High-Nuclear Co-Added Polyoxometalate-Based Chain: Electrocatalytic Oxygen Production.

A high-nuclear Co-added polyoxometalate (CoAP) was synthesized via a hydrothermal reaction: H14.5K9Na7.5-{[Co8(µ2-OH)(µ3-OH)2(H2O)2(Co(H2O)GeW6O26)(B-α-GeW9O34)2][BO(OH)2][Co12(µ2-OH)(µ3-OH)5(H2O)3(Co(H2O)GeW6O26)(GeW6O26)(B-α-GeW9O34)]}·46H2O (1). The polyoxoanion of 1 contains a large Co20 cluster gathered by lacunary GeW6O26 and GeW9O34 subunits. 1 represents a one-dimensional (1D) chain formed by adjacent polyoxoanions coupling through a CoO6 double bridge, showing the first example of a high-nuclear CoAP-based inorganic chain. 1 served as an efficient electrocatalyst in oxygen evolution reactions (OERs).

Switching CO2 Electroreduction toward Ethanol by Delocalization State-Tuned Bond Cleavage.

The electrochemical CO2 reduction reaction by copper-based catalysts features a promising approach to generate value-added multicarbon (C2+) products. However, due to the unfavored formation of oxygenate intermediates on the catalyst surface, the selectivity of C2+ alcohols like ethanol remains unsatisfactory compared to that of ethylene. The bifurcation point (i.e., the CH2═CHO* intermediate adsorbed on Cu via a Cu-O-C linkage) is critical to the C2+ product selectivity, whereas the subsequent cleavage of the Cu-O or the O-C bond determines the ethanol or ethylene pathway. Inspired by the hard-soft acid-base theory, in this work, we demonstrate an electron delocalization tuning strategy of the Cu catalyst by a nitrene surface functionalization approach, which allows weakening and cleaving of the Cu-O bond of the adsorbed CH2═CHO*, as well as accelerating hydrogenation of the C═C bond along the ethanol pathway. As a result, the nitrene-functionalized Cu catalyst exhibited a much-enhanced ethanol Faradaic efficiency of 45% with a peak partial current density of 406 mA·cm-2, substantially exceeding that of unmodified Cu or amide-functionalized Cu. When assembled in a membrane electrode assembly electrolyzer, the catalyst presented a stable CO2-to-ethanol conversion for >300 h at an industrial current density of 400 mA·cm-2.

Nontarget Discovery of Per- and Polyfluoroalkyl Sulfonyl Halides in Soils by Integration of Derivatization and Specific Fragment-Based Liquid Chromatography-High Resolution Mass Spectrometry Screening.

Though long recognized as synthetic precursors to other poly- and perfluoroalkyl substances (PFASs), most poly- and perfluoroalkyl sulfonyl halides (PASXs) cannot be directly measured and have generally received minimal attention. Inspired by the redox reaction between sulfonyl halide groups and p-toluenethiol in organic chemistry, we developed a novel nontarget analysis strategy for PASXs by intergrating derivatization and specific fragment-based liquid chromatography-high resolution mass spectrometry screening for m/z 82.961 [SO2F-] and m/z 95.934 [S2O2-]. By using this strategy, we discovered 11 PASXs, namely, perfluoroalkyl sulfonyl fluorides (5), polyfluoroalkyl sulfonyl fluorides (2), unsaturated perfluoroalkyl sulfonyl fluoride (1), and perfluoroalkyl sulfonyl chlorides (3) in soil samples collected from an abandoned fluorochemical manufacturing park. These average ∑PASXs concentrations were 1120 µg kg-1 (range: 9.7-9860 µg kg-1), which were very likely to be the key intermediates and undesired byproducts of electrochemical fluorination processes. Spatial variation in the mass ratio of ∑PASXs to ∑PFSAs (range: 0.7-795%) also indicates their different transportation pathways. More importantly, the decline of PASXs and increase of perfluoroalkyl sulfonates (when compared to a prior study at this site) suggest the continued hydrolysis of PASXs and the relatively fast environmental transformation rates in the abandoned fluorochemical park soils. Overall, these findings demonstrated the utility of a novel nontarget analysis strategy, which may change most PASXs from inferred precursors to measured intermediates and further could be adapted for structures, distribution, and transformation studies of PFASXs in other matrices.

