Search

In the field of computational oncology, patient status is often assessed using radiology-genomics, which includes two key technologies and data, such as radiology and genomics. Recent advances in deep learning have facilitated the integration of radiology-genomics data, and even new omics data, significantly improving the robustness and accuracy of clinical predictions. These factors are driving artificial intelligence (AI) closer to practical clinical applications. In particular, deep learning models are crucial in identifying new radiology-genomics biomarkers and therapeutic targets, supported by explainable AI (xAI) methods. This review focuses on recent developments in deep learning for radiology-genomics integration, highlights current challenges, and outlines some research directions for multimodal integration and biomarker discovery of radiology-genomics or radiology-omics that are urgently needed in computational oncology.

Abstract Cancer is a complex disease composing systemic alterations in multiple scales. In this study, we develop the Tumor Multi-Omics pre-trained Network (TMO-Net) that integrates multi-omics pan-cancer datasets for model pre-training, facilitating cross-omics interactions and enabling joint representation learning and incomplete omics inference. This model enhances multi-omics sample representation and empowers various downstream oncology tasks with incomplete multi-omics datasets. By employing interpretable learning, we characterize the contributions of distinct omics features to clinical outcomes. The TMO-Net model serves as a versatile framework for cross-modal multi-omics learning in oncology, paving the way for tumor omics-specific foundation models.

Purpose This study investigated the topological structural characteristics of systemic lupus erythematosus (SLE) with and without neuropsychiatric symptoms (NPSLE and non-NPSLE), and explore their clinical implications.Methods We prospectively recruited 50 patients with SLE (21 non-NPSLE and 29 NPSLE) and 32 age-matched healthy controls (HCs), using MRI diffusion tensor imaging. Individual structural networks were constructed using fibre numbers between brain areas as edge weights. Global metrics (eg, small-worldness, global efficiency) and local network properties (eg, degree centrality, nodal efficiency) were computed. Group comparisons of network characteristics were conducted. Clinical correlations were assessed using partial correlation, and differentiation between non-NPSLE and NPSLE was performed using support vector classification.Results Patients with oth non-NPSLE and NPSLE exhibited significant global and local topological alterations compared with HCs. These changes were more pronounced in NPSLE, particularly affecting the default mode and sensorimotor networks. Topological changes in patients with SLE correlated with lesion burdens and clinical parameters such as disease duration and the systemic lupus international collaborating clinics damage index. The identified topological features enabled accurate differentiation between non-NPSLE and NPSLE with 87% accuracy.Conclusion Structural networks in patients SLE may be altered at both global and local levels, with more pronounced changes observed in NPSLE, notably affecting the default mode and sensorimotor networks. These alterations show promise as biomarkers for clinical diagnosis.

The addition of baked Qingke improves the flavor profile of beer. In this study, beer was brewed using Qingke baked at various temperatures. The beer produced with Qingke baked at 180 °C achieved the highest sensory score (40/50), an alcohol content of 6.92% (v/v), a total phenolic content of 446.42 mg/L, melanoidin concentration of 98.22 g/L, a color value of 10.88 EBC, and exhibited satisfactory antioxidant activity. Analysis of volatile compounds using HS-SPME-GC–MS revealed 48 compounds, of which esters accounted for 63% and alcohols accounted for 27% of the total content. The flavor profile of the beer varied across different baking temperatures. Pyrazines and aldehydes were predominantly present in samples baked at higher temperatures (T3, T4, and T5). Correlation analysis showed that the baking flavor in the beer was primarily correlated with 2, 5-dimethyl-pyrazine, trimethyl-pyrazine, phenylacetaldehyde, and ethyl 9-decenoate (R > 0.9).

