English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Dermatological diseases pose a major threat to the global health, affectingalmost one-third of the world's population. Various studies have demonstratedthat early diagnosis and intervention are often critical to prognosis andoutcome. To this end, the past decade has witnessed the rapid evolvement ofdeep learning based smartphone apps, which allow users to conveniently andtimely identify issues that have emerged around their skins. In order tocollect sufficient data needed by deep learning and at the same time protectpatient privacy, federated learning is often used, where individual clientsaggregate a global model while keeping datasets local. However, existingfederated learning frameworks are mostly designed to optimize the overallperformance, while common dermatological datasets are heavily imbalanced. Whenapplying federated learning to such datasets, significant disparities indiagnosis accuracy may occur. To address such a fairness issue, this paperproposes a fairness-aware federated learning framework for dermatologicaldisease diagnosis. The framework is divided into two stages: In the first in-FLstage, clients with different skin types are trained in a federated learningprocess to construct a global model for all skin types. An automatic weightaggregator is used in this process to assign higher weights to the client withhigher loss, and the intensity of the aggregator is determined by the level ofdifference between losses. In the latter post-FL stage, each client fine-tuneits personalized model based on the global model in the in-FL stage. To achievebetter fairness, models from different epochs are selected for each client tokeep the accuracy difference of different skin types within 0.05. Experimentsindicate that our proposed framework effectively improves both fairness andaccuracy compared with the state-of-the-art.

Achieving Fairness in Dermatological Disease Diagnosis through Automatic Weight Adjusting Federated Learning and Personalization