Forest fire susceptibility assessment under small sample scenario: A semi-supervised learning approach using transductive support vector machine.

Forest fires threaten global ecosystems, socio-economic structures, and public safety. Accurately assessing forest fire susceptibility is critical for effective environmental management. Supervised learning methods dominate this assessment, relying on a substantial dataset of forest fire occurrences for model training. However, obtaining precise forest fire location data remains challenging. To address this issue, semi-supervised learning emerges as a viable solution, leveraging both a limited set of collected samples and unlabeled data containing environmental factors for training. Our study employed the transductive support vector machine (TSVM), a key semi-supervised learning method, to assess forest fire susceptibility in scenarios with limited samples. We conducted a comparative analysis, evaluating its performance against widely used supervised learning methods. The assessment area for forest fire susceptibility lies in Dayu County, Jiangxi Province, China, renowned for its vast forest cover and frequent fire incidents. We analyzed and generated maps depicting forest fire susceptibility, evaluating prediction accuracies for both supervised and semi-supervised learning methods across various small sample scenarios (e.g., 4, 8, 12, 16, 20, 24, 28, and 32 samples). Our findings indicate that TSVM exhibits superior prediction accuracy compared to supervised learning with limited samples, yielding more plausible forest fire susceptibility maps. For instance, at sample sizes of 4, 16, and 28, TSVM achieves prediction accuracies of approximately 0.8037, 0.9257, and 0.9583, respectively. In contrast, random forests, the top performers in supervised learning, demonstrate accuracies of approximately 0.7424, 0.8916, and 0.9431, respectively, for the same small sample sizes. Additionally, we discussed three key aspects: TSVM parameter configuration, the impact of unlabeled sample size, and performance within typical sample sizes. Our findings support semi-supervised learning as a promising approach compared to supervised learning for forest fire susceptibility assessment and mapping, particularly in scenarios with small sample sizes.

Compilation of a self-management assessment scale for postoperative patients with aortic dissection.

Objective: The aim of this research was to compile a self-management assessment scale for patients with aortic dissection (AD). The questionnaire is useful in making the patient aware of the need for post-operative care in order to contribute to improving the outcome and quality of life. Methods: The initial version of the "postoperative self-management assessment scale for patients with aortic dissection" was developed using the Delphi expert consultation method based on qualitative research results, consultation of self-management-related literature, reference to the existing self-management scale, and self-efficacy theory, combined with the disease characteristics of AD. By using the convenience sampling method, a total of 201 patients with AD who had undergone surgery were selected as the research participants. The initial version of the scale was used for follow-up investigation, and the scale entries were evaluated and exploratory factor analysis carried out to form the formal version of the "postoperative self-management assessment scale for patients with aortic dissection." A total of 214 patients with AD after surgery were selected as the research participants. The formal version of the scale was used for follow-up investigation, and its reliability and validity were evaluated. Results: The formal version of the scale had 6 dimensions and 35 entries. The Cronbach's α coefficient for the total scale was 0.908, the split-half reliability was 0.790, and the test-retest reliability after 2 weeks was 0.471. The content validity index of the total scale was 0.963. Exploratory factor analysis yielded six common factors, and the cumulative contribution rate of variance was 66.303%. Confirmatory factor analysis showed that except for the incremental fit index, Tucker-Lewis index, and comparative fit index >0.85, slightly lower than 0.90, χ 2/df <3, root mean square of approximation <0.08, parsimonious goodness-of-fit index, and parsimonious normed fit index >0.50; all other model fitting requirements were satisfied, indicating that the model fitting was acceptable. Conclusion: We compiled the postoperative self-management assessment scale for patients with AD, which has demonstrated excellent reliability and validity and can be used as a tool to evaluate the postoperative self-management level in patients with aortic dissection.

Germline assessment for alloHSCT candidates over 50 years: A 'Fast-Track' screening in myeloid neoplasms.

