Ophiopogonin D alleviates acute lung injury by regulating inflammation via the STAT3/A20/ASK1 axis.

BACKGROUND: Acute lung injury (ALI) is characterized by acute pulmonary inflammatory infiltration. Alveolar epithelial cells (AECs) release numerous pro-inflammatory cytokines, which result in the pathological changes seen in ALI. Ophiopogonin D (OD), extracted from the roots of Ophiopogon japonicus (Thunb.) Ker Gawl. (Liliaceae), reduces inflammation; however, the efficacy of OD in ALI has not been reported and the underlying molecular mechanisms remain unclear. PURPOSE: This study investigated the anti-inflammatory effects of OD, as well as the underlying mechanisms, in AECs and a mouse ALI model. METHODS: Lipopolysaccharide (LPS) and tumor necrosis factor-α (TNF-α) were used to stimulate macrophages and A549 cells, and a mouse ALI model was established by intratracheal LPS administration. The anti-inflammatory effects and mechanisms of OD in the TNF-α-induced in vitro inflammation model was evaluated using real-time quantitative polymerase chain reaction qPCR), enzyme-linked immunosorbent assay (ELISA), western blotting, nuclear and cytoplasmic protein extraction, and immunofluorescence. The in vivo anti-inflammatory activity of OD was evaluated using hematoxylin and eosin staining, qPCR, ELISA, and western blotting. RESULTS: The bronchoalveolar lavage fluid and lung tissue of LPS-induced ALI mice exhibited increased TNF-α expression. TNF-α induced a significantly greater pro-inflammatory effect in AECs than LPS. OD reduced inflammation and mitogen-activated protein kinase (MAPK) and transcription factor p65 phosphorylation in vivo and in vitro and promoted signal transducer and activator of transcription 3 (STAT3) phosphorylation and A20 expression, thereby inducing apoptosis signal-regulating kinase 1 (ASK1) proteasomal degradation. CONCLUSION: OD exerts an anti-inflammatory effect by promoting STAT3-dependent A20 expression and ASK1 degradation. OD may therefore have therapeutic value in treating ALI and other TNF-α-related inflammatory diseases.

The role of iron overload and ferroptosis in arrhythmia pathogenesis.

Ferroptosis is a newly discovered form of programmed cell death triggered by intracellular iron overload, which leads to the accumulation of lipid peroxides in various cells. It has been implicated in the pathogenesis and progression of various diseases, including tumors, neurological disorders, and cardiovascular diseases. The intricate mechanism underlying ferroptosis involves an imbalance between the oxidation and antioxidant systems, disturbances in iron metabolism, membrane lipid peroxidation, and dysregulation of amino acid metabolism. We highlight the key molecular mechanisms governing iron overload and ferroptosis, and discuss potential molecular pathways linking ferroptosis with arrhythmias.

Results of a Pilot External Quality Assessment Scheme for Genetic Testing of Newborns with Spinal Muscular Atrophy.

BACKGROUND: This study aims to evaluate the ability of laboratories to perform spinal muscular atrophy (SMA) genetic testing in newborns based on dried blood spot (DBS) samples, and to provide reference data and advance preparation for establishing the pilot external quality assessment (EQA) scheme for SMA genetic testing of newborns in China. METHODS: The pilot EQA scheme contents and evaluation principles of this project were designed by National Center for Clinical Laboratories (NCCL), National Health Commission. Two surveys were carried out in 2022, and 5 batches of blood spots were submitted to the participating laboratory each time. All participating laboratories conducted testing upon receiving samples, and test results were submitted to NCCL within the specified date. RESULTS: The return rates were 75.0% (21/28) and 95.2% (20/21) in the first and second surveys, respectively. The total return rate of the two examinations was 83.7% (41/49). Nineteen laboratories (19/21, 90.5%) had a full score passing on the first survey, while in the second survey twenty laboratories (20/20, 100%) scored full. CONCLUSIONS: This pilot EQA survey provides a preliminary understanding of the capability of SMA genetic testing for newborns across laboratories in China. A few laboratories had technical or operational problems in testing. It is, therefore, of importance to strengthen laboratory management and to improve testing capacity for the establishment of a national EQA scheme for newborn SMA genetic testing.

