Clinically Applicable AI System for Accurate Diagnosis, Quantitative Measurements, and Prognosis of COVID-19 Pneumonia Using Computed Tomography.

Many COVID-19 patients infected by SARS-CoV-2 virus develop pneumonia (called novel coronavirus pneumonia, NCP) and rapidly progress to respiratory failure. However, rapid diagnosis and identification of high-risk patients for early intervention are challenging. Using a large computed tomography (CT) database from 3,777 patients, we developed an AI system that can diagnose NCP and differentiate it from other common pneumonia and normal controls. The AI system can assist radiologists and physicians in performing a quick diagnosis especially when the health system is overloaded. Significantly, our AI system identified important clinical markers that correlated with the NCP lesion properties. Together with the clinical data, our AI system was able to provide accurate clinical prognosis that can aid clinicians to consider appropriate early clinical management and allocate resources appropriately. We have made this AI system available globally to assist the clinicians to combat COVID-19.

Monolithic MOF-Based Metal-Insulator-Metal Resonator for Filtering and Sensing.

Metal-insulator-metal (MIM) configurations based on Fabry-Pérot resonators have advanced the development of color filtering through interactions between light and matter. However, dynamic color changes without breaking the structure of the MIM resonator upon environmental stimuli are still challenging. Here, we report monolithic metal-organic framework (MOF)-based MIM resonators with tunable bandwidth that can boost both dynamic optical filtering and active chemical sensing by laser-processing microwell arrays on the top metal layer. Programmable tuning of the reflection color of the MOF-based MIM resonator is achieved by controlling the MOF layer thicknesses, which is demonstrated by simulation of light-matter interactions on subwavelength scales. Laser-processed microwell arrays are used to boost sensing performance by extending the pathway for diffusion of external chemicals into nanopores of the MOFs. Both experiments and molecular dynamics simulations demonstrate that tailoring the period and height of the microwell array on the MIM resonator can advance the high detection sensitivity of chemicals.

A stable solid-state lithium-oxygen battery enabled by heterogeneous silicon-based interface.

Solid-state lithium-metal batteries using inorganic solid-state electrolyte (SSE) instead of liquid-electrolyte, especially lithium-oxygen (Li-O2) battery, have attracted much more attention due to their high-energy density and safety. However, the poor interface contact between electrodes and SSEs makes these batteries lose most of their capacity and power during cycling. Here we report that by coating a heterogeneous silicon carbide on lithium metal anode and Li1.5Al0.5Ge1.5P3O12(LAGP)-SSE, a good interface contact is created between the electrode and electrolyte that can effectively reduce the interface impedance and improve the cycle performance of the assembled battery. As a result, the solid-sate Li-O2battery demonstrates a cycle lifespan of ∼78 cycles being at least 3-times higher than the solid-state Li-O2battery without silicon carbide with a capacity limitation of 1000 mAh g-1at 250 mA g-1. The characterization of discharge products indicates a typical two-electron convention of oxygen-to-lithium oxide for the solid-state Li-O2battery system. This work paves a way for developing high-energy long-cycle solid-state lithium-metal battery. The work provides insights into the interface between the Li-metal and SSE to develop high-energy long-cycle all solid-state Li-metal batteries.

Impaired lipid metabolism by age-dependent DNA methylation alterations accelerates aging.

Epigenetic alterations and metabolic dysfunction are two hallmarks of aging. However, the mechanism of how their interaction regulates aging, particularly in mammals, remains largely unknown. Here we show ELOVL fatty acid elongase 2 (Elovl2), a gene whose epigenetic alterations are most highly correlated with age prediction, contributes to aging by regulating lipid metabolism. Impaired Elovl2 function disturbs lipid synthesis with increased endoplasmic reticulum stress and mitochondrial dysfunction, leading to key accelerated aging phenotypes. Restoration of mitochondrial activity can rescue age-related macular degeneration (AMD) phenotypes induced by Elovl2 deficiency in human retinal pigmental epithelial (RPE) cells. We revealed an epigenetic-metabolism axis contributing to aging and potentially to antiaging therapy.

Construction of Core-Shell NanoMOFs@microgel for Aqueous Lubrication and Thermal-Responsive Drug Release.

