Specificity-Aware Federated Learning With Dynamic Feature Fusion Network for Imbalanced Medical Image Classification.

Recently, federated learning has become a powerful technique for medical image classification due to its ability to utilize datasets from multiple clinical clients while satisfying privacy constraints. However, there are still some obstacles in federated learning. Firstly, most existing methods directly average the model parameters collected by medical clients on the server, ignoring the specificities of the local models.

Toward Multicenter Skin Lesion Classification Using Deep Neural Network With Adaptively Weighted Balance Loss.

Recently, deep neural network-based methods have shown promising advantages in accurately recognizing skin lesions from dermoscopic images. However, most existing works focus more on improving the network framework for better feature representation but ignore the data imbalance issue, limiting their flexibility and accuracy across multiple scenarios in multi-center clinics. Generally, different clinical centers have different data distributions, which presents challenging requirements for the network's flexibility and accuracy.

Technical Note: An anthropomorphic phantom with implanted neurostimulator for investigation of MRI safety.

Purpose: The objective of this work was to design and construct an improved anthropomorphic phantom for use in studying magnetic resonance imaging (MRI) radiofrequency (RF) safety at 3 T related to deep brain stimulation (DBS), and especially the role of DBS lead trajectories.

Method: Based on a computer-aided design including reasonable representation of human features, the phantom was fabricated by three-dimensional (3D) printing and then fully assembled with a human skull, a commercial DBS device implanted using the surgical standard at our institution, and fiber-optic temperature sensors embedded in two tissue mimicking solutions (e.g.

Inflow effects on hemodynamic responses characterized by event-related fMRI using gradient-echo EPI sequences.

The purpose of this study is to determine whether blood inflow impacts the temporal behavior of BOLD-contrast fMRI signal changes in a typical event-related paradigm. The inflow contributions in the hemodynamic response to repeated single trials of short visual stimulation were assessed with a gradient-echo EPI sequence by altering the flip angle (FA) from 30 degrees to 90 degrees at a repetition time of 1 s. For each FA condition (30 degrees, 60 degrees, and 90 degrees), 30 trials were performed on 15 healthy volunteers on a 3T MRI scanner.