MultiFuseYOLO: Redefining Wine Grape Variety Recognition through Multisource Information Fusion.

Based on the current research on the wine grape variety recognition task, it has been found that traditional deep learning models relying only on a single feature (e.g., fruit or leaf) for classification can face great challenges, especially when there is a high degree of similarity between varieties. In order to effectively distinguish these similar varieties, this study proposes a multisource information fusion method, which is centered on the SynthDiscrim algorithm, aiming to achieve a more comprehensive and accurate wine grape variety recognition. First, this study optimizes and improves the YOLOV7 model and proposes a novel target detection and recognition model called WineYOLO-RAFusion, which significantly improves the fruit localization precision and recognition compared with YOLOV5, YOLOX, and YOLOV7, which are traditional deep learning models. Secondly, building upon the WineYOLO-RAFusion model, this study incorporated the method of multisource information fusion into the model, ultimately forming the MultiFuseYOLO model. Experiments demonstrated that MultiFuseYOLO significantly outperformed other commonly used models in terms of precision, recall, and F1 score, reaching 0.854, 0.815, and 0.833, respectively. Moreover, the method improved the precision of the hard to distinguish Chardonnay and Sauvignon Blanc varieties, which increased the precision from 0.512 to 0.813 for Chardonnay and from 0.533 to 0.775 for Sauvignon Blanc. In conclusion, the MultiFuseYOLO model offers a reliable and comprehensive solution to the task of wine grape variety identification, especially in terms of distinguishing visually similar varieties and realizing high-precision identifications.

A novel PCDscore based on programmed cell death-related genes can effectively predict prognosis and therapy responses of colon adenocarcinoma.

Emerging evidence suggests a correlation between oncogenesis and programmed cell death (PCD). However, comprehensive studies that incorporate all identified PCD-related genes to guide colon adenocarcinoma (COAD) prognosis and precision treatment strategies are lacking. In this study, a series of bioinformatics analyses were comprehensively conducted using data from the TCGA-COAD, GSE17538, and GSE39582 cohorts. A total of 21 PCD-associated prognostic genes were identified through univariate Cox analysis. LASSO and multivariate Cox methods were employed to establish a prognostic gene signature (ALOX12, HSPA1A, IL13, MID2, RFFL, and SLC39A8) and the corresponding scoring system, termed PCDscore, which exhibited robust predictive ability. The ssGSEA and ESTIMATE algorithms were utilized to evaluate the tumor microenvironment of COAD. The high PCDscore group demonstrated a poorer prognosis, characterized by lower CD4+ T cell infiltration and a higher stromal score. In contrast, the low PCDscore group exhibited sensitivity to common chemotherapy drugs such as Cisplatin and 5-Fluorouracil. Single-cell sequencing analysis further revealed that the high-PCDscore group displayed a lower proportion of CD4+ T cells. Colorectal cancer samples from the years 2013-2017 were employed to validate the PCDscore, while those from 2018 to 2019 served as a temporal external validation set for the PCDscore. In vitro experimental results indicated that the overexpression of SLC39A8 inhibited the proliferation and invasion of colorectal cancer cells. The study developed a novel PCDscore system based on the analysis of genes related to all identified PCD types, providing valuable insights into clinical prognosis and drug sensitivity for patients with COAD.

Robust optical edge detection enabled by a twisted reflective q-plate.

Optical edge detection significantly reduces the image information load and is highly sought after in instant image processing. Robustness to the wavelength and polarization of light as well as mechanical vibration is a key requirement for practical applications. Here, a robust optical edge detector is proposed and demonstrated based on a reflective twisted liquid crystal q-plate. The device is composed of a mirror and a 1.46-µm-thick liquid crystal layer with a twist angle of 69.2°. The backtracking of the light inside the twisted medium forms a mirror symmetric twisted design and thus leads to a broadband self-compensated spiral phase modulation. By this means, an optical edge detector with excellent wavelength and polarization independence is presented for both coherent and partially coherent light sources. Additionally, the reflective design makes the system more compact and stable. This work supplies a practical design for robust optical edge detection, which may upgrade existing image processing techniques.