Patients aged 50 or above diagnosed with myeloid neoplasms (MNs) are typically not candidates for germline testing. However, approximately 8% carry pathogenic germline variants. Allogeneic haematopoietic stem cell transplantation (alloHSCT) remains an option for those aged over 50; neglecting germline testing could mask the risk for relative donor cell-derived MN. We propose a germline-augmented somatic panel (GASP), combining MN predisposition genes with a myeloid somatic panel for timely germline variant identification when initial testing is not indicated. Out of our 133 whole-exome-sequenced MN cases aged over 50 years, 9% had pathogenic/likely variants. GASP detected 92%, compared to 50% with somatic-only panel. Our study highlights the relevance of germline screening in MN, particularly for alloHSCT candidates without established germline-testing recommendations.

Advancing nutrition risk assessment in middle-aged and older individuals with diverse food cultures: A data-driven personalized approach to predict incident hypertension, diabetes and mortality.

BACKGROUND: Personalized nutrition risk assessment is crucial in addressing the association between healthy dietary habits across the life course and the development of disease, functional capacity, and healthy aging, as specific dietary pattern recommendations may not be suitable for diverse food cultures. OBJECTIVE: To develop a data-driven, personalized nutrition risk assessment algorithm linked to incident hypertension, diabetes, and all-cause mortality utilizing the food frequency questionnaire among middle-aged and older individuals. METHODS: A retrospective, population-based cohort study conducted between 1999 and 2015 utilized the nationally representative Taiwan Longitudinal Study on Aging (TLSA) survey to examine personalized dietary risk clusters and their associations with health outcomes. Latent class analysis was performed to derive the dietary diversity clusters among community-dwelling middle-aged and older individuals. Outcomes were defined as new-onset hypertension, diabetes mellitus and all-cause mortality at 4-, 8-, 12- and 16-year follow-ups. RESULTS: Data from 1,811 participants (58.14% males, 43.90% aged 50-64 years) showed that around one-third of participants reported being illiterate, 21.98% widowed, and 51.46% engaging in regular physical exercise. Four dietary diversity clusters were identified: "least diverse", "fish and meat", "dairy, fruit, and vegetable", and "most diverse". The "most diverse" cluster was characterized by a high consumption of protein-rich foods, while the "dairy, fruit, and vegetable" cluster had the highest consumption of dairy products and beans/legumes. The "least diverse" cluster had the lowest intake of protein-rich foods, and dark-colored vegetables and fruits. The "most diverse" cluster had a significantly lower risk of hypertension development at the 4-year (aOR 0.58; p < 0.02) and 8-year (aOR 0.57; p < 0.01) follow-up and diabetes at the 4-year (aOR 0.44; p < 0.03) follow-up. Participants in the "most diverse" clusters exhibited lower risks of 8-year, 12-year, and 16-year mortality than those in the "least diverse" cluster (aOR 0.67, p < 0.05; 0.67, p < 0.03; and 0.50, p < 0.01, respectively). CONCLUSION: The personalized nutrition risk assessment algorithm from the food frequency questionnaire can effectively stratify personal health risks among diverse middle-aged and older individuals, making it a valuable tool in lifestyle modification and intervention studies.

Implementing a continuous quality-improvement framework for tuberculosis infection prevention and control in healthcare facilities in China, 2017-2019.

BACKGROUND: Tuberculosis (TB) infection prevention and control (IPC) in healthcare facilities is key to reducing transmission risk. A framework for systematically improving TB IPC through training and mentorship was implemented in 9 healthcare facilities in China from 2017 to 2019. METHODS: Facilities conducted standardized TB IPC assessments at baseline and quarterly thereafter for 18 months. Facility-based performance was assessed using quantifiable indicators for IPC core components and administrative, environmental, and respiratory protection controls, and as a composite of all control types We calculated the percentage changes in scores over time and differences by IPC control type and facility characteristics. RESULTS: Scores for IPC core components increased by 72% during follow-up when averaged across facilities. The percentage changes for administrative, environmental, and respiratory protection controls were 39%, 46%, and 30%, respectively. Composite scores were 45% higher after the intervention. Overall, scores increased most during the first 6 months. There was no association between IPC implementation and provincial economic development or volume of TB services. CONCLUSIONS: TB IPC policies and practices showed most improvement early during implementation and did not differ consistently by facility characteristics. The training component of the project helped increase the capacity of healthcare professionals to manage TB transmission risks. Lessons learned here will inform national TB IPC guidance.