A radiograph-based deep learning model improves radiologists' performance for classification of histological types of primary bone tumors: A multicenter study.

PURPOSE: To develop a deep learning (DL) model for classifying histological types of primary bone tumors (PBTs) using radiographs and evaluate its clinical utility in assisting radiologists. METHODS: This retrospective study included 878 patients with pathologically confirmed PBTs from two centers (638, 77, 80, and 83 for the training, validation, internal test, and external test sets, respectively). We classified PBTs into five categories by histological types: chondrogenic tumors, osteogenic tumors, osteoclastic giant cell-rich tumors, other mesenchymal tumors of bone, or other histological types of PBTs. A DL model combining radiographs and clinical features based on the EfficientNet-B3 was developed for five-category classification. The area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity were calculated to evaluate model performance. The clinical utility of the model was evaluated in an observer study with four radiologists. RESULTS: The combined model achieved a macro average AUC of 0.904/0.873, with an accuracy of 67.5 %/68.7 %, a macro average sensitivity of 66.9 %/57.2 %, and a macro average specificity of 92.1 %/91.6 % on the internal/external test set, respectively. Model-assisted analysis improved accuracy, interpretation time, and confidence for junior (50.6 % vs. 72.3 %, 53.07[s] vs. 18.55[s] and 3.10 vs. 3.73 on a 5-point Likert scale [P < 0.05 for each], respectively) and senior radiologists (68.7 % vs. 75.3 %, 32.50[s] vs. 21.42[s] and 4.19 vs. 4.37 [P < 0.05 for each], respectively). CONCLUSION: The combined DL model effectively classified histological types of PBTs and assisted radiologists in achieving better classification results than their independent visual assessment.

A universal recombinant adenovirus type 5 vector-based COVID-19 vaccine.

A universal recombinant adenovirus type-5 (Ad5) vaccine against COVID19 (Ad-US) was constructed, and immunogenicity and broad-spectrum of Ad5-US were evaluated with both intranasal and intramuscular immunization routes. The humoral immune response of Ad5-US in serum and bronchoalveolar lavage fluid were evaluated by the enzyme-linked immunosorbent assay (ELISA), recombinant vesicular stomatitis virus based pseudovirus neutralization assay, and angiotensin-converting enzyme-2 (ACE2) -binding inhibition assay. The cellular immune response and Th1/Th2 biased immune response of Ad5-US were evaluated by the IFN-γ ELISpot assay, intracellular cytokine staining, and Meso Scale Discovery (MSD) profiling of Th1/Th2 cytokines. Intramuscular priming followed by an intranasal booster with Ad5-US elicited the broad-spectrum and high levels of IgG, IgA, pseudovirus neutralizing antibody (PNAb), and Th1-skewing of the T-cell response. Overall, the adenovirus type-5 vectored universal SARS-CoV-2 vaccine Ad5-US was successfully constructed, and Ad5-US was highly immunogenic and broad spectrum. Intramuscular priming followed by an intranasal booster with Ad5-US induced the high and broad spectrum systemic immune responses and local mucosal immune responses.

Analysis of the value of potential biomarker S100-A8 protein in the diagnosis and pathogenesis of spinal tuberculosis.