Correlating the Microstructure and Current Density of the Li/Garnet Interface.

Solid-state lithium batteries hold great promise for next-generation energy storage systems. However, the formation of lithium filaments within the solid electrolyte remains a critical challenge. In this study, we investigate the crucial role of morphology in determining the resistance of garnet-type electrolytes to lithium filaments. By proposing a new test method, namely, cyclic linear sweep voltammetry, we can effectively evaluate the electrolyte resistance against lithium filaments. Our findings reveal a strong correlation between the microscopic morphology of the solid electrolyte and its resistance to lithium filaments. Samples with reduced pores and multiple grain boundaries demonstrate remarkable performance, achieving a critical current density of up to 3.2 mA cm-2 and excellent long-term cycling stability. Kelvin probe force microscopy and finite element method simulation results shed light on the impact of grain boundaries and electrolyte pores on lithium-ion transport and filament propagation. To inhibit lithium penetration, minimizing pores and achieving a uniform morphology with small grains and plenty of grain boundaries are essential.

Water-solubility croconic acid-bisindole dye with morpholine ring for tumor NIRF/PA imaging and photothermal therapy activated by lysosome pH-response.

A novel croconic acid-bisindole dye CR-630 with a morpholine ring showed good water-solubility and obvious lysosome-targeting. The protonation of the nitrogen atom in the indole and lysosome-targeting of morpholine ring let it exhibit stronger pH-responsive NIR/PA imaging and photothermal effect in the lysosome acidic microenvironment (pH 4.0-5.5) than in the tumor acidic microenvironment. In the animal study, compound CR-630 could NIRF/PA image in the tumor tissues in 1.5-2.0 h, effectively inhibit the growth of the tumor, and even ablate the tumor at the drug dose of 1 mg/kg. It also demonstrated good biosafety. This study gives a new idea to develop water-solubility organic dyes with lysosome targeting, stronger pH-responsive NIRF/PA imaging and PTT for breast cancer.

Context Label Learning: Improving Background Class Representations in Semantic Segmentation.

Background samples provide key contextual information for segmenting regions of interest (ROIs). However, they always cover a diverse set of structures, causing difficulties for the segmentation model to learn good decision boundaries with high sensitivity and precision. The issue concerns the highly heterogeneous nature of the background class, resulting in multi-modal distributions. Empirically, we find that neural networks trained with heterogeneous background struggle to map the corresponding contextual samples to compact clusters in feature space. As a result, the distribution over background logit activations may shift across the decision boundary, leading to systematic over-segmentation across different datasets and tasks. In this study, we propose context label learning (CoLab) to improve the context representations by decomposing the background class into several subclasses. Specifically, we train an auxiliary network as a task generator, along with the primary segmentation model, to automatically generate context labels that positively affect the ROI segmentation accuracy. Extensive experiments are conducted on several challenging segmentation tasks and datasets. The results demonstrate that CoLab can guide the segmentation model to map the logits of background samples away from the decision boundary, resulting in significantly improved segmentation accuracy. Code is available at https://github.com/ZerojumpLine/CoLab.

Causality-Inspired Single-Source Domain Generalization for Medical Image Segmentation.

Deep learning models usually suffer from the domain shift issue, where models trained on one source domain do not generalize well to other unseen domains. In this work, we investigate the single-source domain generalization problem: training a deep network that is robust to unseen domains, under the condition that training data are only available from one source domain, which is common in medical imaging applications. We tackle this problem in the context of cross-domain medical image segmentation. In this scenario, domain shifts are mainly caused by different acquisition processes. We propose a simple causality-inspired data augmentation approach to expose a segmentation model to synthesized domain-shifted training examples. Specifically, 1) to make the deep model robust to discrepancies in image intensities and textures, we employ a family of randomly-weighted shallow networks. They augment training images using diverse appearance transformations. 2) Further we show that spurious correlations among objects in an image are detrimental to domain robustness. These correlations might be taken by the network as domain-specific clues for making predictions, and they may break on unseen domains. We remove these spurious correlations via causal intervention. This is achieved by resampling the appearances of potentially correlated objects independently. The proposed approach is validated on three cross-domain segmentation scenarios: cross-modality (CT-MRI) abdominal image segmentation, cross-sequence (bSSFP-LGE) cardiac MRI segmentation, and cross-site prostate MRI segmentation. The proposed approach yields consistent performance gains compared with competitive methods when tested on unseen domains.