Abstract Aim The aim of this study was to modify and test the psychometric properties of the Chinese version of the stressor scale for emergency nurses. Design The methodological design was carried out in two phases: (a) form the Chinese version by Delphi method and (b) test the psychometric properties by cross‐sectional survey. Methods The translated scale was administered to 420 nurses in Qingdao. Validity was assessed in terms of content validity, calibration correlation validity and construct validity using exploratory factor analysis and confirmatory factor analysis. Internal consistency was estimated using Cronbach α coefficients. Results The Chinese version of the SSEN retains 27 items, four common factors were extracted by exploratory factor analysis, and the factor cumulative variance contribution rate was 78.463%. The fitting indexes of the four‐factor model of CFA were all in the acceptable range[χ2 = 711.30, df = 312, p

Abstract Background The recurrence of hepatocellular carcinoma (HCC) after radiofrequency ablation (RFA) remains a major clinical problem. Cells that survive the sublethal heat stress that is induced by incomplete RFA are the main source of HCC relapse. Heat stress has long been reported to increase intracellular reactive oxygen species (ROS) generation. Although ROS can induce apoptosis, a pro-survival effect of ROS has also been demonstrated. However, the role of ROS in HCC cells exposed to sublethal heat stress remains unclear. Methods HepG2 and HuH7 cells were used for this experiment. Insufficient RFA was performed in cells and in a xenograft model. ROS and antioxidant levels were measured. Apoptosis was analyed by Annexin-V/PI staining and flow cytometry. Protein expression was measured using western blotting. Colocalization of lysosomes and mitochondria was analyzed to assess mitophagy. Corresponding activators or inhibitors were applied to verify the function of specific objectives. Results Here,we showed that sublethal heat stress induced a ROS burst, which caused acute oxidative stress. This ROS burst was generated by mitochondria, and it was initiated by upregulated NOX4 expression in the mitochondria. n-acetylcysteine (NAC) decreased HCC cell survival under sublethal heat stress conditions in vivo and in vitro. NOX4 triggers the production of mitochondrial ROS (mtROS), and NOX4 inhibitors or siNOX4 also decreased HCC cell survival under sublethal heat stress conditions in vitro. Increased mtROS trigger PINK1-dependent mitophagy to eliminate the mitochondria that are damaged by sublethal heat stress and to protect cells from apoptosis. Nrf2 expression was elevated in response to this ROS burst and mediated the ROS burst-induced increase in PINK1 expression after sublethal heat stress. Conclusion These data confirmed that the ROS burst that occurs after iRFA exerted a pro-survival effect. NOX4 increased the generation of ROS by mitochondria. This short-term ROS burst induced PINK1-dependent mitophagy to eliminate damaged mitochondria by increasing Nrf2 expression.

Large language models have demonstrated impressive capabilities in many domains. But they sometimes generate irrelevant or nonsensical text, or produce outputs that deviate from the provided input, an occurrence commonly referred to as hallucination. To mitigate this issue, we introduce a novel decoding method that incorporates both factual and hallucination prompts (DFHP). It applies contrastive decoding to highlight the disparity in output probabilities between factual prompts and hallucination prompts. Experiments on both multiple-choice and text generation tasks show that our approach significantly improves factual accuracy of large language models without additional training. On the TruthfulQA dataset, the DFHP method significantly improves factual accuracy of the LLaMA model, with an average improvement of 6.4% for the 7B, 13B, 30B, and 65B versions. Its high accuracy in factuality makes it an ideal choice for high reliability tasks like medical diagnosis and legal cases. [ABSTRACT FROM AUTHOR]

With the increasing utilization of clean energy, the development and utilization of hydrogen energy has become a research topic of great significance. Cobalt selenide (CS) is an electrocatalyst with great potential for oxygen evolution reaction (OER). In this paper, a nitrogen-doped biomass carbon (1NC@3)-based composite cobalt selenide (CS) heterojunction was prepared via a solvothermal method using kelp as the raw material. Structural, morphological, and electrochemical analyses were conducted to evaluate its performance. The electrochemical test results demonstrate that the overpotential of the CS/1NC@3 catalyst in the OER process was 292 mV, with a Tafel slope of 98.71 mV·dec−1 at a current density of 10 mA·cm−2. The electrochemical performance of the CS/1NC@3 catalyst was further confirmed by theoretical calculations, which revealed that the presence of the biomass carbon substrate enhanced the charge transport speed of the OER process and promoted the OER process. This study provides a promising strategy for the development of efficient electrocatalysts for OER applications.