The construction of porous nanocarriers with good lubricating performance and stimuli-responsive drug release is significant for the synergetic therapy of osteoarthritis (OA). Although metal-organic framework nanoparticles (nanoMOFs) as carriers can support drug delivery, achieving the synergy of aqueous lubrication and stimuli-responsive drug release is challenging. In this work, a core-shell nanoMOFs@poly(N-isopropylacrylamide) (PNIPAm) microgel hybrid via one-pot soap-free emulsion polymerization is developed. Programmable growth of the PNIPAm microgel layer on the surface of nanoMOFs is achieved by tuning the concentration of the monomer and the crosslinker in the reaction. Reversible swelling-collapsing behaviors of the hybrid are realized by tuning the temperature below and above the lower critical solution temperature. When used as water lubrication additives, the hybrid enables reductions in both the coefficient of friction and wear volume. In vitro thermal-responsive drug release is demonstrated on the diclofenac sodium-loaded hybrid by controlling the swelling and collapsing states of the PNIPAm nanolayer. Moreover, the good biocompatibility of the hybrid is verified by culturing toward HeLa and BEAS-2B cells. These results establish a nanoMOFs@microgel hybrid that can achieve friction and wear reduction and thermal-responsive drug release.

Regulation of ROS in myeloid-derived suppressor cells through targeting fatty acid transport protein 2 enhanced anti-PD-L1 tumor immunotherapy.

Despite the remarkable success and efficacy of immune checkpoint blockade (ICB) therapy against the PD-1/PD-L1 axis, it induces sustained responses in a sizeable minority of cancer patients due to the activation of immunosuppressive factors such as myeloid-derived suppressor cells (MDSCs). Inhibiting the immunosuppressive function of MDSCs is critical for successful cancer ICB therapy. Interestingly, lipid metabolism is a crucial factor in modulating MDSCs function. Fatty acid transport protein 2 (FATP2) conferred the function of PMN-MDSCs in cancer via the upregulation of arachidonic acid metabolism. However, whether regulating lipid accumulation in MDSCs by targeting FATP2 could block MDSCs reactive oxygen species (ROS) production and enhance PD-L1 blockade-mediated tumor immunotherapy remains unexplored. Here we report that FATP2 regulated lipid accumulation, ROS, and immunosuppressive function of MDSCs in tumor-bearing mice. Tumor cells-derived granulocyte macrophage-colony stimulating factor (GM-CSF) induced FATP2 expression in MDSCs by activation of STAT3 signaling pathway. Pharmaceutical blockade of FATP2 expression in MDSCs by lipofermata decreased lipid accumulation, reduced ROS, blocked immunosuppressive activity, and consequently inhibited tumor growth. More importantly, lipofermata inhibition of FATP2 in MDSCs enhanced anti-PD-L1 tumor immunotherapy via the upregulation of CD107a and reduced PD-L1 expression on tumor-infiltrating CD8+T-cells. Furthermore, the combination therapy blocked MDSC's suppressive role on T- cells thereby enhanced T-cell's ability for the production of IFN-γ. These findings indicate that FATP2 plays a key role in modulating lipid accumulation-induced ROS in MDSCs and targeting FATP2 in MDSCs provides a novel therapeutic approach to enhance anti-PD-L1 cancer immunotherapy.

Global Analyses of Selective Insulin Resistance in Hepatocytes Caused by Palmitate Lipotoxicity.

Obesity is tightly linked to hepatic steatosis and insulin resistance. One feature of this association is the paradox of selective insulin resistance: insulin fails to suppress hepatic gluconeogenesis but activates lipid synthesis in the liver. How lipid accumulation interferes selectively with some branches of hepatic insulin signaling is not well understood. Here we provide a resource, based on unbiased approaches and established in a simple cell culture system, to enable investigations of the phenomenon of selective insulin resistance. We analyzed the phosphoproteome of insulin-treated human hepatoma cells and identified sites in which palmitate selectively impairs insulin signaling. As an example, we show that palmitate interferes with insulin signaling to FoxO1, a key transcription factor regulating gluconeogenesis, and identify altered FoxO1 cellular compartmentalization as a contributing mechanism for selective insulin resistance. This model system, together with our comprehensive characterization of the proteome, phosphoproteome, and lipidome changes in response to palmitate treatment, provides a novel and useful resource for unraveling the mechanisms underlying selective insulin resistance.

Role of G protein-coupled receptor 1 in choriocarcinoma progression.