Enhancing MR image segmentation with realistic adversarial data augmentation.

The success of neural networks on medical image segmentation tasks typically relies on large labeled datasets for model training. However, acquiring and manually labeling a large medical image set is resource-intensive, expensive, and sometimes impractical due to data sharing and privacy issues. To address this challenge, we propose AdvChain, a generic adversarial data augmentation framework, aiming at improving both the diversity and effectiveness of training data for medical image segmentation tasks. AdvChain augments data with dynamic data augmentation, generating randomly chained photo-metric and geometric transformations to resemble realistic yet challenging imaging variations to expand training data. By jointly optimizing the data augmentation model and a segmentation network during training, challenging examples are generated to enhance network generalizability for the downstream task. The proposed adversarial data augmentation does not rely on generative networks and can be used as a plug-in module in general segmentation networks. It is computationally efficient and applicable for both low-shot supervised and semi-supervised learning. We analyze and evaluate the method on two MR image segmentation tasks: cardiac segmentation and prostate segmentation with limited labeled data. Results show that the proposed approach can alleviate the need for labeled data while improving model generalization ability, indicating its practical value in medical imaging applications.

Clinical evaluation of prone position ventilation in the treatment of acute respiratory distress syndrome induced by sepsis.

BACKGROUND: Acute respiratory distress syndrome (ARDS) is an acute, diffuse, inflammatory lung injury. Previous studies have shown prone position ventilation (PPV) to be associated with improvement in oxygenation. However, its role in patients with ARDS caused by sepsis remains unknown. AIM: To analyze the clinical effects of PPV in patients with ARDS caused by sepsis. METHODS: One hundred and two patients with ARDS were identified and divided into a control group (n = 55) and a PPV treatment group (n = 47). Outcomes included oxygenation index, lung compliance (Cst) and platform pressure (Pplat), which were compared between the two groups after ventilation. Other outcomes included heart rate (HR), mean arterial pressure (MAP), central venous pressure (CVP), left ventricular ejection fraction (LVEF), the length of mechanical ventilation time and intensive care unit (ICU) stay, and levels of C-reactive protein (CRP), procalcitonin (PCT), and interleukin-6 (IL-6) after ventilation. Finally, mortality rate was also compared between the two groups. RESULTS: On the first day after ventilation, the oxygenation index and Cst were higher and Pplat level was lower in the PPV group than in the conventional treatment group (P < 0.05). There were no significant differences in oxygenation index, Cst, and Pplat levels between the two groups on the 2nd, 4th, and 7th day after ventilation (P > 0.05). There were no significant differences in HR, MAP, CVP, LVEF, duration of mechanical ventilation and ICU stay, and the levels of CRP, PCT, and IL-6 between the two groups on the first day after ventilation (all P > 0.05). The mortality rates on days 28 and 90 in the PPV and control groups were 12.77% and 29.09%, and 25.53% and 45.45%, respectively (P < 0.05). CONCLUSION: PPV may improve respiratory mechanics indices and may also have mortality benefit in patients with ARDS caused by sepsis. Finally, PPV was not shown to cause any adverse effects on hemodynamics and inflammation indices.

Cardiac segmentation on late gadolinium enhancement MRI: A benchmark study from multi-sequence cardiac MR segmentation challenge.