MiniMounter: A low-cost miniaturized microscopy development toolkit for image quality control and enhancement.

Head-mounted miniaturized fluorescence microscopy (Miniscope) has emerged as a significant tool in neuroscience, particularly for behavioral studies in awake rodents. However, the challenges of image quality control and standardization persist for both Miniscope users and developers. In this study, we propose a cost-effective and comprehensive toolkit named MiniMounter. This toolkit comprises a hardware platform that offers customized grippers and four-degree-of-freedom adjustment for Miniscope, along with software that integrates displacement control, image quality evaluation, and enhancement of 3D visualization. Our toolkit makes it feasible to accurately characterize Miniscope. Furthermore, MiniMounter enables auto-focusing and 3D imaging for Miniscope prototypes that possess solely a 2D imaging function, as demonstrated in phantom and animal experiments. Overall, the implementation of MiniMounter effectively enhances image quality, reduces the time required for experimental operations and image evaluation, and consequently accelerates the development and research cycle for both users and developers within the Miniscope community.

Relationship Between Freshmen's Psychological Health and Family Economic Status in Chinese Universities: A Latent Profile Analysis.

Purpose: With the rapid social changes, psychological health problems among freshmen in universities have become increasingly serious. Many researchers have studied the psychological health status of college freshmen using quantitative methods. However, most researchers have studied the psychological health of college freshmen by treating variables as the central concept and analysis unit, which has limitations on the variable-centered research method. Therefore, this study aims to explore the impact of family economic status and demographic variables on the psychological health of college freshmen, as well as the potential types of psychological health among college freshmen. Methods: Based on the SCL-90 and the self-evaluation of the family economic status of college freshmen, latent profile analysis (LPA) was used to analyze the psychological health of 1497 college freshmen selected through a whole-group sampling method. Freshmen completed the questionnaire from October 18 to October 20, 2022. Then, multinomial logistic regression analysis and variance analysis were conducted using SPSS 25.0 to analyze the impact of family economic status and demographic variables on different psychological health subtypes of college freshmen from a university in Eastern China. Results: There is obvious heterogeneity in the psychological health of college freshmen, which can be divided into three groups: psychological health group (69.54%), psychological distress group (24.65%), and psychological risk group (5.81%). The latent categories of psychological health among college freshmen differ by gender and family economic status. The psychological health status of female students is significantly better than that of male students. In addition, the lower the self-evaluation of family economic status of college freshmen, the lower their psychological health level. Conclusion: Three categories of mental health exist for college freshmen, with significant group heterogeneity. Gender and family economic status significantly influenced mental health outcomes with female students faring better. The study's significance lies in providing useful recommendations for educational administrators and developing targeted interventions for college freshmen. Given the practical realities of mental health education efforts, personalized strategies that utilize peer support, counseling, and psychological services can be adopted to assist freshmen in acclimating to university life and maintaining mental wellness. The findings suggest considering the role of family economic status and gender in promoting mental health and the need for further research to develop comprehensive and effective interventions.

Evaluation on coupling coordinated development of population economy and eco-geological environment in the twin-city economic circle of Chengdu-Chongqing region.