Objectives: The objective of this study is to evaluate the value of S100-A8 protein as a diagnostic marker for spinal tuberculosis and to explore its role in the potential pathogenesis of spinal tuberculosis (STB). Methods: The peripheral blood of 100 spinal tuberculosis patients admitted to the General Hospital of Ningxia Medical University from September 2018 to June 2021 were collected as the observation group, and the peripheral blood of 30 healthy medical examiners were collected as the control group. Three samples from the observation group and three samples from the control group were selected for proteomics detection and screening of differential proteins. Kyoto Encyclopedia of Genes (KEGG) was used to enrich and analyze related signaling pathways to confirm the target protein. The serum expression levels of the target proteins were determined and compared between the two groups using enzyme-linked immunosorbent assay (ELISA). Statistical methods were used to evaluate the value of target protein as a diagnostic marker for STB. A macrophage model of Mycobacterium tuberculosis infection was constructed and S100-A8 small interfering RNA was used to investigate the molecular mechanism of the target protein. Results: S100-A8 protein has the value of diagnosing spinal tuberculosis (AUC = 0.931, p < 0.001), and the expression level in the peripheral blood of the observation group (59.04 ± 19.37 ng/mL) was significantly higher than that of the control group (43.16 ± 10.07 ng/mL) (p < 0.05). S100-A8 protein expression showed a significant positive correlation with both CRP and ESR values (p < 0.01). Its AUCs for combined bacteriological detection, T-SPOT results, diagnostic imaging, antacid staining results, and pathological results were 0.705 (p < 0.05), 0.754 (p < 0.01), 0.716 (p < 0.01), 0.656 (p < 0.05), and 0.681 (p < 0.01), respectively. Lack of S100-A8 leads to a significant decrease in the expression levels of TLR4 and IL-17A in infected macrophages. Conclusion: S100-A8 protein is differentially expressed in the peripheral blood of patients with spinal tuberculosis and healthy individuals and may be a novel candidate biomarker for the diagnosis of spinal tuberculosis. The feedback loop on the S100-A8-TLR4-IL-17A axis may play an important role in the inflammatory mechanism of spinal tuberculosis.

Performance Evaluation of Noninvasive Prenatal Testing in Screening Chromosome Disorders: A Single-Center Observational Study of 15,304 Consecutive Cases in China.

Objective: This study was to evaluate the performance of noninvasive prenatal testing (NIPT) in detecting fetal chromosome disorders in pregnant women. Methods: From October 1st, 2017, to December 31th, 2022, a total of 15,304 plasma cell free DNA-NIPT samples were collected for fetal chromosome disorders screening. The results of NIPT were validated by confirmatory invasive testing or clinical outcome follow-up. Further, NIPT performance between low-risk and high-risk groups, as well as singleton pregnancy and twin pregnancy groups was compared. Besides, analysis of 111 false-positive cases was performed. Results: Totally, NIPT was performed on 15,086 eligible venous blood samples, of which 179 (1.19%) showed positive NIPT results and 68 were further validated to be true positive samples via confirmatory invasive testing or follow-up of clinical outcomes. For common chromosome aneuploidies, sex chromosome abnormalities (SCA) and other chromosomal aneuploidies, the detection sensitivities of NIPT were all 100%, the specificities were 99.87%, 99.70%, and 99.68% and the positive predictive values (PPVs) were 65.45%, 31.82%, and 10.91%, respectively. No statistically significant variance in detection performance was observed among 2987 high-risk and 12,099 low-risk subjects, as well as singleton and twin pregnancy subjects. The concentration of cell-free fetal DNA of 111 false-positive cases ranged from 5.5% to 33.7%, which was higher than the minimum requirement of NIPT. Conclusion: With stringent protocol, NIPT shows high sensitivity and specificity for detecting fetal chromosome disorders in a large-scale clinical service, helping improving overall pregnancy management.

Fermenting Distiller's Grains by the Domesticated Microbial Consortium To Release Ferulic Acid.

The microbial consortium FA12 that can release ferulic acid (FA) by fermenting distiller's grains was screened from Daqu. Taibaiella, Comamonadaceae, and Ochrobacum were highly abundant in FA12 by 16S rRNA gene sequencing. In the process of long-term acclimation with distiller's grains as a medium, the biomass of FA12 remained stable, and the pH value of fermentation liquid was also relatively stable. Meanwhile, the activities of cellulase, xylanase, and feruloyl esterase secreted by FA12 were stable in the ranges of 0.2350-0.4470, 0.1917-0.3078, and 0.1103-0.1595 U/mL, respectively, and the release of FA could reach 133.77 µg/g. It is proven that the microbial consortium has good genetic stability. In addition, the structural changes of lignocellulose in distiller's grains before and after fermentation were analyzed by scanning electron microscopy (SEM), X-ray diffraction (XRD), and Fourier transform infrared spectroscopy (FTIR), and the changes of distiller's grains weight and lignocellulose content before and after fermentation were also detected. These results all confirmed that FA12 had the function of degrading distiller's grains. In this study, we explored a method to use microbial communities to release FA from distiller's grains and degrade lignocellulose in the waste, which opened up a new way for the application of the high value of lost waste.