Accurate computing, analysis and modeling of the ventricles and myocardium from medical images are important, especially in the diagnosis and treatment management for patients suffering from myocardial infarction (MI). Late gadolinium enhancement (LGE) cardiac magnetic resonance (CMR) provides an important protocol to visualize MI. However, compared with the other sequences LGE CMR images with gold standard labels are particularly limited. This paper presents the selective results from the Multi-Sequence Cardiac MR (MS-CMR) Segmentation challenge, in conjunction with MICCAI 2019. The challenge offered a data set of paired MS-CMR images, including auxiliary CMR sequences as well as LGE CMR, from 45 patients who underwent cardiomyopathy. It was aimed to develop new algorithms, as well as benchmark existing ones for LGE CMR segmentation focusing on myocardial wall of the left ventricle and blood cavity of the two ventricles. In addition, the paired MS-CMR images could enable algorithms to combine the complementary information from the other sequences for the ventricle segmentation of LGE CMR. Nine representative works were selected for evaluation and comparisons, among which three methods are unsupervised domain adaptation (UDA) methods and the other six are supervised. The results showed that the average performance of the nine methods was comparable to the inter-observer variations. Particularly, the top-ranking algorithms from both the supervised and UDA methods could generate reliable and robust segmentation results. The success of these methods was mainly attributed to the inclusion of the auxiliary sequences from the MS-CMR images, which provide important label information for the training of deep neural networks. The challenge continues as an ongoing resource, and the gold standard segmentation as well as the MS-CMR images of both the training and test data are available upon registration via its homepage (www.sdspeople.fudan.edu.cn/zhuangxiahai/0/mscmrseg/).

Chiral RuII -PtII Complexes Inducing Telomere Dysfunction against Cisplatin-Resistant Cancer Cells.

The application of G-quadruplex stabilizers presents a promising anticancer strategy. However, the molecular crowding conditions within cells diminish the potency of current G-quadruplex stabilizers. Herein, chiral RuII -PtII dinuclear complexes were developed as highly potent G-quadruplex stabilizers even under challenging molecular crowding conditions. The compounds were encapsulated with biotin-functionalized DNA cages to enhance sub-cellular localization and provide cancer selectivity. The nanoparticles were able to efficiently inhibit the endogenous activities of telomerase in cisplatin-resistant cancer cells and cause cell death by apoptosis. The nanomaterials demonstrated high antitumor activity towards cisplatin-resistant tumor cells as well as tumor-bearing mice. To the best of our knowledge, this study presents the first example of a RuII -PtII dinuclear complex as a G-quadruplex stabilizer with an anti-cancer effect towards drug-resistant tumors inside an animal model.

Self-Supervised Learning for Few-Shot Medical Image Segmentation.

Fully-supervised deep learning segmentation models are inflexible when encountering new unseen semantic classes and their fine-tuning often requires significant amounts of annotated data. Few-shot semantic segmentation (FSS) aims to solve this inflexibility by learning to segment an arbitrary unseen semantically meaningful class by referring to only a few labeled examples, without involving fine-tuning. State-of-the-art FSS methods are typically designed for segmenting natural images and rely on abundant annotated data of training classes to learn image representations that generalize well to unseen testing classes. However, such a training mechanism is impractical in annotation-scarce medical imaging scenarios. To address this challenge, in this work, we propose a novel self-supervised FSS framework for medical images, named SSL-ALPNet, in order to bypass the requirement for annotations during training. The proposed method exploits superpixel-based pseudo-labels to provide supervision signals. In addition, we propose a simple yet effective adaptive local prototype pooling module which is plugged into the prototype networks to further boost segmentation accuracy. We demonstrate the general applicability of the proposed approach using three different tasks: organ segmentation of abdominal CT and MRI images respectively, and cardiac segmentation of MRI images. The proposed method yields higher Dice scores than conventional FSS methods which require manual annotations for training in our experiments.

Noniridescent structural color from enhanced electromagnetic resonances of particle aggregations and its applications for reconfigurable patterns.