Choriocarcinoma is characterized by malignant proliferation and transformation of trophoblasts and is currently treated with systemic chemotherapeutic agents. The lack of specific targets for chemotherapeutic agents results in indiscriminate drug distribution. In our study, we aimed to delineate the mechanism by which G protein-coupled receptor 1 (GPR1) regulates the development of choriocarcinoma and thus investigated GPR1 as a prospective chemotherapeutic target. In this study, GPR1 expression levels were examined in several trophoblast cell lines. We found significantly higher GPR1 expression in choriocarcinoma cells (JEG3 and BeWo) than in normal trophoblast cells (HTR-8/SVneo). Additionally, we studied the role of GPR1 in choriocarcinoma in vitro and in vivo. GPR1 knockdown suppressed proliferation, invasion, and Akt and ERK phosphorylation in vitro and slowed tumor growth in vivo. Interestingly, GPR1 overexpression promoted increased proliferation, invasion, and Akt and ERK phosphorylation in vitro. Furthermore, we identified a specific GPR1-binding seven-amino acid peptide, LRH7-G3, that might also suppress choriocarcinoma in vitro and in vivo through phage display. Our study is the first to report that GPR1 may play a role in regulating choriocarcinoma progression through the Akt and ERK pathways. GPR1 could be a promising potential pharmaceutical target for choriocarcinoma.

Mice lacking lipid droplet-associated hydrolase, a gene linked to human prostate cancer, have normal cholesterol ester metabolism.

Variations in the gene LDAH (C2ORF43), which encodes lipid droplet-associated hydrolase (LDAH), are among few loci associated with human prostate cancer. Homologs of LDAH have been identified as proteins of lipid droplets (LDs). LDs are cellular organelles that store neutral lipids, such as triacylglycerols and sterol esters, as precursors for membrane components and as reservoirs of metabolic energy. LDAH is reported to hydrolyze cholesterol esters and to be important in macrophage cholesterol ester metabolism. Here, we confirm that LDAH is localized to LDs in several model systems. We generated a murine model in which Ldah is disrupted but found no evidence for a major function of LDAH in cholesterol ester or triacylglycerol metabolism in vivo, nor a role in energy or glucose metabolism. Our data suggest that LDAH is not a major cholesterol ester hydrolase, and an alternative metabolic function may be responsible for its possible effect on development of prostate cancer.

[Determination of Bi in Soils and Sediments by Atomic Fluorescence Spectrometry].

A method for determination of the contents of Bi in soils and sediments by atomic fluorescence spectrophotometry (AFS) was established by using aqua regia as the dissolved medium. In this paper, the instrument parameters, load flow and reducing agent concentration were optimized. Compared with microwave digestion and electric heating plate digestion, water bath digestion demonstrated the better digestion efficiency and was most commonly used. Under the optimal experimental conditions, the detection limit (LOD) was 0.01 mg·kg-1 (sample quantity 0.500 0 g, sample volume 50 mL), and the limit of quantitation (LOQ) was 0.04 mg·kg-1. The results were in good agreement with the centified value, and the relative error was -4.7%~-2.0%. For determination of soil and sediment actual samples, the relative standard deviation (RSD) were 2.5%~3.4% and 3.1%~3.4%, respectively, and the recoveries of the method respectively ranged from 97.6%~102% and 99.5%~104%.

Predictive machine learning models for anticipating loss to follow-up in tuberculosis patients throughout anti-TB treatment journey.

Loss to follow-up (LTFU) in tuberculosis (TB) management increases morbidity and mortality, challenging effective control strategies. This study aims to develop and evaluate machine learning models to predict loss to follow-up in TB patients, improving treatment adherence and outcomes. Retrospective data encompassing tuberculosis patients who underwent treatment or registration at the National Center for Clinical Medical Research on Infectious Diseases from January 2017 to December 2021 were compiled. Employing machine learning techniques, namely SVM, RF, XGBoost, and logistic regression, the study aimed to prognosticate LTFU. A comprehensive cohort of 24,265 tuberculosis patients underwent scrutiny, revealing a LTFU prevalence of 12.51% (n = 3036). Education level, history of hospitalization, alcohol consumption, outpatient admission, and prior tuberculosis history emerged as precursors for pre-treatment LTFU. Employment status, outpatient admission, presence of chronic hepatitis/cirrhosis, drug adverse reactions, alternative contact availability, and health insurance coverage exerted substantial influence on treatment-phase LTFU. XGBoost consistently surpassed alternative models, boasting superior discriminative ability with an average AUC of 0.921 for pre-treatment LTFU and 0.825 for in-treatment LTFU. Our study demonstrates that the XGBoost model provides superior predictive performance in identifying LTFU risk among tuberculosis patients. The identification of key risk factors highlights the importance of targeted interventions, which could lead to significant improvements in treatment adherence and patient outcomes.

