04) for CTV2, 0.79 (±0.02) and 0.68 (±0.03) for CTVnd, respectively. For the 95th Hausdorff distance, the corresponding values were 6.30±3.55 mm and 9.34±3.39 mm for CTV2, 7.09±2.27 mm and 14.33±3.98 mm for CTVnd. Besides, volume overlap error and relative volume difference could also predict the same situations. Statistical analyses showed significant difference between the two auto-segmentation methods (p<0.01). CONCLUSION: Compared with the Atlas-based segmentation approach, the deep-learning-based segmentation method performed better both in accuracy and stability for meaningful anatomical areas other than organs at risk.

[Using stacked neural network to improve the auto-segmentation accuracy of Graves' ophthalmopathy target volumes for radiotherapy].

Compared with the previous automatic segmentation neural network for the target area which considered the target area as an independent area, a stacked neural network which uses the position and shape information of the organs around the target area to regulate the shape and position of the target area through the superposition of multiple networks and fusion of spatial position information to improve the segmentation accuracy on medical images was proposed in this paper. Taking the Graves' ophthalmopathy disease as an example, the left and right radiotherapy target areas were segmented by the stacked neural network based on the fully convolutional neural network. The volume Dice similarity coefficient (DSC) and bidirectional Hausdorff distance (HD) were calculated based on the target area manually drawn by the doctor. Compared with the full convolutional neural network, the stacked neural network segmentation results can increase the volume DSC on the left and right sides by 1.7% and 3.4% respectively, while the two-way HD on the left and right sides decrease by 0.6. The results show that the stacked neural network improves the degree of coincidence between the automatic segmentation result and the doctor's delineation of the target area, while reducing the segmentation error of small areas. The stacked neural network can effectively improve the accuracy of the automatic delineation of the radiotherapy target area of Graves' ophthalmopathy.

[Feasibility Study on Location of CT Images Using Convolutional Neural Networks].

OBJECTIVE: To locate CT images by using the deep learning model based on convolutional neural network. METHODS: The AlexNet network was used as a deep learning model, which was preset by the transfer learning approach. Training samples were divided into 4 categories according to the vertebral body parts and labeled, and the data augmentation was used to improve the classification accuracy. RESULTS: The accuracy of image classification after augmentation increased from 94.95% to 97.72%, and the testing time increased from 2.05 s to 3.03 s. CONCLUSIONS: It is feasible to use the convolutional neural network to locate CT images. The data augmentation approach can increase the classification accuracy but also increase the training and testing time.

Exploring the community phylogenetic structure along the slope aspect of subalpine meadows in the eastern Qinghai-Tibetan Plateau, China.

Exploring the community assembly has been important for explaining the maintenance mechanisms of biodiversity and species coexistence, in that it is a central issue in community ecology. Here, we examined patterns of the community phylogenetic structure of the subalpine meadow plant community along the slope gradient in the Qinghai-Tibetan Plateau of China. We surveyed all species and constructed the phylogenetic tree of the plant community based on data from the Angiosperm Phylogeny Group III. We selected the net relative index (NRI) and evaluated the community phylogenetic structure along the five slope plants communities. We found that the phylogenetic structure varied from phylogenetic clustering to phylogenetic overdispersion with the slope aspect from north to south. In the north slope, the community phylogenetically cluster indicated that the limiting similarity played a leading role in the community assembly and the maintenance of biodiversity. Community phylogenetic overdispersion in the east, southeast, and south slopes indicated that habitat filtration was the driving force for community assembly. The NRI index of the northeast slope was close to zero, implying random dispersion. But it may be driven by the neutral process or limiting similarity, in that the community assembly process was the result of a combination of several ecological factors and thus required further study.

Degradation of bisphenol A in aqueous solution by persulfate activated with ferrous ion.

Degradation of bisphenol A (BPA) in aqueous solution was studied with high-efficiency sulfate radical (SO4(-·)), which was generated by the activation of persulfate (S2O8(2-)) with ferrous ion (Fe(2+)). S2O8(2-) was activated by Fe(2+) to produce SO4(-·), and iron powder (Fe(0)) was used as a slow-releasing source of dissolved Fe(2+). The major oxidation products of BPA were determined by liquid chromatography-mass spectrometer. The mineralization efficiency of BPA was monitored by total organic carbon (TOC) analyzer. BPA removal efficiency was improved by the increase of initial S2O8(2-) or Fe(2+) concentrations and then decreased with excess Fe(2+) concentration. The adding mode of Fe(2+) had significant impact on BPA degradation and mineralization. BPA removal rates increased from 49 to 97% with sequential addition of Fe(2+), while complete degradation was observed with continuous diffusion of Fe(2+), and the latter achieved higher TOC removal rate. When Fe(0) was employed as a slow-releasing source of dissolved Fe(2+), 100% of BPA degradation efficiency was achieved, and the highest removal rate of TOC (85%) was obtained within 2 h. In the Fe(0)-S2O8(2-) system, Fe(0) as the activator of S2O8(2-) could offer sustainable oxidation for BPA, and higher TOC removal rate was achieved. It was proved that Fe(0)-S2O8(2-) system has perspective for future works.

Photocatalytic degradation of 4-tert-octylphenol in a spiral photoreactor system.

A spiral photoreactor system (SPS) was developed for the degradation of 4-tert-octylphenol (4-t-OP) in aqueous phase. 4-t-OP was previously considered as a endocrine disrupting compound frequently present in water. The direct photodegradation reaction caused by the SPS was found to accord with the characteristic of apparent first-order reaction with reaction rate constant k = 4.8 x 10(-2) min(-1). However, the direct photodegradation reaction could not make the 4-t-OP mineralized. The photodegradation efficiency increased from 88% to 91.2% in 45 min irradiation period after the internal surface of SPS was sintered with TiO2 thin film as catalyst. Catalyst concentration, number of catalyst coating layers and initial concentration of 4-t-OP were proven to be the factors affecting the photocatalytic degradation performance of the SPS on aqueous 4-t-OP. The degradation mechanism was investigated and the byproducts were analyzed using total organic carbon analyzer (TOC) and LC-MS. The possible chemical structures of the products were suggested. SPS with single layer of TiO2 prepared by sintering 13.6% of TiO2 precursor was proven to be more efficient than most of previous systems for removal of 4-t-OP from aqueous phase. 28.3% of the 4-t-OP was mineralized in 45 min according to the decreased amount of TOC value.