HYPOTHESIS: The conventional noniridescent structural colors refer to the coherent scattering of visible light by the short-range ordered structures assembled from the small colloids (100-250 nm). Our hypothesis is that noniridescent structural color can be generated by the random aggregations of large silica particles through the enhanced electromagnetic resonances. EXPERIMENTS: The random aggregations of large silica particles (350-475 nm) were prepared through the infiltration of silica particles solution with the porous substrate. The mechanism of the structural color is investigated. Reconfigurable patterns are prepared. FINDINGS: Dissimilar to the conventional noniridescent colors, the angle-independent colors of silica aggregations originate from the enhanced electromagnetic resonances due to the random aggregation of the particles. The colors (blue, green, and red) and corresponding reflection peak positions of the particle aggregations can be well controlled by simply altering the size of the silica particles. Compared to the traditional prints with permanent patterns, reconfigurable patterns with large-area and multicolor can be fabricated by the repeatedly selective spray of water on the substrate pre-coated with noniridescent colors. This work provides new insight and greenway for the fabrication of noniridescent structural colors and reconfigurable patterns, and will promote their applications in soft display, green printing, and anti-counterfeiting.

Simple and efficient fabrication of multi-stage color-changeable photonic prints as anti-counterfeit labels.

Color changeable photonic prints (CCPPs) show their potential applications in high-level information storage and anti-counterfeiting, but usually suffer from the complex fabrication process and limited color variation. Here, a simple and efficient method is developed to generate CCPPs with multilevel tunable color contrasts by packing the solvent responsive photonic crystals with diverse cross-linking degrees and desired way. The key to the successful fabrication is to create and control over the optical response of each part of the CCPPs through altering the cross-linking degree of PCs and thus the affinity between the CCPPs and solvents. A CCPPs based anti-fake label with the encrypted information functionality which originates from reversible color change between dried state and swelling with the mixture of acetic acid and ethanol is investigated. Compared with conventional CCPPs, the as-prepared CCPPs can reveal multistage information depending on the volume fraction of ethanol. This work provides a new insight for the simple fabrication of CCPPs and will facilitate their applications in the information protection and high-level anti-counterfeiting.

Supramolecular Assembly of An Organoplatinum(II) Complex with Ratiometric Dual Emission for Two-Photon Bioimaging.

The organoplatinum(II) complex [Pt(C^N^N)(Cl)] (C^N^N=5,6-diphenyl-2,2'-bipyridine, Pt1) can assemble into nanoaggregates via π-π stacking and complementary hydrogen bonds, rather than Pt-Pt interactions. Pt1 exhibits ratiometric dual emission, including rare blue emission (λem =445 nm) and assembly-induced yellow emission (λem =573 nm), under one- and two-photon excitation. Pt1 displays blue emission in cells with an intact membrane due to its low cellular uptake. In cells where the membrane is disrupted, uptake of the complex is increased and at higher concentrations yellow emission is observed. The ratio of yellow to blue emission shows a linear relationship to the loss of cell membrane integrity. Pt1 is, to our knowledge, the first example of an assembly-induced two-photon ratiometric dual emission organoplatinum complex. The excellent and unique characteristics of the complex enabled its use for the tracking of cell apoptosis, necrosis, and the inflammation process in zebrafish.

Correction: Synthesis, characterization and anticancer mechanism studies of fluorinated cyclometalated ruthenium(ii) complexes.

Correction for 'Synthesis, characterization and anticancer mechanism studies of fluorinated cyclometalated ruthenium(ii) complexes' by Ya Wen et al., Dalton Trans., 2020, DOI: .

Synthesis, characterization and anticancer mechanism studies of fluorinated cyclometalated ruthenium(ii) complexes.

The drug-resistance of cancer cells has become a major obstacle to the development of clinical drugs for chemotherapy. In order to overcome cisplatin-resistance, seven cyclometalated ruthenium(ii) complexes were synthesized with a varying degree of fluorine substitution, for use as anticancer agents. A cytotoxicity assay testified that the complexes possessed a more cytotoxic effect than cisplatin towards the cisplatin-resistant cell line A549R. The number of fluorine atoms regulated the lipophilicity of the complexes, but the relationship was not linear. Ru1 containing one fluorine atom had the highest lipophilicity and the best therapeutic effect. The complexes enter cells through an energy-dependent pathway and then localize in the nuclei and mitochondria. The complexes induced nuclear dysfunction by the inhibition of DNA replication as well as mitochondrial dysfunction by the loss of membrane potential. The damage to these vital organelles leads to cell apoptosis via the caspase 3/7 pathway. Our results indicated that the modulation of the number of fluorine atoms in therapeutic agents can have a profound effect and Ru1 is a complex with a high potential as a drug for the treatment of cisplatin-resistant cancer.