Using the CRITIC method, comprehensive index evaluation method, VAR model, coupling coordination model and other methods, this paper evaluates the comprehensive development of the population economy and eco-geological environment composite system in the twin-city economic circle of Chengdu-Chongqing region from 2000 to 2020, verifies the dynamic coupling relationship between subsystems and measures the coupling coordination degree of complex system. Meanwhile, the differences in the development process of each subsystem are discussed and the spatial-temporal evolution characteristics and laws of coupling coordination of the composite system during the research period are analyzed. The results reveal the following: (1) There is a long-term coordinated relationship between population, economy and eco-geological environment in Chengdu-Chongqing region, which have the conditions for emergence and generation. (2) The subsystems of population, economy and eco-geological environment in Chengdu-Chongqing region show an overall upward trend, among which the Sichuan part obviously outperforms the Chongqing part. Besides, the growth rate of the economy subsystem is significantly higher than that of the population, eco-geological environment subsystem. (3) The coupling coordinated development of the composite system has shown a benign upward development trend, gradually changing from "composite spiral structure" to "two cores outstanding, peripheral collapse, the west superior to the east", while the main coordination state has developed from "Basic synergy-Economic lag" in 2000 to "Basic synergy-Population lag" in 2022. In addition, Cheng-De-Mian and Yuzhong districts with the better level of economic development have shown "Advanced synergy-Ecological lag".

Assessment of the structure-activity relationship and antileukemic activity of diacylpyramide compounds as human ClpP agonists.

Human caseinolytic protease P (ClpP) is required for the regulatory hydrolysis of mitochondrial proteins. Allosteric ClpP agonists dysfunctionally activate mitochondrial ClpP in antileukemic therapies. We previously developed ZG111, a potent ClpP agonist derived from ICG-001, inhibits the proliferation of pancreatic ductal adenocarcinoma cell lines in vitro and in vivo by degrading respiratory chain complex proteins. Herein, we studied the structure-activity relationships of ICG-001 analogs as antileukemia agents. Compound ZG36 exhibited improved stabilization effects on the thermal stability of ClpP in acute myeloid leukemia (AML) cell lines compared with the stabilization effects of ZG111, indicating a direct binding between ZG36 and ClpP. Indeed, the resolved ZG36/ClpP structural complex reveals the mode of action of ZG36 during ClpP binding. Compound ZG36 nonselectively degrades respiratory chain complexes and decreases the mitochondrial DNA, eventually leading to the collapse of mitochondrial function and leukemic cell death. Finally, ZG36 treatment inhibited 3-D cell growth in vitro and suppressed the tumorigenesis of AML cells in xenografted mice models. Collectively, we developed a new class of human ClpP agonists that can be used as potential antileukemic therapies.

Assessment of the relative benefits of monotherapy and combination therapy approaches to the treatment of hospital-acquired Stenotrophomonas maltophilia pneumonia: a multicenter, observational, real-world study.

PURPOSE: Stenotrophomonas maltophilia is a Gram-negative pathogen that most commonly causes hospital-acquired infections that can be extremely challenging to treat, contributing to underrecognized mortality throughout the world. The relative benefits of monotherapy as compared to combination therapy in patients diagnosed with S. maltophilia pneumonia, however, have yet to be established. METHODS: Data from 307 patients diagnosed with S. maltophilia hospital-acquired pneumonia (HAP) across four Chinese teaching hospitals from 2016 to 2022 were retrospectively analyzed. RESULTS: Of the analyzed patients, 55.7% (171/307) were administered combination definitive therapy, with a 30-day all-cause mortality rate of 41.0% (126/307). A propensity score weighting analysis revealed that compared with monotherapy, combination definitive therapy was associated with a comparable 30-day mortality risk in the overall patient cohort (OR 1.124, 95% CI 0.707-1.786, P = 0.622), immunocompetent patients (OR 1.349, 95% CI 0.712-2.554, P = 0.359), and patients with APACHE II scores < 15 (OR 2.357, 95% CI 0.820-6.677, P = 0.111), whereas it was associated with a decreased risk of death in immunocompromised patients (OR 0.404, 95% CI .170-0.962, P = 0.041) and individuals with APACHE II scores ≥ 15 (OR 0.494, 95% CI 0.256-0.951, P = 0.035). CONCLUSION: The present data suggest that when treating S. maltophilia-HAP, immunocompromised patients and individuals with APACHE II scores ≥ 15 may potentially benefit from combination therapy.

Assessment of ferroptosis-associated gene signatures as potential biomarkers for differentiating latent from active tuberculosis in children.