LSTM-Based Prediction Model for Tuberculosis Among HIV-Infected Patients Using Structured Electronic Medical Records: A Retrospective Machine Learning Study.

Background: Both HIV and TB are chronic infectious diseases requiring long-term treatment and follow-up, resulting in extensive electronic medical records. With the exponential growth of health and medical big data, effectively extracting and analyzing these data has become the research hotspot. As a fundamental aspect of artificial intelligence, machine learning has been extensively applied in medical research, encompassing diagnosis, treatment, patient monitoring, drug development, and epidemiological investigations. This significantly enhances medical information systems and facilitates the interoperability of medical data. Methods: In our study, we analyzed longitudinal data from the electronic health records of 4540 patients, gathered from the National Clinical Research Center for Infectious Diseases in Shenzhen, China, spanning from 2017 to 2021. Initially, we employed the fine-tuned ChatGLM to structure the electronic medical records. Subsequently, we utilized a multi-layer perceptron to classify each patient and determined the presence of tuberculosis in HIV patients. Using machine learning-based natural language processing, we structured these records to build a specialized database for HIV and TB co-infection. We studied the epidemiological characteristics, focusing on incidence patterns, patient characteristics, and influencing factors, to uncover the transmission characteristics of these diseases in Shenzhen. Additionally, we used Long Short-Term Memory to create a predictive model for TB co-infection among HIV patients, based on their medical records. This model predicted the risk of TB co-infection, providing scientific evidence for clinical decision-making and enabling early detection and precise intervention. Results: Based on the refined ChatGLM model tailored for structured electronic health records, the accuracy of symptom extraction consistently surpassed 0.95 precision. Key symptoms such as diarrhea and normal showed precision rates exceeding 0.90. High scores were also achieved in recall and F1 scores. Among 4540 HIV patients, 758 were diagnosed with concurrent tuberculosis, indicating a 16.7% co-infection rate, while syphilis co-infection affected 25.1%, underscoring the prevalence of concurrent infections among HIV patients. Utilizing electronic health records, a Multilayer Perceptron classifier was developed as a benchmark against Long Short-Term Memory to predict high-risk groups for HIV and tuberculosis co-infections. The Multilayer Perceptron classifier demonstrated predictive ability with AUROC values ranging from 0.616 to 0.682 on the test set, suggesting opportunities for further optimization and generalization despite its accuracy in identifying HIV-TB co-infections. In tuberculosis intelligent diagnosis based on laboratory results, the Long Short-Term Memory showed consistent performance across 5-fold cross-validation, with AUROC values ranging from 0.827 to 0.850, indicating reliability and consistency in tuberculosis prediction. Furthermore, by optimizing classification thresholds, the model achieved an overall accuracy of 81.18% in distinguishing HIV co-infected tuberculosis from simple HIV infection. Conclusion: Combining the Multilayer Perceptron classifier with Long Short-Term Memory represented an advanced approach for effectively extracting electronic health records and utilizing it for disease prediction. This underscored the superior performance of deep learning techniques in managing both structured and unstructured medical data. Models leveraging laboratory time-series data demonstrated notably better performance compared to those relying solely on electronic health records for predicting tuberculosis incidence. This emphasized the benefits of deep learning in handling intricate medical data and provided valuable insights for healthcare providers exploring the use of deep learning in disease prediction and management.

Dose-effect relationship of copolymer on enhancing aqueous lubrication of a hybrid osteoarthritis drug delivery nanocarrier.

Developing stimulus-responsive properties of drug delivery nanocarriers combined with enhanced joint lubrication is an effective synergistic strategy for treating osteoarthritis. Poly(N-isopropylacrylamide) (PNIPAm) is a typical thermo-responsive polymer, which can achieve drug delivery by transition from swollen state to collapsed state. However, undesired transition temperature, limited drug loading capacity, and weakened mechanical properties in joint present obstacles to use as drug delivery nanocarriers. In this work, we demonstrate dose-effect relationship between the PNIPAm-based copolymer and nanoscale metal-organic frameworks on enhancing both aqueous lubrication and drug delivery performance of a hybrid osteoarthritis (OA) nanocarrier. A series of NIPAm and poly(ethylene glycol)methacrylate (PEGMa) copolymer microgels with different feeding content are optimized to grow on the surface of MIL-101(Cr) nanoparticles via one-pot soap-free emulsion copolymerization method. By changing the feeding mass ratio of NIPAm and PEGMa, MIL-101(Cr)@P(NIPAm-g-PEGMax) (x = 0, 1, 2, 3, and 4, named MPNPx) hybrids can ameliorate the lower critical solution temperature to match with OA and enhance the aqueous lubrication performance. Among the as-synthesized hybrids, MPNP3 hybrids manifested the notable enhanced thermo-responsive tribological performance due to the synergistic effect of "hydration lubrication" and "ball-bearing" function of the optimized copolymer microgel layer on the surface of metal-organic frameworks (MOFs). Anti-inflammatory drug loading is enabled by the high surface area and porosity of the MOFs, and the MPNP3 drug delivery nanocarriers achieve thermo-responsive release in vitro. Our work establishes the dose-effect relationship between thermo-responsive NIPAm and hydrophilic PEGMa of the copolymer grown on the surface of MOFs, providing valuable insights for improving the versatility of stimuli-responsive for biomedical application.