Electrochemical water splitting to produce green hydrogen offers a promising technology for renewable energy conversion and storage, as well as realizing carbon neutrality. The efficiency, stability, and cost of electrocatalysts toward hydrogen evolution reaction (HER) and electrocatalytic overall water splitting (EOWS) at large current densities are essential for practical application. In this review, the key factors that determine the catalytic performance of electrocatalysts at large current densities are summarized from the angel of thermodynamic and kinetic correlation. The corresponding design strategies are presented. The electronic structure and density of active sites that affect the adsorption/desorption of intermediates are considered as the thermodynamic aspects, while charge transfer and mass transport capabilities closely associated with electrode resistance and intermediate diffusion are assigned as kinetic effects. Recent development of bifunctional and integrated electrocatalysts toward EOWS is also discussed in detail. Finally, the perspective and direction on the electrocatalytic water splitting under large current density are proposed. This comprehensive overview will offer profound insights and guidance for the continued advancement of this field. [ABSTRACT FROM AUTHOR]

Purpose: This retrospective study aimed to investigate whether the repair of medial meniscus posterior root tears (MMPRTs) is effective for improving clinical outcomes and return to sports rates in young patients (50 years old or younger) with medial compartment knee osteoarthritis (KOA) and MMPRTs. Methods: Between 2016 and 2019, 153 patients with KOA and MMPRTs who underwent open‐wedge high tibial osteotomy (OWHTO) were retrospectively included. The patients were divided into OWHTO combined with MMPRT repair (n = 73) and isolated OWHTO (n = 80) groups. Lysholm scores, Hospital for Special Surgery (HHS) scores, Tegner scores, flexion contracture, range of knee flexion, return to sports rates and postoperative complications were compared. Radiological outcomes, including hip–knee–ankle angle (HKA), medial proximal tibial angle (MPTA), joint line convergence angle (JLCA) and Kellgren–Lawrence (K‐L) grade, were compared between the two groups. Results: After a mean follow‐up of 30.1 ± 3.0 months, the OWHTO + Repair group observed better clinical outcomes compared with the OWHTO group (Lysholm score: 86.7 ± 7.4 vs. 81.6 ± 6.9, p = 0.023. HHS score: 85.4 ± 8.20 vs. 80.5 ± 7.1, p = 0.039). The OWHTO + Repair group had higher Tegner scores and return to sports rates than the OWHTO group (Tegner score: 6 vs. 5, p = 0.020; return to sports rates: 38% vs. 15%, p = 0.001). No fracture or major complications occurred. Radiological outcomes showed no significant differences between the two groups (HKA: 181.1 ± 2.7 vs. 180.1 ± 1.5 n.s; MPTA: 90.1 ± 1.8 vs. 89.2 ± 1.4, n.s; JLCA:1.9 ± 0.7 vs. 2.1 ± 0.7, n.s). Conclusions: Additional MMPRT repair during OWHTO was associated with better clinical outcomes and higher rates of return to sports in young patients with medial compartment KOA and MMPRTs. Level of Evidence: Level III. [ABSTRACT FROM AUTHOR]

Complex products (CPs) modeling design has a long development cycle and high cost, and it is difficult to accurately meet the needs of enterprises and users. At present, the Kansei Engineering (KE) method based on back-propagated (BP) neural networks is applied to solve the modeling design problem that meets users’ affective preferences for simple products quickly and effectively. However, the modeling feature data of CPs have a wide range of dimensions, long parameter codes, and the characteristics of time series. As a result, it is difficult for BP neural networks to recognize the affective preferences of CPs from an overall visual perception level as humans do. To address the problems above and assist designers with efficient and high-quality design, a CP modeling design method based on Long Short-Term Memory (LSTM) neural network and KE (CP-KEDL) was proposed. Firstly, the improved MA method was carried out to transform the product modeling features into feature codes with sequence characteristics. Secondly, the mapping model between perceptual images and modeling features was established based on the LSTM neural network to predict the evaluation value of the product’s perceptual images. Finally, the optimal feature sets were calculated by a Genetic Algorithm (GA). The experimental results show that the MSE of the LSTM model is only 0.02, whereas the MSE of the traditional Deep Neural Networks (DNN) and Convolutional Neural Networks (CNN) neural network models are 0.30 and 0.23, respectively. The results verified that the proposed method can effectively grapple with the CP modeling design problem with the timing factor, improve design satisfaction and shorten the R&D cycle of CP industrial design.