A mitochondria-targeting dinuclear Ir-Ru complex as a synergistic photoactivated chemotherapy and photodynamic therapy agent against cisplatin-resistant tumour cells.

A mitochondria-targeting hetero-binuclear Ir(iii)-Ru(ii) complex was developed as a photoactivated chemotherapy (PACT) and photodynamic therapy (PDT) bifunctional agent to achieve a synergistic effective therapeutic outcome for the therapy of cisplatin-resistant tumour cells.

The stepwise photodamage of organelles by two-photon luminescent ruthenium(ii) photosensitizers.

Ru(ii) polypyridyl complexes, containing a morpholine moiety, and possessing two-photon absorption properties and pH dependent singlet oxygen production were used for stepwise lysosomes-to-mitochondria photodamage of cancer cells.

Parallel beamlet dose calculation via beamlet contexts in a distributed multi-GPU framework.

PURPOSE: Dose calculation is one of the most computationally intensive, yet essential tasks in the treatment planning process. With the recent interest in automatic beam orientation and arc trajectory optimization techniques, there is a great need for more efficient model-based dose calculation algorithms that can accommodate hundreds to thousands of beam candidates at once. Foundational work has shown the translation of dose calculation algorithms to graphical processing units (GPUs), lending to remarkable gains in processing efficiency. But these methods provide parallelization of dose for only a single beamlet, serializing the calculation of multiple beamlets and under-utilizing the potential of modern GPUs. In this paper, the authors propose a framework enabling parallel computation of many beamlet doses using a novel beamlet context transformation and further embed this approach in a scalable network of multi-GPU computational nodes. METHODS: The proposed context-based transformation separates beamlet-local density and TERMA into distinct beamlet contexts that independently provide sufficient data for beamlet dose calculation. Beamlet contexts are arranged in a composite context array with dosimetric isolation, and the context array is subjected to a GPU collapsed-cone convolution superposition procedure, producing the set of beamlet-specific dose distributions in a single pass. Dose from each context is converted to a sparse representation for efficient storage and retrieval during treatment plan optimization. The context radius is a new parameter permitting flexibility between the speed and fidelity of the dose calculation process. A distributed manager-worker architecture is constructed around the context-based GPU dose calculation approach supporting an arbitrary number of worker nodes and resident GPUs. Phantom experiments were executed to verify the accuracy of the context-based approach compared to Monte Carlo and a reference CPU-CCCS implementation for single beamlets and broad beams composed by addition of beamlets. Dose for representative 4π beam sets was calculated in lung and prostate cases to compare its efficiency with that of an existing beamlet-sequential GPU-CCCS implementation. Code profiling was also performed to evaluate the scalability of the framework across many networked GPUs. RESULTS: The dosimetric accuracy of the context-based method displays <1.35% and 2.35% average error from the existing serialized CPU-CCCS algorithm and Monte Carlo simulation for beamlet-specific PDDs in water and slab phantoms, respectively. The context-based method demonstrates substantial speedup of up to two orders of magnitude over the beamlet-sequential GPU-CCCS method in the tested configurations. The context-based framework demonstrates near linear scaling in the number of distributed compute nodes and GPUs employed, indicating that it is flexible enough to meet the performance requirements of most users by simply increasing the hardware utilization. CONCLUSIONS: The context-based approach demonstrates a new expectation of performance for beamlet-based dose calculation methods. This approach has been successful in accelerating the dose calculation process for very large-scale treatment planning problems - such as automatic 4π IMRT beam orientation and VMAT arc trajectory selection, with hundreds of thousands of beamlets - in clinically feasible timeframes. The flexibility of this framework makes it as a strong candidate for use in a variety of other very large-scale treatment planning tasks and clinical workflows.