Hefei Institute of Technology Innovation

This study aimed to identify a novel plasma lncRNA biomarker and establish a model based on lncRNAs and radiomics for predicting the response of non-small cell lung cancer (NSCLC) to chemo- and radio-therapy. Next-generation sequencing and integrated bioinformatics analysis were used to identify lncRNAs associated with the response of NSCLC to chemo- and radio-therapy. RT-qPCR was utilized to detect MIF-AS1 expression in the plasma of NSCLC patients. Radiomics analysis was performed on CT images of NSCLC patients. The model was constructed by multiple logistic regression. The expression of the lncRNA MIF-AS1 was up-regulated in the plasma of patients with chemo- and radio-resistant NSCLC, as validated by RT-qPCR in 124 NSCLC patients. Furthermore, in vitro experiments demonstrated that knockdown of MIF-AS1 expression significantly reduced cellular proliferation and invasion, and increased the sensitivity of NSCLC cells to the chemotherapeutic drug cisplatin. Using the ceRNA network, a DNA-damage repair related protein RAD21 was identified as a target gene of MIF-AS1. Finally, two radiomic features were found to be associated with the response of NSCLC to chemo- and radio-therapy. Combining the MIF-AS1 level and the two radiomic features, a model was established to predict the response of NSCLC to chemo- and radio-therapy, with a high AUC of 0.808. MIF-AS1 could be a novel biomarker for predicting the response of NSCLC to chemo- and radio-therapy. This model, which uses both CT radiomics and MIF-AS1 levels, increases the accuracy of predicting therapeutic response in NSCLC patients.

The topological noncollinear antiferromagnet (AFM) Mn3Sn exhibits a giant anomalous Hall conductance (AHC) originating from its nonvanishing Berry curvature. Conventionally, the two AHC states are regarded as time-reversal pairs coupled to the magnetic octupole moment, and their control has relied on reversing this moment by external magnetic fields or electric currents. Here, an alternative mechanism is demonstrated-the chirality-selected noncollinear antiferromagnetic state-in which the AHC polarity is defined by the vector spin chirality (VSC) of the Kagome lattice. By constructing Mn3Sn/Pt heterostructures, a Fert-Levy-type Dzyaloshinskii-Moriya interaction (DMI) is introduced that sets the lattice chirality. The induced DMI changes the VSC from counterclockwise (CCW) to clockwise (CW), resulting in a corresponding inversion of the AHC sign. This behavior is confirmed by symmetry analysis and atomistic simulations that link the polarity inversion to the competition between DMI energy and intrinsic anisotropy. These findings establish a chirality-defined route for controlling noncollinear antiferromagnetic order and highlight DMI engineering as a powerful means of tailoring Berry-curvature-driven transport in AFMs.

High-precision autonomous localization remains a critical challenge for intelligent mining vehicles in GNSS-denied and unstructured coal mine roadways, where traditional odometry-based methods suffer from severe cumulative drift and perceptual aliasing. Inspired by the synergy between mammalian visual cues and cognitive neural mechanisms, this paper proposes a robust biomimetic localization framework that integrates multi-source perception with a prior cognitive map. The core contributions are three-fold: First, a semantic-enhanced biomimetic localization method is developed, leveraging roadway sign data as absolute spatial anchors to suppress long-distance cumulative errors. Second, an optimized head direction (HD) cell model is formulated by incorporating a speed balance factor, kinematic constraints, and a drift correction influence factor, significantly improving the precision of angular perception. Third, boundary-adaptive and sign-based semantic constraint terms are integrated into a continuous attractor network (CAN)-based path integration model, effectively preventing trajectory deviation into non-navigable regions. Comprehensive evaluations conducted in large-scale underground scenarios demonstrate that the proposed framework consistently outperforms conventional IMU-odometry fusion, representative 3D SLAM solutions, and baseline biomimetic algorithms. By effectively integrating semantic landmarks as spatial anchors, the system exhibits superior resilience against cumulative drift, maintaining high localization precision where standard methods typically diverge. The results confirm that our approach significantly enhances both trajectory consistency and heading stability across extensive distances, validating its robustness and scalability in handling the inherent complexities of unstructured coal mine environments for enhanced intrinsic safety.

Self-supervised learning has shown potential in fine-grained visual recognition (FGVR). However, existing self-supervised learning methods are often susceptible to irrelevant patterns during training and lack the ability to capture the critical subtle differences in FGVR, leading to suboptimal performance. Moreover, existing approaches focus primarily on uni-modal visual concepts. Despite the emergence of powerful vision-language models (VLMs) in various high-level vision tasks, their potential in self-supervised FGVR remains largely unexplored. To this end, we propose a novel self-supervised learning (LearnMat) framework, that effectively filters out irrelevant feature interference and extracts more important and subtle discriminative features during training. Specifically, LearnMat consists of two key modules: the semantic awareness module (SAM) and the insight extraction module (IEM). In the SAM, we introduce a novel vision-language-grounded semantic distillation strategy using a corpus of generic, category-agnostic textual attributes, that injects explicit semantic constraints into self-supervised training and improves robustness to background interference. Complementarily, the IEM exploits gradient-based signals from the input image to highlight subtle differences and localize key discriminative regions, mitigating inter-class similarity and intra-class variation, and enhancing fine-grained discrimination. Extensive experiments across multiple popular FGVR datasets show that LearnMat significantly outperforms recent state-of-the-art methods, highlighting its marked effectiveness. Our code is avaliable at https://github.com/Heng-CHY/LearnMat.