Metabolic reprogramming is a feature of cancer cells and crucial for tumor growth and metastasis. Interferon-γ (IFNγ) is a cytokine that plays a pivotal role in host antitumor immunity. However, little is known about the roles of metabolic reprogramming in immune responses. Here, we show that colon cancer cells reprogram metabolism to coordinate proper cellular responses to IFNγ by downregulating mitochondrial pyruvate carrier (MPC)1 and 2 via STAT3 signaling. Forced overexpression of MPC promote the production of reactive oxygen species and enhance the apoptosis induced by IFNγ in colon cancer cells. Moreover, inhibiting STAT3 sensitize the antitumor efficacy of IFN-γ against colon cancer cells. Our findings present a previously unrecognized mechanism that colon cancer manipulate to resist IFNγ mediated antitumor immunity that have implications for targeting a unique aspect of this disease. Keywords: Interferon-γ, Mitochondrial pyruvate carriers, Redox, Colon cancer, Glycolysis

Tuberculosis remains a serious challenge to human health worldwide. para-Aminosalicylic acid (PAS) is an important anti-tuberculosis drug, which requires sequential activation by Mycobacterium tuberculosis (M. tuberculosis) dihydropteroate synthase and dihydrofolate synthase (DHFS, FolC). Previous studies showed that loss of function mutations of a thymidylate synthase coding gene thyA caused PAS resistance in M. tuberculosis, but the mechanism is unclear. Here we showed that deleting thyA in M. tuberculosis resulted in increased content of tetrahydrofolate (H4PteGlu) in bacterial cells as they rely on the other thymidylate synthase ThyX to synthesize thymidylate, which produces H4PteGlu during the process. Subsequently, data of in vitro enzymatic activity experiments showed that H4PteGlu hinders PAS activation by competing with hydroxy dihydropteroate (H2PtePAS) for FolC catalysis. Meanwhile, over-expressing folC in ΔthyA strain and a PAS resistant clinical isolate with known thyA mutation partially restored PAS sensitivity, which relieved the competition between H4PteGlu and H2PtePAS. Thus, loss of function mutations in thyA led to increased H4PteGlu content in bacterial cells, which competed with H2PtePAS for catalysis by FolC and hence hindered the activation of PAS, leading to decreased production of hydroxyl dihydrofolate (H2PtePAS-Glu) and finally caused PAS resistance. On the other hand, functional deficiency of thyA in M. tuberculosis pushes the bacterium switch to an unidentified dihydrofolate reductase for H4PteGlu biosynthesis, which might also contribute to the PAS resistance phenotype. Our study revealed how thyA mutations confer PAS resistance in M. tuberculosis and provided new insights into studies on the folate metabolism of the bacterium. [ABSTRACT FROM AUTHOR]

Deep learning requires a large amount of data, and enhanced processing of the data is particularly important. This study aims to investigate the enhancement of the dataset from the perspective of the underlying data distribution and solve the problem of unbalanced data samples. In this study, a novel approach called the single-sample sampling variant autoencoder (S3VAE) was proposed to generate data which were then compared. Experimental results demonstrate that, under the same data discard rate, the data generated by the S3VAE architecture exhibit a test accuracy closer to that of the original data, which proves the ability of the S3VAE architecture to generate results closer to the original data. Furthermore, the reconstruction abilities of C-VAE and S3VAE were compared using two public datasets and conducted five different discard rate experiments. As observed, the test accuracy of S3VAE is higher than that of C-VAE in all cases. With an increase in the data discard rate, the advantage of S3VAE becomes more pronounced. When the data discard rate is 97.5%, the test accuracy of S3VAE is 2.7% higher than that of C-VAE. These results confirm that the method has a significant positive effect on data enhancement and can be effectively used in practical scenarios. Moreover, this method can be extended to most advanced variant autoencoders. [ABSTRACT FROM AUTHOR]