A comparison between centralized and asynchronous federated learning approaches for survival outcome prediction using clinical and PET data from non-small cell lung cancer patients.

BACKGROUND AND OBJECTIVE: Survival analysis plays an essential role in the medical field for optimal treatment decision-making. Recently, survival analysis based on the deep learning (DL) approach has been proposed and is demonstrating promising results. However, developing an ideal prediction model requires integrating large datasets across multiple institutions, which poses challenges concerning medical data privacy. METHODS: In this paper, we propose FedSurv, an asynchronous federated learning (FL) framework designed to predict survival time using clinical information and positron emission tomography (PET)-based features. This study used two datasets: a public radiogenic dataset of non-small cell lung cancer (NSCLC) from the Cancer Imaging Archive (RNSCLC), and an in-house dataset from the Chonnam National University Hwasun Hospital (CNUHH) in South Korea, consisting of clinical risk factors and F-18 fluorodeoxyglucose (FDG) PET images in NSCLC patients. Initially, each dataset was divided into multiple clients according to histological attributes, and each client was trained using the proposed DL model to predict individual survival time. The FL framework collected weights and parameters from the clients, which were then incorporated into the global model. Finally, the global model aggregated all weights and parameters and redistributed the updated model weights to each client. We evaluated different frameworks including single-client-based approach, centralized learning and FL. RESULTS: We evaluated our method on two independent datasets. First, on the RNSCLC dataset, the mean absolute error (MAE) was 490.80±22.95 d and the C-Index was 0.69±0.01. Second, on the CNUHH dataset, the MAE was 494.25±40.16 d and the C-Index was 0.71±0.01. The FL approach achieved centralized method performance in PET-based survival time prediction and outperformed single-client-based approaches. CONCLUSIONS: Our results demonstrated the feasibility and effectiveness of employing FL for individual survival prediction in NSCLC patients, using clinical information and PET-based features.

Amorphous-Crystalline Interfaces on Hollow Nanocubes Derived from Ir-Doped Ni-Fe-Zn Prussian Blue Analog Enables High Capability of Alkaline/Acidic/Saline Water Oxidations.

Development of highly efficient and robust electrocatalysts for oxygen evolution reaction (OER) under specific electrolyte is a key to actualize commercial low-temperature water electrolyzers. Herein, a rational catalyst design strategy is first reported based on amorphous-crystalline (a-c) interfacial engineering to achieve high catalytic activity and durability under diverse electrolytes that can be used for all types of low-temperature water electrolysis. Abundant a-c interface (ACI) is implemented into a hollow nanocubic (pre)-electrocatalyst which is derived from Ir-doped Ni-Fe-Zn Prussian blue analogues (PBA). The implemented c-a interface is well maintained during prolonged OER in alkaline, alkalized saline, and acidic electrolytes demonstrating its diverse functionality for water electrolysis. Notably, the final catalyst exhibits superior catalytic activity with excellent durability for OER compared to that of benchmark IrO2 catalyst, regardless of chemical environment of electrolytes. Hence, this work can be an instructive guidance for developing the ACI engineered electroctalyst which can be diversely used for different types of low-temperature electrolyzers.

Design Strategies of Li-Si Alloy Anode for Mitigating Chemo-Mechanical Degradation in Sulfide-Based All-Solid-State Batteries.

Composite anodes of Li3 PS4  glass+Li-Si alloy (Type 1) and Li3 N+LiF+Li-Si alloy (Type 2) are prepared for all-solid-state batteries with Li3 PS4 (LPS) glass electrolyte and sulfur/LPS glass/carbon composite cathode. Using a three-electrode system, the anode and cathode potentials are separated, and their polarization resistances are individually traced. Even under high-cutoff-voltage conditions (3.7 V), Type 1 and 2 cells are stably cycled without voltage noise for >200 cycles. Although cathode polarization resistance drastically increases after 3.7 V charge owing to LPS oxidation, LPS redox behavior is fairly reversible upon discharge-charge unlike the non-composite alloy anode cell. Time-of-flight secondary ion mass spectrometry analysis reveals that the enhanced cyclability is attributed to uniform Li-Si alloying throughout the composite anode, providing more pathways for lithium ions even when these ions are over-supplied via LPS oxidation. These results imply that LPS-based cells can be reversibly cycled with LPS redox even under high-cutoff voltages, as long as non-uniform alloying (lithium dendrite growth) is prevented. Type 1 and 2 cells exhibit similar performance and stability although reduction product is formed in Type 1. This work highlights the importance of alloy anode design to prevent chemo-mechanical failure when cycling the cell outside the electrochemical stability window.