Automatic segmentation of fat metaplasia on sacroiliac joint MRI using deep learning.

OBJECTIVE: To develop a deep learning (DL) model for segmenting fat metaplasia (FM) on sacroiliac joint (SIJ) MRI and further develop a DL model for classifying axial spondyloarthritis (axSpA) and non-axSpA. MATERIALS AND METHODS: This study retrospectively collected 706 patients with FM who underwent SIJ MRI from center 1 (462 axSpA and 186 non-axSpA) and center 2 (37 axSpA and 21 non-axSpA). Patients from center 1 were divided into the training, validation, and internal test sets (n = 455, 64, and 129). Patients from center 2 were used as the external test set. We developed a UNet-based model to segment FM. Based on segmentation results, a classification model was built to distinguish axSpA and non-axSpA. Dice Similarity Coefficients (DSC) and area under the curve (AUC) were used for model evaluation. Radiologists' performance without and with model assistance was compared to assess the clinical utility of the models. RESULTS: Our segmentation model achieved satisfactory DSC of 81.86% ± 1.55% and 85.44% ± 6.09% on the internal cross-validation and external test sets. The classification model yielded AUCs of 0.876 (95% CI: 0.811-0.942) and 0.799 (95% CI: 0.696-0.902) on the internal and external test sets, respectively. With model assistance, segmentation performance was improved for the radiological resident (DSC, 75.70% vs. 82.87%, p < 0.05) and expert radiologist (DSC, 85.03% vs. 85.74%, p > 0.05). CONCLUSIONS: DL is a novel method for automatic and accurate segmentation of FM on SIJ MRI and can effectively increase radiologist's performance, which might assist in improving diagnosis and progression of axSpA. CRITICAL RELEVANCE STATEMENT: DL models allowed automatic and accurate segmentation of FM on sacroiliac joint MRI, which might facilitate quantitative analysis of FM and have the potential to improve diagnosis and prognosis of axSpA. KEY POINTS: • Deep learning was used for automatic segmentation of fat metaplasia on MRI. • UNet-based models achieved automatic and accurate segmentation of fat metaplasia. • Automatic segmentation facilitates quantitative analysis of fat metaplasia to improve diagnosis and prognosis of axial spondyloarthritis.

Transformer fault diagnosis method based on SMOTE and NGO-GBDT.

In order to improve the accuracy of transformer fault diagnosis and improve the influence of unbalanced samples on the low accuracy of model identification caused by insufficient model training, this paper proposes a transformer fault diagnosis method based on SMOTE and NGO-GBDT. Firstly, the Synthetic Minority Over-sampling Technique (SMOTE) was used to expand the minority samples. Secondly, the non-coding ratio method was used to construct multi-dimensional feature parameters, and the Light Gradient Boosting Machine (LightGBM) feature optimization strategy was introduced to screen the optimal feature subset. Finally, Northern Goshawk Optimization (NGO) algorithm was used to optimize the parameters of Gradient Boosting Decision Tree (GBDT), and then the transformer fault diagnosis was realized. The results show that the proposed method can reduce the misjudgment of minority samples. Compared with other integrated models, the proposed method has high fault identification accuracy, low misjudgment rate and stable performance.

Variants in 3p24.3 predicts the risk of early neurological deterioration in large artery atherosclerotic stroke.