Efficient, Color-Stable, Pure-Blue Light-Emitting Diodes Based on Aromatic Ligand-Engineered Perovskite Nanoplatelets.

Perovskite nanoplatelets (NPLs) show great potential for high-color-purity light-emitting diodes (LEDs) due to their narrow line width and high exciton binding energy. However, the performance of perovskite NPL LEDs lags far behind perovskite quantum dot-/film-based LEDs, owing to their material instability and poor carrier transport. Here, we achieved efficient and stable pure blue-emitting CsPbBr3 NPLs with outstanding optical and electrical properties by using an aromatic ligand, 4-bromothiophene-2-carboxaldehyde (BTC). The BTC ligands with thiophene groups can guide two-dimensional growth and inhibit out-of-plane ripening of CsPbBr3 NPLs, which, meanwhile, increases their structural stability via strongly interacting with PbBr64- octahedra. Moreover, aromatic structures with delocalized π-bonds facilitate charge transport, diminish band tail states, and suppress Auger processes in CsPbBr3 NPLs. Consequently, the LEDs demonstrate efficient and color-stable blue emissions at 465 nm with a narrow emission line width of 17 nm and a maximum external quantum efficiency (EQE) of 5.4%, representing the state-of-the-art CsPbBr3 NPL LEDs.

Patterns and organ treatment response of Erdheim-Chester disease with cardiac involvement.

OBJECTIVE: To evaluate the heart response of Erdheim-Chester disease (ECD) through continuous follow-up within our large cohort, for which there is a lack of understanding. METHODS: We conducted a retrospective analysis of clinical data from patients with ECD with cardiac involvement diagnosed at our centre between January 2010 and August 2023. We assessed the heart response by integrating pericardial effusion and metabolic responses. RESULTS: A total of 40 patients were included, with a median age of 51.5 years (range: 29-66) and a BRAFV600E mutation rate of 56%. The most common imaging manifestations observed were pericardial effusion (73%), right atrium (70%) and right atrioventricular sulcus infiltration (58%). Among 21 evaluable patients, 18 (86%) achieved a heart response including 5 (24%) complete response (CR) and 13 (62%) partial response (PR). The CR rate of pericardial effusion response was 33%, while the PR rate was 56%. Regarding the cardiac mass response, 33% of patients showed PR. For cardiac metabolic response, 32% and 53% of patients achieved complete and partial metabolic response, respectively. There was a correlation between pericardial effusion response and cardiac metabolic response (r=0.73 (95% CI 0.12 to 0.83), p<0.001). The median follow-up was 50.2 months (range: 1.0-102.8 months). The estimated 5-year overall survival was 78.9%. The median progression-free survival was 59.4 months (95% CI 26.2 to 92.7 months). Patients who received BRAF inhibitors achieved better heart response (p=0.037) regardless of treatment lines. CONCLUSION: We pioneered the evaluation of heart response of ECD considering both pericardial effusion and cardiac metabolic response within our cohort, revealing a correlation between these two indicators. BRAF inhibitors may improve heart response, regardless of the treatment lines.

Valproic acid regulates the miR-155/Jarid2 axis by affecting miR-155 promoter methylation in glioma.