The Effect of Co-Doping at the A-Site on the Structure and Oxide Ion Conductivity in (Ba0.5-xSrx)La0.5InO3-δ: A Molecular Dynamics Study.

A molecular dynamics simulation was used to investigate the structural and transport properties of a (Ba0.5-xSrx)La0.5InO3-δ (x = 0, 0.1, 0.2) oxygen ion conductor. Previous studies reported that the ionic conductivity of Ba-doped LaInO3 decreases because Ba dopant forms a narrow oxygen path in the lattice, which could hinder the diffusion of oxygen ions. In this study, we reveal the mechanism to improve ionic conductivity by Ba and Sr co-doping on an La site in LaInO3 perovskite oxide. The results show that the ionic conductivity of (Ba0.5-xSrx)La0.5InO3-δ increases with an increasing number of Sr ions because oxygen diffusion paths which contain Sr ions have a larger critical radius than those containing Ba ions. The radial distribution function (RDF) calculations show that the peak heights in compositions including Sr ions were lower and broadened, meaning that the oxygen ions moved easily into other oxygen sites.

Oxygen deficient layered double perovskite as an active cathode for CO2 electrolysis using a solid oxide conductor.

A-site ordered PrBaMn2O(5+δ) was investigated as a potential cathode for CO2 electrolysis using a La(0.9)Sr(0.1)Ga(0.8)Mg(0.2)O3 (LSGM) electrolyte. The A-site ordered layered double perovskite, PrBaMn2O(5+δ), was found to enhance electrocatalytic activity for CO2 reduction on the cathode side since it supports mixed valent transition metal cations such as Mn, which could provide high electrical conductivity and maintain a large oxygen vacancy content, contributing to fast oxygen ion diffusion. It was found that during the oxidation of the reduced PrBaMn2O(5+δ) (O5 phase) to PrBaMn2O(6-δ) (O6 phase), a reversible oxygen switchover in the lattice takes place. In addition, here the successful CO2 electrolysis was measured in LSGM electrolyte with this novel oxide electrode. It was found that this PrBaMn2O(5+δ), layered perovskite cathode exhibits a performance with a current density of 0.85 A cm(-2) at 1.5 V and 850 °C and the electrochemical properties were also evaluated by impedance spectroscopy.

Layered oxygen-deficient double perovskite as an efficient and stable anode for direct hydrocarbon solid oxide fuel cells.

Different layered perovskite-related oxides are known to exhibit important electronic, magnetic and electrochemical properties. Owing to their excellent mixed-ionic and electronic conductivity and fast oxygen kinetics, cation layered double perovskite oxides such as PrBaCo2O5 in particular have exhibited excellent properties as solid oxide fuel cell oxygen electrodes. Here, we show for the first time that related layered materials can be used as high-performance fuel electrodes. Good redox stability with tolerance to coking and sulphur contamination from hydrocarbon fuels is demonstrated for the layered perovskite anode PrBaMn2O5+δ (PBMO). The PBMO anode is fabricated by in situ annealing of Pr0.5Ba0.5MnO3-δ in fuel conditions and actual fuel cell operation is demonstrated. At 800 °C, layered PBMO shows high electrical conductivity of 8.16 S cm(-1) in 5% H2 and demonstrates peak power densities of 1.7 and 1.3 W cm(-2) at 850 °C using humidified hydrogen and propane fuels, respectively.

Self-recovery of Pd nanoparticles that were dispersed over La(Sr)Fe(Mn)O3 for intelligent oxide anodes of solid-oxide fuel cells.