The rate of early neurological deterioration (END) differs in different subtypes of ischaemic stroke. Previous studies showed PLCL2 gene is a novel susceptibility locus for the occurrence of atherosclerosis and thrombotic events. The objective of this research is to examine the efficacy that PLCL2 may have on the risk of END in large artery atherosclerotic (LAA) stroke. Tagged single nucleotide polymorphisms (SNPs) were identified by a strategy of fine-mapping. The genotyping of the selected SNPs was performed by SNPscan. The impact of PLCL2 on indicating the susceptibility of END in LAA patients was evaluated by binary logistic regression. The SNP-SNP interactions of PLCL2 for END was assessed by generalized multifactor dimensionality reduction (GMDR). A total of 1527 LAA stroke patients were recruited, 582 patients (38 %) experienced END. Compared to participants without END, participants experienced END were much older (P = 0.018), more likely to suffer pre-existing diabetes mellitus (P = 0.036), higher frequent in active tobacco users (P = 0.022) and had much higher median NIHSS on admission (P < 0.001). Rs4685423 was identified to be a predictor to the risk of END: the frequency of END in AA genotype patients is lower than that in AC or CC genotype patients (multivariate-adjusted, OR 0.63; 95 % CI 0.49-0.80; P < 0.001). The SNP-SNP interactions analysis indicates rs4685423 has the greatest impacton the risk of END for LAA patients. The time from admission diagnosis to END onset in AA genotype patients is much later than that in CA or CC genotype patients (log-rank, P = 0.005). In summary, the PLCL2 rs4685423 SNP is probably associated with the END risk in LAA stroke patients.

Effects of Exercise Habits and Gender on Sports e-Learning Behavior: Evidence from an Eye-Tracking Study.

Background/Objective: In the post-epidemic era, an increasing number of individuals were accustomed to learning sports and physical activity knowledge online for fitness and health demands. However, most previous studies have examined the influence of e-learning materials and resources on learners and have neglected intrinsic factors such as experience and physiological characteristics. Therefore, we conducted a study to investigate the effect of exercise habits and gender on sports e-learning behavior via eye-tracking technology. Methods: We recruited a sample of 60 undergraduate students (mean age = 19.6) from a university in Nanjing, China. They were randomly assigned into 4 groups based on 2 genders × 2 exercise habits. Their gaze behavior was collected by an eye-tracking device during the experiment. The cognitive Load Test and Learning Effect Test were conducted at the end of the individual experiment. Results: (1) Compared to the non-exercise habit group, the exercise habit group had a higher fixation count (P<0.05), a shorter average fixation duration (P<0.05), a smaller average pupil diameter (P<0.05), and a lower subjective cognitive load (P<0.05) and better learning outcome (P<0.05). (2) Male participants showed a greater tendency to process information from the video area of interest (AOIs), and had lower subjective cognitive load (P < 0.05) and better learning outcomes (P < 0.05). (3) There was no interaction effect between exercise habits and gender for any of the indicators (P > 0.05). Conclusion: Our results indicate that exercise habits effectively enhance sports e-learning outcomes and reduce cognitive load. The exercise habits group showed significant improvements in fixation counts, average fixation duration, and average pupil diameter. Furthermore, male subjects exhibited superior learning outcomes, experienced lower cognitive load, and demonstrated greater attentiveness to dynamic visual information. These conclusions are expected to improve sports e-learning success and address educational inequality.

JAK/STAT3 signaling in cardiac fibrosis: a promising therapeutic target.

Cardiac fibrosis is a serious health problem because it is a common pathological change in almost all forms of cardiovascular diseases. Cardiac fibrosis is characterized by the transdifferentiation of cardiac fibroblasts (CFs) into cardiac myofibroblasts and the excessive deposition of extracellular matrix (ECM) components produced by activated myofibroblasts, which leads to fibrotic scar formation and subsequent cardiac dysfunction. However, there are currently few effective therapeutic strategies protecting against fibrogenesis. This lack is largely because the molecular mechanisms of cardiac fibrosis remain unclear despite extensive research. The Janus kinase/signal transducer and activator of transcription (JAK/STAT) signaling cascade is an extensively present intracellular signal transduction pathway and can regulate a wide range of biological processes, including cell proliferation, migration, differentiation, apoptosis, and immune response. Various upstream mediators such as cytokines, growth factors and hormones can initiate signal transmission via this pathway and play corresponding regulatory roles. STAT3 is a crucial player of the JAK/STAT pathway and its activation is related to inflammation, malignant tumors and autoimmune illnesses. Recently, the JAK/STAT3 signaling has been in the spotlight for its role in the occurrence and development of cardiac fibrosis and its activation can promote the proliferation and activation of CFs and the production of ECM proteins, thus leading to cardiac fibrosis. In this manuscript, we discuss the structure, transactivation and regulation of the JAK/STAT3 signaling pathway and review recent progress on the role of this pathway in cardiac fibrosis. Moreover, we summarize the current challenges and opportunities of targeting the JAK/STAT3 signaling for the treatment of fibrosis. In summary, the information presented in this article is critical for comprehending the role of the JAK/STAT3 pathway in cardiac fibrosis, and will also contribute to future research aimed at the development of effective anti-fibrotic therapeutic strategies targeting the JAK/STAT3 signaling.