The most frequent primary brain tumor in adults is glioma, yet no effective curative treatments are currently available. Our previous study demonstrated the enhancing effects of JARID2 on glioma sensitivity to TMZ treatment. In this study, miR-155 is predicted to target JARID2. miR-155 is overexpressed in clinical glioma specimens and cell lines. miR-155 overexpression in glioma cells enhances cell viability and represses cell apoptosis. Through targeting, miR-155 inhibits JARID2 expression. miR-155 inhibition inhibits glioma cell viability and enhances cell apoptosis, whereas JARID2 knockdown enhances cell viability and inhibits cell apoptosis; JARID2 knockdown partially reverses miR-155 inhibition effects on glioma phenotypes. miR-155 inhibition reduces but knockdown of JARID2 promotes the tumor formation ability of glioma cells in vivo. Valproic acid (VPA) upregulates JARID2 expression, inhibits glioma cell viability and enhances cell apoptosis. VPA downregulates the expression level of miR-155 by increasing the methylation level of the miR-155 promoter, suggesting that the miR-155/JARID2 axis is implicated in VPA inhibition of glioma cell viability and enhancement of glioma cell apoptosis. This study demonstrates a new mechanism of VPA treatment of gliomas by affecting the miR-155/JARID2 axis, which could be regarded as a new strategy for the prevention and treatment of glioma.

Metabolic and Microbial Profiling of Soil Microbial Community under Per- and Polyfluoroalkyl Substance (PFAS) Stress.

Per- and polyfluoroalkyl substances (PFAS) represent significant stress to organisms and are known to disrupt microbial community structure and function. Nevertheless, a detailed knowledge of the soil microbial community responding to PFAS stress at the metabolism level is required. Here we integrated UPLC-HRMS-based metabolomics data with 16S rRNA and ITS amplicon data across soil samples collected adjacent to a fluoropolymer production facility to directly identify the biochemical intermediates in microbial metabolic pathways and the interactions with microbial community structure under PFAS stress. A strong correlation between metabolite and microbial diversity was observed, which demonstrated significant variations in soil metabolite profiles and microbial community structures along with the sampling locations relative to the facility. Certain key metabolites were identified in the metabolite-PFAS co-occurrence network, functioning on microbial metabolisms including lipid metabolism, amino acid metabolism, and secondary metabolite biosynthesis. These results provide novel insights into the impacts of PFAS contamination on soil metabolomes and microbiomes. We suggest that soil metabolomics is an informative and useful tool that could be applied to reinforce the chemical evidence on the disruption of microbial ecological traits.

Tellurium-Doped 0D Organic-Inorganic Hybrid Lead-Free Perovskite for X-ray Imaging.

The application of X-ray imaging in military, industrial flaw detection, and medical examination is inseparable from the wide application of scintillator materials. In order to substitute for lead, lower costs, and reduce self-absorption, organic-inorganic hybrid lead-free perovskite scintillators are emerging as a new option. In this work, novel (TEA)2Zr1-xTexCl6 perovskite microcrystals (MCs) were successfully synthesized by a hydrothermal method, with Te4+ doping, which leads to yellow triplet-state self-trapped excitons emission. The emission peak of (TEA)2Zr1-xTexCl6 located at 605 nm under X-ray excitation, which was applied to X-ray imaging, shows a clear wiring structure inside the USB connector. The detection limit (DL) of 820 nGyair/s for (TEA)2Zr0.9Te0.1Cl6 is well below the dose rate corresponding to a standard medical X-ray diagnosis is 5.5 µGyair/s. This work opens up a new path for organic-inorganic hybrid lead-free scintillators.

Electronic State Engineering in Perovskite-Cerium-Composite Nanocrystals toward Enhanced Triplet Annihilation Upconversion.