Hydnum (Hydnaceae, Cantharellales), one of the edible ectomycorrhizal fungi, is characterized by a spine-bearing hymenophore. It is widely distributed in temperate regions and forms stable symbiotic relationships with Fagaceae and Pinaceae. During a survey of macrofungi in the Dabie Mountains region of China, ten specimens of Hydnum were collected. Based on morphological characteristics and phylogenetic analysis using three genetic markers (ITS + nrLSU + tef1-α), three new species (H. luteoalbum, H. albodentum, and H. albotomentosum) were identified and described, and two species newly recorded from the Dabie Mountains (H. berkeleyanum and H. pallidomarginatum) were reported. H. luteoalbum is distinguished by a white pileus covered with white tomentum, dagger-shaped or sword-like spines, and broadly ellipsoid basidiospores. H. albodentum is characterized by a pale brown pileus and subelliptical basidiospores (8.0-8.5 × 6.0-7.0 μm; av. Q = 1.17). H. albotomentosum features smaller basidiocarps, extremely short spines (0.5-2 mm), and globose to subglobose basidiospores. This study enriches the known taxonomic diversity of Hydnum and provides a dichotomous key to the species of Hydnum in China to facilitate species identification.

Accurate Body mass index (BMI) prediction is essential for early prevention and health management. In this study, we developed an interpretable machine learning model to predict adolescents’ BMI using readily available questionnaire data. This 1-year longitudinal study gathered initial data through physical measurements and questionnaires, with BMI re-evaluated after 12 months. We first conducted univariate analysis to identify predictive factors, then developed seven machine learning models for BMI prediction. Model performance was evaluated using Mean Squared Error (MSE), Mean Absolute Error (MAE), and coefficient of determination (R2). Model calibration was performed to further improve prediction accuracy. SHapley Additive exPlanations (SHAP) was applied to interpret the model. Among the 1,798 students in the study cohort, the initial average BMI was 21.12 ± 3.45 kg/m2, which increased to 21.50 ± 3.47 kg/m2 after one year (average change: 0.38 ± 1.36 kg/m2). Based on univariate analysis, 21 predictive factors were identified. The calibrated CatBoost (CB) model achieved the best predictive performance on the test set, achieving an MSE of 0.998 (95% confidence interval [CI]: 0.846–1.173), MAE of 0.769 (95% CI: 0.707–0.835), and R2 of 0.919 (95% CI: 0.900–0.936). SHAP analysis revealed how daily information affects BMI changes and offered valuable insights for the development of personalized weight and health management strategies for adolescents.

Stroke often leads to persistent upper limb dysfunction that impairs activities of daily living, yet objective biomechanical indicators for precise assessment remain limited. This study aimed to characterize phase-specific impairments in energy output, torque stability, and muscle coordination during the hand-to-mouth (HTM) task and to explore their potential for improving rehabilitation evaluation. Motion data from 20 stroke patients and 20 healthy controls were recorded using wearable surface electromyography and inertial measurement unit systems. A musculoskeletal model was applied to calculate joint torque, mechanical work, torque smoothness, and a novel torque-based co-contraction index across four movement subphases. These phase-specific metrics demonstrated significant correlations with clinical motor impairment scores, confirming their clinical validity. Significant dynamic features were then selected to construct machine learning models for group classification. Stroke patients showed reduced output capacity, increased torque fluctuations, and abnormal co-contraction patterns that varied across subphases. Among the classifiers, the quadratic support vector machine achieved the best performance, with an accuracy of 84.6% and an AUC of 0.853, surpassing models based on whole-task features. These findings demonstrate that phase-specific biomechanical features sensitively capture neuromuscular deficits in stroke survivors and highlight the potential of phase-specific biomechanics to inform future individualized rehabilitation assessment and treatment planning.

The quality testing of traditional Chinese medicine injection (TCMI) provides quantitative indicators for its safety evaluation, improving quality control and reducing adverse reaction rates. However, no automated Microtox-based testing equipment currently exists globally, necessitating manual intervention in the entire process. To address this, our team developed an automated TCMI quality testing system using Microtox technology. By integrating manual workflows and standardizing reusable technologies, the instrument comprises three modules: a mixing-resuscitation module, a mechanical pipetting module, and a light intensity measurement module. Centralized computer control enables fully automated TCMI quality testing. Experimental results demonstrate strong consistency between manual and automated testing, with reduced human-induced measurement errors and enhanced data reliability for TCMI quality assessment.