Risk factors for electrical storms following percutaneous coronary intervention in patients with acute myocardial infarction: A meta-analysis.

Electrical storms (ESs) following percutaneous coronary intervention (PCI) in acute myocardial infarction (AMI) patients pose a significant challenge, affecting prognostic outcomes and increasing mortality. This meta-analysis synthesized data from 11 studies involving 9,666 AMI patients to identify risk factors associated with ES following PCI. Our findings revealed an average ES incidence of 7.70%, with identified risk factors including low thrombolysis in myocardial infarction (TIMI) flow grades (0-1), elevated cardiac troponin I levels, persistent hypotension, reperfusion arrhythmias, the right coronary artery being the infarct-related artery, increased diameter of the infarct-related artery, renal dysfunction, elevated creatine kinase-MB, and bradycardia. Notably, the use of ß-blockers was found to significantly reduce the risk of ES. The study underscores the importance of early identification and management of these risk factors in AMI patients undergoing PCI to prevent the occurrence of ES, highlighting the protective role of ß-blockers. This research provides a foundation for future strategies aimed at reducing the incidence and improving the prognosis of ES in this patient population.

A deep learning fusion network trained with clinical and high-frequency ultrasound images in the multi-classification of skin diseases in comparison with dermatologists: a prospective and multicenter study.

Background: Clinical appearance and high-frequency ultrasound (HFUS) are indispensable for diagnosing skin diseases by providing internal and external information. However, their complex combination brings challenges for primary care physicians and dermatologists. Thus, we developed a deep multimodal fusion network (DMFN) model combining analysis of clinical close-up and HFUS images for binary and multiclass classification in skin diseases. Methods: Between Jan 10, 2017, and Dec 31, 2020, the DMFN model was trained and validated using 1269 close-ups and 11,852 HFUS images from 1351 skin lesions. The monomodal convolutional neural network (CNN) model was trained and validated with the same close-up images for comparison. Subsequently, we did a prospective and multicenter study in China. Both CNN models were tested prospectively on 422 cases from 4 hospitals and compared with the results from human raters (general practitioners, general dermatologists, and dermatologists specialized in HFUS). The performance of binary classification (benign vs. malignant) and multiclass classification (the specific diagnoses of 17 types of skin diseases) measured by the area under the receiver operating characteristic curve (AUC) were evaluated. This study is registered with www.chictr.org.cn (ChiCTR2300074765). Findings: The performance of the DMFN model (AUC, 0.876) was superior to that of the monomodal CNN model (AUC, 0.697) in the binary classification (P = 0.0063), which was also better than that of the general practitioner (AUC, 0.651, P = 0.0025) and general dermatologists (AUC, 0.838; P = 0.0038). By integrating close-up and HFUS images, the DMFN model attained an almost identical performance in comparison to dermatologists (AUC, 0.876 vs. AUC, 0.891; P = 0.0080). For the multiclass classification, the DMFN model (AUC, 0.707) exhibited superior prediction performance compared with general dermatologists (AUC, 0.514; P = 0.0043) and dermatologists specialized in HFUS (AUC, 0.640; P = 0.0083), respectively. Compared to dermatologists specialized in HFUS, the DMFN model showed better or comparable performance in diagnosing 9 of the 17 skin diseases. Interpretation: The DMFN model combining analysis of clinical close-up and HFUS images exhibited satisfactory performance in the binary and multiclass classification compared with the dermatologists. It may be a valuable tool for general dermatologists and primary care providers. Funding: This work was supported in part by the National Natural Science Foundation of China and the Clinical research project of Shanghai Skin Disease Hospital.