Wavelength conversion based on hybrid inorganic-organic sensitized triplet-triplet annihilation upconversion (TTA-UC) is promising for applications such as photovoltaics, light-emitting-diodes, photocatalysis, additive manufacturing, and bioimaging. The efficiency of TTA-UC depends on the population of triplet excitons involved in triplet energy transfer (TET), the driving force in TET, and the coupling strength between the donor and acceptor. Consequently, achieving highly efficient TTA-UC necessitates the precise control of the electronic states of inorganic donors. However, conventional covalently bonded nanocrystals (NCs) face significant challenges in this regard. Herein, a novel strategy to exert control over electronic states is proposed, thereby enhancing TET and TTA-UC by incorporating ionic-bonded CsPbBr3 and lanthanide Ce3+ ions into composite NCs. These composite-NCs exhibit high photoluminescence quantum yield, extended single-exciton lifetime, quantum confinement, and uplifted energy levels. This engineering strategy of electronic states engendered a comprehensive impact, augmenting the population of triplet excitons participating in the TET process, enhancing coupling strength and the driving force, ultimately leading to an unconventional, dopant concentration-dependent nonlinear enhancement of UC efficiency. This work not only advances fundamental understanding of hybrid TTA-UC but also opens a door for the creation of other ionic-bonded composite NCs with tunable functionalities, promising innovations for next-generation optoelectronic applications.

Water-Resistant Subwavelength Perovskite Lasing from Transparent Silica-Based Nanocavity.

Great research efforts are devoted to exploring the miniaturization of chip-scale coherent light sources possessing excellent lasing performance. Despite the indispensable role in Si photonics, SiO2 is generally considered not contributing to the starting up and operation of integrated lasers. Here, this work demonstrates an extraordinary-performance subwavelength-scale perovskite vertical cavity laser with all-transparent SiO2 cavity, whose cavity is ultra-simple and composed of only two parallel SiO2 plates. By introducing a ligand-assisted thermally co-evaporation strategy, highly luminescent perovskite film with high reproducibility and excellent optical gain is grown directly on SiO2 . Benefitting from their high-refractive-index contrast, low-threshold, high-quality factor, and single-mode lasing is achieved in subwavelength range of ≈120 nm, and verified by long-range coherence distance (115.6 µm) and high linear polarization degree (82%). More importantly, the subwavelength perovskite laser device could operate in water for 20 days without any observable degradation, exhibiting ultra-stable water-resistant performance. These findings would provide a simple but robust and reliable strategy for the miniaturized on-chip lasers compatible with Si photonics.

Interfacial Water Tuning by Intermolecular Spacing for Stable CO2 Electroreduction to C2+ Products.

Electroreduction of CO2 to multi-carbon (C2+ ) products is a promising approach for utilization of renewable energy, in which the interfacial water quantity is critical for both the C2+ product selectivity and the stability of Cu-based electrocatalytic sites. Functionalization of long-chain alkyl molecules on a catalyst surface can help to increase its stability, while it also tends to block the transport of water, thus inhibiting the C2+ product formation. Herein, we demonstrate the fine tuning of interfacial water by surface assembly of toluene on Cu nanosheets, allowing for sustained and enriched CO2 supply but retarded water transfer to catalytic surface. Compared to bare Cu with fast cathodic corrosion and long-chain alkyl-modified Cu with main CO product, the toluene assembly on Cu nanosheet surface enabled a high Faradaic efficiency of 78 % for C2+ and a partial current density of 1.81 A cm-2 . The toluene-modified Cu catalyst further exhibited highly stable CO2 -to-C2 H4 conversion of 400 h in a membrane-electrode-assembly electrolyzer, suggesting the attractive feature for both efficient C2+ selectivity and excellent stability.

Vertical migration and dissipation of oxytetracycline induces the recoverable shift in microbial community and antibiotic resistance.