Colorimetric Fabry-Pérot Sensor with Hetero-Structured Dielectric for Humidity Monitoring.

A full-color colorimetric humidity sensor with high brightness is proposed by using a hetero-structured dielectric film in a metal-insulator-metal (MIM) resonator. A humidity-responsive polymer is designed to graft on top of a metal-organic frameworks (MOFs) thin film (MOFs-Polymer) as insulator layer in the resonator. Programmable tuning of reflected color is achieved by controlling the polymer thicknesses, and finite difference time domain simulation of light-matter interactions at subwavelength scales proves the dependence of the reflected wavelength on dielectric layer thickness of the resonator. Vivid full-color changing is realized during tracking humidity process due to swelling of the stimuli-responsive polymer. Ultrafast response (≈0.75 s) is achieved for tracking trace H2O from H2O/methanol mixture, which is ≈104 faster than that of the pure polymer-based MIM resonator. Meanwhile, the study observes significant spectral redshift because the porous MOFs film facilitates the preconcentration of external stimulus and improves the detection sensitivity of the resonator. Further, double-channel anti-counterfeiting multiplexing imaging is devised on the MIM resonator by photomask technology. Patterned encoding for security label is achieved on the MIM resonator by engineering humidity-tunable pixels of Au/MOFs-Polymer/Au and humidity-invalid pixels of Au/MOFs/Au.

A deep-learning pipeline for the diagnosis and grading of common blinding ophthalmic diseases based on lesion-focused classification model.

Background: Glaucoma (GLAU), Age-related Macular Degeneration (AMD), Retinal Vein Occlusion (RVO), and Diabetic Retinopathy (DR) are common blinding ophthalmic diseases worldwide. Purpose: This approach is expected to enhance the early detection and treatment of common blinding ophthalmic diseases, contributing to the reduction of individual and economic burdens associated with these conditions. Methods: We propose an effective deep-learning pipeline that combine both segmentation model and classification model for diagnosis and grading of four common blinding ophthalmic diseases and normal retinal fundus. Results: In total, 102,786 fundus images of 75,682 individuals were used for training validation and external validation purposes. We test our model on internal validation data set, the micro Area Under the Receiver Operating Characteristic curve (AUROC) of which reached 0.995. Then, we fine-tuned the diagnosis model to classify each of the four disease into early and late stage, respectively, which achieved AUROCs of 0.597 (GL), 0.877 (AMD), 0.972 (RVO), and 0.961 (DR) respectively. To test the generalization of our model, we conducted two external validation experiments on Neimeng and Guangxi cohort, all of which maintained high accuracy. Conclusion: Our algorithm demonstrates accurate artificial intelligence diagnosis pipeline for common blinding ophthalmic diseases based on Lesion-Focused fundus that overcomes the low-accuracy of the traditional classification method that based on raw retinal images, which has good generalization ability on diverse cases in different regions.

Promotion of rice seedlings growth and enhancement of cadmium immobilization under cadmium stress with two types of organic fertilizer.

Cadmium (Cd)-contaminated soil poses a severe threat to crop production and human health, while also resulting in a waste of land resources. In this study, two types of organic fertilizer (ZCK: Low-content available iron; Z2: High-content available iron) were applied to Cd-contaminated soil for rice cultivation, and the effects of the fertilizer on rice growth and Cd passivation were investigated in conjunction with soil microbial analysis. Results showed that Z2 could alter the composition, structure, and diversity of microbial communities, as well as enhance the complexity and stability of the microbial network. Both 2% and 5% Z2 significantly increased the fresh weight and dry weight of rice plants while suppressing Cd absorption. The 2% Z2 exhibited the best Cd passivation effect. Gene predictions suggested that Z2 may promote plant growth by regulating microbial production of organic acids that dissolve phosphorus and potassium. Furthermore, it is suggested that Z2 may facilitate the absorption and immobilization of soil cadmium through the regulation of microbial cadmium efflux and uptake systems, as well as via the secretion of extracellular polysaccharides. In summary, Z2 can promote rice growth, suppress Cd absorption by rice, and passivate soil Cd by regulating soil microbial communities.