Diosgenin biosynthesis pathway and its regulation in Dioscorea cirrhosa L.

Dioscorea cirrhosa L. (D. cirrhosa) tuber is a traditional medicinal plant that is abundant in various pharmacological substances. Although diosgenin is commonly found in many Dioscoreaceae plants, its presence in D. cirrhosa remained uncertain. To address this, HPLC-MS/MS analysis was conducted and 13 diosgenin metabolites were identified in D. cirrhosa tuber. Furthermore, we utilized transcriptome data to identify 21 key enzymes and 43 unigenes that are involved in diosgenin biosynthesis, leading to a proposed pathway for diosgenin biosynthesis in D. cirrhosa. A total of 3,365 unigenes belonging to 82 transcription factor (TF) families were annotated, including MYB, AP2/ERF, bZIP, bHLH, WRKY, NAC, C2H2, C3H, SNF2 and Aux/IAA. Correlation analysis revealed that 22 TFs are strongly associated with diosgenin biosynthesis genes (-r2- > 0.9, P < 0.05). Moreover, our analysis of the CYP450 gene family identified 206 CYP450 genes (CYP450s), with 40 being potential CYP450s. Gene phylogenetic analysis revealed that these CYP450s were associated with sterol C-22 hydroxylase, sterol-14-demethylase and amyrin oxidase in diosgenin biosynthesis. Our findings lay a foundation for future genetic engineering studies aimed at improving the biosynthesis of diosgenin compounds in plants.

Automatic detection, segmentation, and classification of primary bone tumors and bone infections using an ensemble multi-task deep learning framework on multi-parametric MRIs: a multi-center study.

OBJECTIVES: To develop an ensemble multi-task deep learning (DL) framework for automatic and simultaneous detection, segmentation, and classification of primary bone tumors (PBTs) and bone infections based on multi-parametric MRI from multi-center. METHODS: This retrospective study divided 749 patients with PBTs or bone infections from two hospitals into a training set (N = 557), an internal validation set (N = 139), and an external validation set (N = 53). The ensemble framework was constructed using T1-weighted image (T1WI), T2-weighted image (T2WI), and clinical characteristics for binary (PBTs/bone infections) and three-category (benign/intermediate/malignant PBTs) classification. The detection and segmentation performances were evaluated using Intersection over Union (IoU) and Dice score. The classification performance was evaluated using the receiver operating characteristic (ROC) curve and compared with radiologist interpretations. RESULT: On the external validation set, the single T1WI-based and T2WI-based multi-task models obtained IoUs of 0.71 ± 0.25/0.65 ± 0.30 for detection and Dice scores of 0.75 ± 0.26/0.70 ± 0.33 for segmentation. The framework achieved AUCs of 0.959 (95%CI, 0.955-1.000)/0.900 (95%CI, 0.773-0.100) and accuracies of 90.6% (95%CI, 79.7-95.9%)/78.3% (95%CI, 58.1-90.3%) for the binary/three-category classification. Meanwhile, for the three-category classification, the performance of the framework was superior to that of three junior radiologists (accuracy: 65.2%, 69.6%, and 69.6%, respectively) and comparable to that of two senior radiologists (accuracy: 78.3% and 78.3%). CONCLUSION: The MRI-based ensemble multi-task framework shows promising performance in automatically and simultaneously detecting, segmenting, and classifying PBTs and bone infections, which was preferable to junior radiologists. CLINICAL RELEVANCE STATEMENT: Compared with junior radiologists, the ensemble multi-task deep learning framework effectively improves differential diagnosis for patients with primary bone tumors or bone infections. This finding may help physicians make treatment decisions and enable timely treatment of patients. KEY POINTS: • The ensemble framework fusing multi-parametric MRI and clinical characteristics effectively improves the classification ability of single-modality models. • The ensemble multi-task deep learning framework performed well in detecting, segmenting, and classifying primary bone tumors and bone infections. • The ensemble framework achieves an optimal classification performance superior to junior radiologists' interpretations, assisting the clinical differential diagnosis of primary bone tumors and bone infections.