Antibiotic resistance gene (ARG) spread in anthropogenic polluted soils is believed to be accelerated by the incidental inputs of antibiotics via fertilizing and irrigation, and endangering food and human health. However, due to the complex nature of substrates and uncertain microbial responses, the primary drivers of ARG dissemination remain unclear. To address this concern, the effects of antibiotic inputs on soil microbes and antibiotic resistance under simulated natural conditions was investigated in this study. Specifically, four flow-through reactors with gravity flow were established, and the oxytetracycline (OTC) a typical antibiotic in agricultural soils was studied at environmental concentrations (i.e. 0.1, 1 and 10 mg/kg) for 31 days. The vertical distribution and dissipation of OTC were profiled by measuring the residuals in layers over time. Correspondingly, the effects of antibiotic exposure on microbial communities and ARG abundances were studied. The results showed that the average exposure intensity of OTC in different soil layers ranged in 0.03-6.45 mg/kg, and resulted in different dissipation kinetics. In addition, top layer was found to be the main site of OTC reduction, where OTC dissipated at magnitude of 74.0-96.6 %, depending on the initial OTC concentration. OTC migration and dissipation resulted in the shift of community composition to the extent of 0.25-0.33 in terms of Bray-Curtis distance, which partially recovered over time. And the achievement of alternative community compositions was supposed to be largely affected by the microbial interaction. Along with the community changes, a short-term accumulation of resistance genes was detected, while the relative abundance of indicator ARGs, i.e. tetG and mexB, rising up to 10-fold higher than the initial, although eventually decayed. Collective findings of this study indicated that antibiotics at environmental concentrations might trigger extra microbial interactions and thereby reducing the demand for ARGs accumulation. It provided valuable understandings in the risk of antibiotic spillage, especially for the incident exposure at the environmentally relevant concentrations.

Ionic Solvent-Assisted MAPbBr3 Perovskite Film for Two-Photon Pumped Single-Mode Laser.

Miniaturized coherent light sources on the nanoscale are highly desired for on-chip photonics integration. However, when approaching the diffraction limit, the sub-wavelength-scale all-dielectric lasers are difficult to realize due to the trade-off between lasing performance and physical size. Especially for a thin-film laser, usually an externally complex cavity is required to provide the necessary optical feedback. Herein, we successfully shrink the MAPbBr3 perovskite thin-film laser to sub-wavelength scale (300 nm) with simplified cavity design using only an ultraviolet glue layer and a quartz glass. The morphology quality and the gain properties of the film are enhanced by introducing ionic liquid. Consequently, the stable and low-threshold single-mode laser with a highly linear polarization degree of 78.6% and a narrow line width of 0.35 nm is achieved under two-photon excitation. The excellent single-mode laser with sub-wavelength scale and ultrasimplified structure could provide a facile and versatile platform for future integrated optoelectronic devices.

Generalized central slice theorem perspective on Fourier-transform spectral imaging at a sub-Nyquist sampling rate.

Fourier-transform spectral imaging captures frequency-resolved images with high spectral resolution, broad spectral range, high photon flux, and low stray light. In this technique, spectral information is resolved by taking Fourier transformation of the interference signals of two copies of the incident light at different time delays. The time delay should be scanned at a high sampling rate beyond the Nyquist limit to avoid aliasing, at the price of low measurement efficiency and stringent requirements on motion control for time delay scan. Here we propose, what we believe to be, a new perspective on Fourier-transform spectral imaging based on a generalized central slice theorem analogous to computerized tomography, using an angularly dispersive optics decouples measurements of the spectral envelope and the central frequency. Thus, as the central frequency is directly determined by the angular dispersion, the smooth spectral-spatial intensity envelope is reconstructed from interferograms measured at a sub-Nyquist time delay sampling rate. This perspective enables high-efficiency hyperspectral imaging and even spatiotemporal optical field characterization of femtosecond laser pulses without a loss of spectral and spatial resolutions.

Towards precise diagnosis time profile of ultrafast electron bunch trains using orthogonal terahertz streak camera.

Ultrafast electron microbunch trains have broad applications in which the individual bunch length and the bunch-to-bunch interval are critical parameters that need to be precisely diagnosed. However, directly measuring these parameters remains challenging. This paper presents an all-optical method that simultaneously measures the individual bunch length and the bunch-to-bunch spacing through an orthogonal THz-driven streak camera. For a 3 MeV electron bunch train, the simulation indicates that the temporal resolution of individual bunch length and the bunch-to-bunch spacing is 2.5 fs and 1 fs, respectively. Through this method, we expect to open a new chapter in the temporal diagnostic of electron bunch trains.

Cascaded Nanozyme with In Situ Enhanced Photothermal Capacity for Tumor-Specific and Self-Replenishing Cancer Therapy.

