One model to unite them all: Personalized federated learning of multi-contrast MRI synthesis.

Curation of large, diverse MRI datasets via multi-institutional collaborations can help improve learning of generalizable synthesis models that reliably translate source- onto target-contrast images. To facilitate collaborations, federated learning (FL) adopts decentralized model training while mitigating privacy concerns by avoiding sharing of imaging data. However, conventional FL methods can be impaired by the inherent heterogeneity in the data distribution, with domain shifts evident within and across imaging sites. Here we introduce the first personalized FL method for MRI Synthesis (pFLSynth) that improves reliability against data heterogeneity via model specialization to individual sites and synthesis tasks (i.e., source-target contrasts). To do this, pFLSynth leverages an adversarial model equipped with novel personalization blocks that control the statistics of generated feature maps across the spatial/channel dimensions, given latent variables specific to sites and tasks. To further promote communication efficiency and site specialization, partial network aggregation is employed over later generator stages while earlier generator stages and the discriminator are trained locally. As such, pFLSynth enables multi-task training of multi-site synthesis models with high generalization performance across sites and tasks. Comprehensive experiments demonstrate the superior performance and reliability of pFLSynth in MRI synthesis against prior federated methods.

A novel pathogenic missense variant in CNNM4 underlying Jalili syndrome: Insights from molecular dynamics simulations.

BACKGROUND: Jalili syndrome (JS) is a rare cone-rod dystrophy (CRD) associated with amelogenesis imperfecta (AI). The first clinical presentation of JS patients was published in 1988 by Jalili and Smith. Pathogenic mutations in the Cyclin and CBS Domain Divalent Metal Cation Transport Mediator 4 (CNNM4) magnesium transporter protein have been reported as the leading cause of this anomaly. METHODS: In the present study, a clinical and genetic investigation was performed in a consanguineous family of Pakistani origin, showing characteristic features of JS. Sanger sequencing was successfully used to identify the causative variant in CNNM4. Molecular dynamics (MD) simulations were performed to study the effect of amino acid change over CNNM4 protein. RESULTS: Sequence analysis of CNNM4 revealed a novel missense variant (c.1220G>T, p.Arg407Leu) in exon-1 encoding cystathionine-ß-synthase (CBS) domain. To comprehend the mutational consequences in the structure, the mutant p.Arg407Leu was modeled together with a previously reported variant (c.1484C>T, p.Thr495Ile) in the same domain. Additionally, docking analysis deciphered the binding mode of the adenosine triphosphate (ATP) cofactor. Furthermore, 60ns MD simulations were carried out on wild type (p.Arg407/p.Thr495) and mutants (p.Arg407Leu/p.Thr495Ile) to understand the structural and energetic changes in protein structure and its dynamic behavior. An evident conformational shift of ATP in the binding site was observed in simulated mutants disrupting the native ATP-binding mode. CONCLUSION: The novel identified variant in CNNM4 is the first report from the Pakistani population. Overall, the study is valuable and may give a novel insight into metal transport in visual function and biomineralization.