This research paper explores ways to apply Federated Learning (FL) and Differential Privacy (DP) techniques to population-scale Electrocardiogram (ECG) data. The study learns a multi-label ECG classification model using FL and DP based on 1,565,849 ECG tracings from 7 hospitals in Alberta, Canada. The FL approach allowed collaborative model training without sharing raw data between hospitals, while building robust ECG classification models for diagnosing various cardiac conditions. These accurate ECG classification models can facilitate the diagnoses while preserving patient confidentiality using FL and DP techniques. Our results show that the performance achieved using our implementation of the FL approach is comparable to that of the pooled approach, where the model is trained over the aggregating data from all hospitals. Furthermore, our findings suggest that hospitals with limited ECGs for training can benefit from adopting the FL model compared to single-site training. In addition, this study showcases the trade-off between model performance and data privacy by employing DP during model training. Our code is available at https://github.com/vikhyatt/Hospital-FL-DP.
Since its discovery in the early 1980s, there has been significant progress in understanding the biology of type 1 human immunodeficiency virus (HIV-1). Structural biologists have made tremendous contributions to this challenge, guiding the development of current therapeutic strategies. Despite our efforts, there are unresolved structural features of the virus and consequently, significant knowledge gaps in our understanding. The superstructure of the HIV-1 matrix (MA) shell has not been elucidated. Evidence by various high-resolution microscopy techniques support a model composed of MA trimers arranged in a hexameric configuration consisting of 6 MA trimers forming a hexagon. In this manuscript we review the mathematical limitations of this model and propose a new model consisting of a 6-lune hosohedra structure, which aligns with available structural evidence. We used geometric and rotational matrix computation methods to construct our model and predict a new mechanism for viral entry that explains the increase in particle size observed during CD4 receptor engagement and the most common HIV-1 ellipsoidal shapes observed in cryo-EM tomograms. A better understanding of the HIV-1 MA shell structure is a key step towards better models for viral assembly, maturation and entry. Our new model will facilitate efforts to improve understanding of the biology of HIV-1.
The existing deterministic neighbor discovery algorithms in WSNs ensure that successful discovery can be obtained within a certain time, but the average discovery delay is long. It is difficult to meet the need for rapid discovery in mobile low duty cycle environments. This paper proposes a group-based fast neighbor discovery algorithm (GBFA). By carrying neighbor information in beacon packet, GBFA knows in advance some potential neighbors and select more energy efficient potential neighbors and actively make nodes wake up to verify whether these potential neighbors are true neighbors, thereby speeding up neighbor discovery and improving energy utilization efficiency. The evaluation results indicate the mean delay and extra energy consumption of the GBFA algorithm are lower than Group-based method up to \(10.58\%\) and \(18.88\%\) respectively.