Ferroptotic cell death is a regulated process that is governed by iron-dependent membrane lipid peroxide accumulation that plays a pathogenic role in several disease-related settings. The use of ferroptosis-related genes (FRGs) to distinguish active tuberculosis (ATB) from latent tuberculosis infection (LTBI) among children, however, remains to be analysed. Tuberculosis-related gene expression data and FRG lists were obtained, respectively, from Gene Expression Omnibus (GEO) and FerrDb. Differentially expressed FRGs (DE-FRGs) detected when comparing samples from paediatric ATB and LTBI patients were explored using appropriate bioinformatics techniques, after which enrichment analyses were performed for these genes and hub genes were identified, with these genes then being used to explore potential drug interactions and construct competing endogenous RNA (ceRNA) networks. The GSE39939 dataset yielded 124 DE-FRGs that were primarily related to responses to oxidative, chemical and extracellular stimulus-associated stress. In total, the LASSO and SVM-RFE algorithms enabled the identification of nine hub genes (MAPK14, EGLN2, IDO1, USP11, SCD, CBS, PARP8, PARP16, CDC25A) that exhibited good diagnostic utility. Functional enrichment analyses of these genes suggested that they may govern ATB transition from LTBI through the control of many pathways, including the immune response, DNA repair, transcription, RNA degradation, and glycan and energy metabolism pathways. The CIBERSORT algorithm suggested that these genes were positively correlated with inflammatory and myeloid cell activity while being negatively correlated with the activity of lymphocytes. A total of 50 candidate drugs targeting 6 hub DE-FRGs were also identified, and a ceRNA network was used to explore the complex interplay among these hub genes. The nine hub FRGs defined in this study may serve as valuable biomarkers differentiating between ATB and LTBI in young patients.

Exploring the impact mechanism of low-carbon multivariate coupling system in Chinese typical cities based on machine learning.

Low-carbon city construction is one of the key issues that must be addressed for China to achieve high-quality economic development and meet the Sustainable Development Goals. This study creates a comprehensive evaluation index system of low-carbon city multivariate system based on carbon emission data from 30 typical Chinese cities from 2006 to 2017 and evaluates and analyzes the trend of city low-carbon levels using the CRITIC-TOPSIS technique and MK method. Meanwhile, the influence mechanism of the multi-coupled system is investigated using the coupling coordination degree model and random forest algorithm.The results show that there are 8 cities with a significant increasing trend of low-carbon level, 19 cities with no significant monotonic change trend, and 3 cities with a decreasing trend of low-carbon level. By analyzing the coupling coordination degree, we found that the coupling coordination degree between low-carbon level and economic development in most cities tends to increase year by year, from the initial antagonistic effect to a good coordination development trend, which confirms the "inverted U-shaped" relationship between economy and carbon emission. In addition, industrial pollutant emissions, foreign direct investment, and economic output are the core drivers of low-carbon levels in cities.

Health care utilization and potentially preventable adverse outcomes of high-need, high-cost middle-aged and older adults: Needs for integrated care models with life-course approach.

PURPOSE OF THE RESEARCH: The success of modern health care increases life expectancy and prolongs the days of having multimorbidity and functional limitations; the so-defined "high need, high cost (HNHC)" state represents the extreme scenarios of care burden and complexity. This study aims to explore health care utilization and the risk of preventable adverse outcomes stratified by age and HNHC state. MATERIALS AND METHODS: We conducted a retrospective cohort study using the National Health Insurance (NHI) database. People aged ≥40 years were included and further stratified by age (middle-aged: 40-64 and older adults: 65) and HNHC state (top 10% of spending). Health care utilization and drug consumption across different groups were obtained. The multimorbidity frailty index (mFI) was developed for further analysis. Cox regression models were used to examine the associations between HNHC and adverse clinical outcomes (preventable hospitalizations, preventable emergency department visits, and mortality). RESULTS: HNHC participants were older, had a higher mFI and drug consumption, and had higher health care utilization. Compared with non-HNHC participants, HNHC participants exhibited a 4.4-fold and 2.4-fold higher risk of preventable hospitalizations in middle-aged (HR=4.41; 95% CI, 4.17-4.65, p<0.01) and older adults (HR=2.44; 95% CI, 2.34-2.55, p<0.01). Similar risks were observed for preventable emergency department visits and mortality (all p<0.01). CONCLUSIONS: The HNHC state substantially increased health care utilization, polypharmacy, and potentially preventable adverse outcomes after adjustment for frailty. Intervention studies developing integrated care models using the life-course approach are needed to improve the quality of health care systems in super-aged societies.

