Loading
Life sciences · Preprint
arXiv · September 4, 2026
Early or partial results. Treat as a signal, not a conclusion.
FedDRAW is a novel aggregation method for federated learning that dynamically weights institutional contributions using reputation derived from parameter similarity and data size, rather than size alone. In simulation across 12 institutional partition scenarios on two public chest radiograph datasets, FedDRAW ranked highest on AUC and sensitivity/specificity by geometric mean, with statistical significance confirmed by Friedman test; however, the work remains a simulation study without real multi-institutional deployment or clinical validation.
Simulation study; algorithm development and comparative benchmark. Data derived from two public chest radiograph datasets (CheXpert and ChestMNIST); artificially partitioned into simulated multi-institutional clients. No real patients or institutions; no real federated deployment.. Intervention: FedDRAW: server-side aggregation method using dual reputation annealing weighting based on data-size prior and cosine similarity of parameters with coupled annealing schedules.. Compared with: Seven federated learning baselines; implicitly includes standard FedAvg (which weights by local sample count)..
FedDRAW achieved highest average rank among eight methods (FedDRAW plus seven baselines) across both AUC and geometric mean of sensitivity and specificity metrics. Friedman test with Nemenyi post-hoc analysis confirmed statistically significant difference in performance between methods. Evaluation conducted on 12 simulated client-partition scenarios across two chest radiograph datasets (CheXpert and ChestMNIST).
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
This work is not yet clinically applicable. It proposes a methodological improvement to federated learning aggregation and demonstrates potential benefit in simulation; real-world validation across actual hospitals, with attention to institutional data quality and clinical decision-making, would be required before adoption in clinical practice.
Early-stage federated learning method evaluated on simulated multi-institutional scenarios using surrogate endpoints (AUC, sensitivity/specificity) without validation on real distributed data or clinical outcomes.
As stated by the source record.
Quoted from the source exactly as published.
This work is not yet clinically applicable. It proposes a methodological improvement to federated learning aggregation and demonstrates potential benefit in simulation; real-world validation across actual hospitals, with attention to institutional data quality and clinical decision-making, would be required before adoption in clinical practice.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
Artificial intelligence models are promising for medical diagnosis, but they require large numbers of unbiased data, which in medicine are distributed across hospitals and cannot be centralized to protect patient privacy. Federated Learning (FL) addresses this, since hospitals train one shared diagnostic model while patient data remain local. Training proceeds in communication rounds, in which each hospital trains the shared model locally and returns it to the server for merging by weighted average. This aggregation weight determines whose institutional knowledge shapes the result. Federated averaging (FedAvg) sets it in proportion to local sample count, so a small but informative hospital is permanently assigned a small influence, andl argest clients could dominate the global model even when they are less informative. We propose Federated Dual Reputation Annealing Weighting (FedDRAW), a server-side aggregation method that combines a data-size prior with the cosine similarity between client and global parameters under two coupled annealing schedules. An inner schedule shifts client reputation from the size prior towards similarity. An outer, deferred annealing schedule on the softmax inverse temperature keeps the weighting selective in the early and middle rounds and relaxes it to uniformity at convergence. We evaluate FedDRAW on 12 simulated client-partition scenarios of two chest radiograph datasets (CheXpert and ChestMNIST), against seven federated baselines under identical local training settings. FedDRAW achieved the highest average rank among all eight methods under both AUC and the geometric mean (GM) of sensitivity and specificity, which a Friedman test with Nemenyi post-hoc analysis confirmed to be a statistically significant difference between the methods. Scheduling two signals, rather than fixing the weights by sample count alone, could enable less biased diagnostic models.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.