Federated Deep Learning Architecture for Personalized Healthcare

Abstract

Using deep learning to advance personalized healthcare requires data about patients to be collected and aggregated from disparate sources that often span institutions and geographies. Researchers regularly come face-to-face with legitimate security and privacy policies that constrain access to these data. In this work, we present a vision for privacy-preserving federated neural network architectures that permit data to remain at a custodian’s institution while enabling the data to be discovered and used in neural network modeling. Using a diabetes dataset, we demonstrate that accuracy and processing efficiencies using federated deep learning architectures are equivalent to the models built on centralized datasets.

Small target detection from DIKW model perspective

“…Different receptive fields under different layers. [19] In addition to the most common convolution, there are empty convolution, upsampling, BatchNorm [21] SeparableConv [22]receptive field; Receptive field is not the only information factor that affects the accuracy. Receptive field is also affected by information such as network depth, width, residual connection, BatchNorm, etc.…”

Novel technical and privacy-preserving technology for artificial intelligence in ophthalmology

“…Studies have consistently shown that FL models outperformed traditional single-institution ML architectures [148,[151][152][153][154][155][156] and may be comparable to models built via central learning (centralized database) [137,138,[143][144][145][146][147][148][149][150]. In some studies, the FL approach has even been shown to be superior to alternative collaborative learning methods [137,145].…”