Self-recovery is one of the most-desirable properties for functional materials. Recently, oxide anodes have attracted significant attention as alternative anode materials for solid-oxide fuel cells (SOFCs) that can overcome reoxidation, deactivation, and coke-deposition. However, the electrical conductivity and surface activity of the most-widely used oxide anodes remain unsatisfactory. Herein, we report the synthesis of an "intelligent oxide anode" that exhibits self-recovery from power-density degradation in the redox cycle by using a Pd-doped La(Sr)Fe-(Mn)O(3) cell as an oxide anode for the SOFCs. We investigated the anodic performance and oxidation-tolerance of the cell by using Pd-doped perovskite as an anode and fairly high maximum power densities of 0.5 and 0.1 W cm(-2) were achieved at 1073 and 873 K, respectively, despite using a 0.3 mm-thick electrolyte. Long-term stability was also examined and the power density was recovered upon exposure of the anode to air. This recovery of the power density can be explained by the formation of Pd nanoparticles, which were self-recovered through reoxidation and reduction. In addition, the self-recovery of the anode by oxidation was confirmed by XRD and SEM and this process was effective for improving the durability of SOFC systems when they were exposed to severe operating conditions.

Doped CeO2-LaFeO3 composite oxide as an active anode for direct hydrocarbon-type solid oxide fuel cells.

Direct utilization of hydrocarbon and other renewable fuels is one of the most important issues concerning solid oxide fuel cells (SOFCs). Mixed ionic and electronic conductors (MIECs) have been explored as anode materials for direct hydrocarbon-type SOFCs. However, electrical conductivity of the most often reported MIEC oxide electrodes is still not satisfactory. As a result, mixed-conducting oxides with high electrical conductivity and catalytic activity are attracting considerable interest as an alternative anode material for noncoke depositing anodes. In this study, we examine the oxide composite Ce(Mn,Fe)O(2)-La(Sr)Fe(Mn)O(3) for use as an oxide anode in direct hydrocarbon-type SOFCs. High performance was demonstrated for this composite oxide anode in direct hydrocarbon-type SOFCs, showing high maximum power density of approximately 1 W cm(-2) at 1073 K when propane and butane were used as fuel. The high power density of the cell results from the high electrical conductivity of the composite oxide in hydrocarbon and the high surface activity in relation to direct hydrocarbon oxidation.

Chromatin regulation during C. elegans germline development.

Recent studies in Caenorhabditis elegans implicate PcG- and NuRD-like chromatin regulators in the establishment and maintenance of germline-soma distinctions. Somatic cells appear to utilize NuRD-related nucleosome-remodeling factors to overwrite germline-specific chromatin states that are specified through PcG-like activities. The germline, in turn, may rely on an asymmetrically inherited inhibitor to prevent chromatin reorganization that would otherwise erase pluripotency.

BTB proteins are substrate-specific adaptors in an SCF-like modular ubiquitin ligase containing CUL-3.

Programmed destruction of regulatory proteins through the ubiquitin-proteasome system is a widely used mechanism for controlling signalling pathways. Cullins are proteins that function as scaffolds for modular ubiquitin ligases typified by the SCF (Skp1-Cul1-F-box) complex. The substrate selectivity of these E3 ligases is dictated by a specificity module that binds cullins. In the SCF complex, this module is composed of Skp1, which binds directly to Cul1, and a member of the F-box family of proteins. F-box proteins bind Skp1 through the F-box motif, and substrates by means of carboxy-terminal protein interaction domains. Similarly, Cul2 and Cul5 interact with BC-box-containing specificity factors through the Skp1-like protein elongin C. Cul3 is required for embryonic development in mammals and Caenorhabditis elegans but its specificity module is unknown. Here we report the identification of a large family of BTB-domain proteins as substrate-specific adaptors for C. elegans CUL-3. Biochemical studies using the BTB protein MEL-26 and its genetic target MEI-1 (refs 12, 13) indicate that BTB proteins merge the functional properties of Skp1 and F-box proteins into a single polypeptide.

MEP-1 and a homolog of the NURD complex component Mi-2 act together to maintain germline-soma distinctions in C. elegans.

A rapid cascade of regulatory events defines the developmental fates of embryonic cells. However, once established, these developmental fates and the underlying transcriptional programs can be remarkably stable. Here, we describe two proteins, MEP-1 and LET-418/Mi-2, required for maintenance of somatic differentiation in C. elegans. In animals lacking MEP-1 and LET-418, germline-specific genes become derepressed in somatic cells, and Polycomb group (PcG) and SET domain-related proteins promote this ectopic expression. MEP-1 and LET-418 interact in vivo with the germline-protein PIE-1. Our findings support a model in which PIE-1 inhibits MEP-1 and associated factors to maintain the pluripotency of germ cells, while at later times MEP-1 and LET-418 remodel chromatin to establish new stage- or cell-type-specific differentiation potential.