Accurate prediction of myopic progression and high myopia by machine learning.

Background: Myopia is a leading cause of visual impairment in Asia and worldwide. However, accurately predicting the progression of myopia and the high risk of myopia remains a challenge. This study aims to develop a predictive model for the development of myopia. Methods: We first retrospectively gathered 612 530 medical records from five independent cohorts, encompassing 227 543 patients ranging from infants to young adults. Subsequently, we developed a multivariate linear regression algorithm model to predict the progression of myopia and the risk of high myopia. Result: The model to predict the progression of myopia achieved an R2 value of 0.964 vs a mean absolute error (MAE) of 0.119D [95% confidence interval (CI): 0.119, 1.146] in the internal validation set. It demonstrated strong generalizability, maintaining consistent performance across external validation sets: R2 = 0.950 vs MAE = 0.119D (95% CI: 0.119, 1.136) in validation study 1, R2 = 0.950 vs MAE = 0.121D (95% CI: 0.121, 1.144) in validation study 2, and R2 = 0.806 vs MAE = -0.066D (95% CI: -0.066, 0.569) in the Shanghai Children Myopia Study. In the Beijing Children Eye Study, the model achieved an R2 of 0.749 vs a MAE of 0.178D (95% CI: 0.178, 1.557). The model to predict the risk of high myopia achieved an area under the curve (AUC) of 0.99 in the internal validation set and consistently high area under the curve values of 0.99, 0.99, 0.96 and 0.99 in the respective external validation sets. Conclusion: Our study demonstrates accurate prediction of myopia progression and risk of high myopia providing valuable insights for tailoring strategies to personalize and optimize the clinical management of myopia in children.

Inhibitory effects and molecular mechanisms of ginsenoside Rg1 on the senescence of hematopoietic stem cells.

Hematopoietic stem cells (HSCs) produce all blood cell lineages and maintain life-long hematopoiesis. However, the self-renewal ability and differentiation capacity of HSCs reduces with age. The senescence of HSCs can lead to the imbalance of hematopoietic homeostasis and immune disorder and induce a variety of age-related diseases. Recent studies have shown that therapeutic interventions targeting the senescence of HSCs may prevent disease progression. Ginsenoside Rg1 (Rg1), extracted from roots or stems of ginseng, has beneficial antiaging activities. It has been reported that Rg1 can inhibit the senescence of HSCs. Here, we reviewed recent advances of Rg1 in inhibiting the senescence of HSCs and discussed related molecular mechanisms. Bioinformatics and network databases have been widely applied to drug discoveries. Here, we predicted potential antiaging targets of Rg1 explored by bioinformatic methods, which may help discover new targets of Rg1 and provide novel strategies for delaying the aging process of HSCs.

Factors associated with loss to follow-up before and after treatment initiation among patients with tuberculosis: A 5-year observation in China.

Background: Loss to follow-up (LTFU) is a significant barrier to the completion of anti-tuberculosis (TB) treatment and a major predictor of TB-associated deaths. Currently, research on LTFU-related factors in China is both scarce and inconsistent. Methods: We collected information from the TB observation database of the National Clinical Research Center for Infectious Diseases. The data of all patients who were documented as LTFU were assessed retrospectively and compared with those of patients who were not LTFU. Descriptive epidemiology and multivariable logistic regression analyses were conducted to identify the factors associated with LTFU. Results: A total of 24,265 TB patients were included in the analysis. Of them, 3,046 were categorized as LTFU, including 678 who were lost before treatment initiation and 2,368 who were lost afterwards. The previous history of TB was independently associated with LTFU before treatment initiation. Having medical insurance, chronic hepatitis or cirrhosis, and providing an alternative contact were independent predictive factors for LTFU after treatment initiation. Conclusion: Loss to follow-up is frequent in the management of patients with TB and can be predicted using patients' treatment history, clinical characteristics, and socioeconomic factors. Our research illustrates the importance of early assessment and intervention after diagnosis. Targeted measures can improve patient engagement and ultimately treatment adherence, leading to better health outcomes and disease control.