Cost-Effective H2 O2 -Regeneration of Powdered Activated Carbon by Isolated Fe Sites.

The reuse of powdered activated carbon (PAC) vitally determines the economics and security of the PAC-based adsorption process, while state-of-the-art PAC regeneration technologies are usually unsatisfactory. Here, it is demonstrated that isolated Fe sites anchored on commercial PAC enable fast H2 O2 activation to produce Fe-based reactive oxygen species for highly efficient PAC regeneration at room temperature. Taking rhodamine B as a representative pollutant, PAC decorated with isolated Fe sites realize H2 O2 based regeneration with negligible adsorption capacity degradation for 10 cycles. Moreover, in terms of the PAC loss rate, this technology is greatly superior to traditional Fenton-based regeneration technology. Further operando experiments and theoretical calculations reveal that the high regeneration performance can be attributed to the isolated HOFeO motifs, which activate H2 O2 via a nonradical reaction pathway. These findings provide a very promising strategy toward reducing the cost of H2 O2 -based PAC regeneration technology.

Non-invasive Assessment of Liver Fibrosis Regression in Patients with Chronic Hepatitis B: A Retrospective Cohort Study.

INTRODUCTION: Non-invasive tests (NITs) have been alternative methods of liver biopsy for the cross-sectional assessment of liver fibrosis in patients with chronic hepatitis B (CHB). However, there are limited data on the longitudinal association between NITs and histological changes of liver fibrosis. This study aimed to evaluate whether NITs can be used to assess liver fibrosis regression (LFR) during anti-HBV treatment. METHODS: This retrospective study included 337 patients with CHB who underwent contemporaneous NITs, such as liver stiffness measurement (LSM), the aspartate aminotransferase to platelet ratio index (APRI), the fibrosis index based on four factors (FIB-4), and the γ-glutamyl transpeptidase to platelet ratio (GPR), and liver biopsy at baseline and followed by a repeated liver biopsy and NITs assessment. The LFR was defined as fibrosis regression by at least one stage assessed by METAVIR scoring system. RESULTS: The median interval between the two paired liver biopsy assessment was 31 months (IQR 24-45). At the first liver biopsy, the fibrosis stage was F2 in 159 (47.2%), F3 in 68 (20.2%), and F4 in 110 (32.6%) patients. At the second liver biopsy, the number of patients with fibrosis stages F0-1, F2, F3, and F4 was 102 (30.3%), 106 (31.5%), 63 (18.7%), and 66 (19.6%), respectively. At follow-up liver biopsy, 169 patients (50.1%) had LFR, 128 patients (38.0%) had no change in fibrosis stage, and 40 patients (11.9%) had liver fibrosis progression on histology. A decrease in liver stiffness measurement (LSM) by 25% is the optimal cutoff for predicting LFR. Patients with a 25% or larger decrease in LSM value had more LFR than those with a less than 25% decrease in LSM value (78.1% vs 22.9%, p < 0.001). CONCLUSION: LSM might be used to monitor regression of liver fibrosis during antiviral treatment using nucleos(t)ide analogues (NUCs) in patients with CHB.

GTKO rabbit: A novel animal model for preclinical assessment of decellularized xenogeneic grafts via in situ implantation.