Reactive oxygen species (ROS)-involved tumor therapeutic strategy, chemodynamic therapy (CDT), has attracted extensive research interest in the scientific community. However, the therapeutic effect of CDT is insufficient and unsustainable owing to the limited endogenous H2 O2 level in the tumor microenvironment. Here, peroxidase (POD)-like RuTe2 nanozyme with the immobilization of glucose oxidase (GOx) and allochroic 3,3',5,5'-tetramethylbenzidine (TMB) molecule have been synthesized to construct RuTe2 -GOx-TMB nanoreactors (RGT NRs) as cascade reaction systems for tumor-specific and self-replenishing cancer therapy. GOx in sequential nanocatalysts can effectively deplete glucose in tumor cells. Meanwhile, a sustainable supply of H2 O2 for subsequent Fenton-like reactions catalyzed by RuTe2 nanozyme is achieved in response to the mild acidic tumor microenvironment. Through this cascade reaction, highly toxic hydroxyl radicals (·OH) are produced, which can further oxidize TMB to trigger tumor-specific "turn-on" photothermal therapy (PTT). In addition, PTT and massive ROS can stimulate the tumor immune microenvironment and activate the systematic anti-tumor immune responses, exerting a notable effect on hindering tumor recurrence and metastasis. This study paves a promising paradigm for synergistic starvation therapy, PTT, and CDT cancer therapy with high efficiency.

The efficacy and safety of anti-PD-1/PD-L1 in treatment of glioma: a single-arm meta-analysis.

Objective: This meta-analysis aimed to evaluate the efficacy and safety of PD-1/PD-L1 inhibitors in patients with glioma. Methods: PubMed, EMBASE, Web of Science, and the Cochrane library were searched from inception to January 2023 without language restriction. Primary outcomes included overall survival (OS), progression-free survival (PFS), objective response rate (ORR), and adverse events (AEs). The risk of bias was assessed by subgroup analysis, sensitivity analysis, and publication bias, including funnel plot, Egger's test, and Begg's test. Results: A total of 20 studies involving 2,321 patients were included in this meta-analysis. In the analysis of the included phase III clinical trials, the forest plot showed that PD-1/PD-L1 inhibitors did not improve the OS (HR=1.15, 95% CI: 1.03-1.29, P=0.02, I2 = 14%) and PFS (HR=1.43, 95% CI: 1.03-1.99, P=0.03, I2 = 87%). In the single-arm analysis, the forest plot demonstrated that the 6-month OS was 71% (95% CI: 57%-83%, I2 = 92%), 1-year OS was 43% (95% CI: 33%-54%, I2 = 93%), and the 2-year OS was 27% (95% CI: 13%-44%, I2 = 97%). The pooled estimate of the median OS was 8.85 months (95% CI: 7.33-10.36, I2 = 91%). Furthermore, the result indicated that the 6-month PFS was 28% (95% CI: 18%-40%, I2 = 95%), 1-year PFS was 15% (95% CI: 8%-23%, I2 = 92%), and the 18-month PFS was 10% (95% CI: 3%-20%, I2 = 93%). The pooled estimate of the median PFS was 3.72 months (95% CI: 2.44-5.00, I2 = 99%). For ORR, the pooled estimate of ORR was 10% (95% CI: 2%-20%, I2 = 88%). We further analyzed the incidence of PD-1/PD-L1 inhibitor-related AEs, and the pooled incidence of AEs was 70% (95% CI: 58%-81%, I2 = 94%). The incidence of AEs ≥ grade 3 was 19% (95% CI: 11%-30%, I2 = 94%). The funnel plot for the median PFS and median OS was symmetric with no significant differences in Egger's test and Begg's test. The sensitivity analysis revealed that our results were stable and reliable. Conclusion: The results of this meta-analysis suggest that anti-PD-1/PD-L1 therapy is relatively safe but could not prolong survival in glioma. More randomized controlled trials are needed to confirm our results. Systematic review registration: https://www.crd.york.ac.uk/prospero/, identifier CRD42023396057.