Wild type (WT) animals cannot be used to objectively assess the immunogenicity of animal tissue-derived biomaterials when used as recipients due to difference with human in α-Gal expression. The purpose of this study is to compare the differences of immunological responses between the GGTA1 gene-knockout (GTKO) rabbits and WT rabbits after implantation with animal tissue-derived biomaterials. The porcine-derived decellularized bone matrix (natural bone material, NBM) and fresh porcine cancellous bone (PCB) were implanted in GTKO rabbits and WT rabbits, respectively, and sham operation was used as control (Con). At 2- and 6-week post-implantation, the related immunological items including antibody levels, serum-mediated cell lysis, cytokines, lymphocyte subtypes, and histopathological changes were assessed. GTKO rabbits exhibited more sensitive immune responses than WT rabbits after PCB implantation, resulted from a significant increase of antibodies (except total antibodies) and cytokines levels, cell lysis ratios, CD4/CD8 proportions, and inflammatory cells infiltration. Immunological factors and inflammatory cells infiltrate in GTKO rabbits after NBM implantation were significantly lower than those in the PCB group. Among the three groups, the NBM group showed the highest contents of new bone formation elements. In conclusion, the GTKO rabbit is a more sensitive alternative model than WT rabbit for preclinical study of xenografts via in situ implantation. Studies on multiple gene-edited animals are also necessary for more comprehensively evaluating xenoimmunologen risks of animal tissue-derived biomaterials in the future. Additionally, the immunogenicity of NBM was remarkably decreased compared to PCB.

Accuracy and prognostic impact of FDG PET/CT and biopsy in bone marrow assessment of follicular lymphoma at diagnosis: A Nation-Wide cohort study.

BACKGOUND: In the workup of follicular lymphoma (FL), bone marrow biopsy (BMB) assessment is a key component of FLIPI and FLIPI2, the most widely used outcome scores. During the previous decade, several studies explored the role of FDG-PET/CT for detecting nodal and extranodal disease, with only one large study comparing both techniques. METHODS: The aim of our study was to evaluate the diagnostic accuracy and the prognostic impact of both procedures in a retrospective cohort of 299 FL patients with both tests performed at diagnosis. In order to avoid a collinearity bias, FLIPI2 was deconstructed in its founding parameters, and the bone marrow involvement (BMI) parameter separately included as: a positive BMB, a positive PET/CT, the combined "PET/CT and BMB positive" or "PET/CT or BMB positive". These variables were also confronted independently with the POD24 in 233 patients treated with intensive regimens. RESULTS: In the total cohort, bone marrow was involved in 124 and 60 patients by BMB and PET/CT, respectively. In terms of overall survival, age > 60 y.o. and the combined "PET/CT or BMB positive" achieved statistical independence as a prognostic factor. In patients treated with an intensive regimen, only the combined "PET/CT or BMB positive" added prognostic value for a shorter overall survival, when confronted with the POD24. CONCLUSION: Our results show that in FL both BMB and PET/CT should be considered at diagnosis, as their combined assessment provides independent prognostic value in the context of the most widely use clinical scores.

Prediction models and associated factors on the fertility behaviors of the floating population in China.

The floating population has been growing rapidly in China, and their fertility behaviors do affect urban management and development. Based on the data set of the China Migrants Dynamic Survey in 2016, the logistic regression model and multiple linear regression model were used to explore the related factors of fertility behaviors among the floating populace. The artificial neural network model, the naive Bayes model, and the logistic regression model were used for prediction. The findings showed that age, gender, ethnic, household registration, education level, occupation, duration of residence, scope of migration, housing, economic conditions, and health services all affected the reproductive behavior of the floating population. Among them, the improvement duration of post-migration residence and family economic conditions positively impacted their fertility behavior. Non-agricultural new industry workers with college degrees or above living in first-tier cities were less likely to have children and more likely to delay childbearing. Among the prediction models, both the artificial neural network model and logistic regression model had better prediction effects. Improving the employment and income of new industry workers, and introducing preferential housing policies might improve their probability of bearing children. The artificial neural network and logistic regression model could predict individual fertility behavior and provide a scientific basis for the urban population management.