Corrigendum: Predicting the recurrence and overall survival of patients with glioma based on histopathological images using deep learning.

[This corrects the article DOI: 10.3389/fneur.2023.1100933.].

Contribution of whole slide imaging-based deep learning in the assessment of intraoperative and postoperative sections in neuropathology.

The pathological diagnosis of intracranial germinoma (IG), oligodendroglioma, and low-grade astrocytoma on intraoperative frozen section (IFS) and hematoxylin-eosin (HE)-staining section directly determines patients' treatment options, but it is a difficult task for pathologists. We aimed to investigate whether whole-slide imaging (WSI)-based deep learning can contribute new precision to the diagnosis of IG, oligodendroglioma, and low-grade astrocytoma. Two types of WSIs (500 IFSs and 832 HE-staining sections) were collected from 379 patients at multiple medical centers. Patients at Center 1 were split into the training, testing, and internal validation sets (3:1:1), while the other centers were the external validation sets. First, we subdivided WSIs into small tiles and selected tissue tiles using a tissue tile selection model. Then a tile-level classification model was established, and the majority voting method was used to determine the final diagnoses. Color jitter was applied to the tiles so that the deep learning (DL) models could adapt to the variations in the staining. Last, we investigated the effectiveness of model assistance. The internal validation accuracies of the IFS and HE models were 93.9% and 95.3%, respectively. The external validation accuracies of the IFS and HE models were 82.0% and 76.9%, respectively. Furthermore, the IFS and HE models can predict Ki-67 positive cell areas with R2 of 0.81 and 0.86, respectively. With model assistance, the IFS and HE diagnosis accuracy of pathologists improved from 54.6%-69.7% and 53.5%-83.7% to 87.9%-93.9% and 86.0%-90.7%, respectively. Both the IFS model and the HE model can differentiate the three tumors, predict the expression of Ki-67, and improve the diagnostic accuracy of pathologists. The use of our model can assist clinicians in providing patients with optimal and timely treatment options.

Predicting the recurrence and overall survival of patients with glioma based on histopathological images using deep learning.

Background: A deep learning (DL) model based on representative biopsy tissues can predict the recurrence and overall survival of patients with glioma, leading to optimized personalized medicine. This research aimed to develop a DL model based on hematoxylin-eosin (HE) stained pathological images and verify its diagnostic accuracy. Methods: Our study retrospectively collected 162 patients with glioma and randomly divided them into a training set (n = 113) and a validation set (n = 49) to build a DL model. The HE-stained slide was segmented into a size of 180 × 180 pixels without overlapping. The patch-level features were extracted by the pre-trained ResNet50 to predict the recurrence and overall survival. Additionally, a light-strategy was introduced where low-size digital biopsy images with clinical information were inputted into the DL model to ensure minimum memory occupation. Results: Our study extracted 512 histopathological features from the HE-stained slides of each glioma patient. We identified 36 and 18 features as significantly related to disease-free survival (DFS) and overall survival (OS), respectively, (P < 0.05) using the univariate Cox proportional-hazards model. Pathomics signature showed a C-index of 0.630 and 0.652 for DFS and OS prediction, respectively. The time-dependent receiver operating characteristic (ROC) curves, along with nomograms, were used to assess the diagnostic accuracy at a fixed time point. In the validation set (n = 49), the area under the curve (AUC) in the 1- and 2-year DFS was 0.955 and 0.904, respectively, and the 2-, 3-, and 5-year OS were 0.969, 0.955, and 0.960, respectively. We stratified the patients into low- and high-risk groups using the median hazard score (0.083 for DFS and-0.177 for OS) and showed significant differences between these groups (P < 0.001). Conclusion: Our results demonstrated that the DL model based on the HE-stained slides showed the predictability of recurrence and survival in patients with glioma. The results can be used to assist oncologists in selecting the optimal treatment strategy in clinical practice.