Semi-Supervised Federated Peer Learning for Skin Lesion Classification

Abstract

Skin cancer is one of the most deadly cancers worldwide. Yet, it can be reduced by early detection. Recent deep-learning methods have shown a dermatologist-level performance in skin cancer classification. Yet, this success demands a large amount of centralized data, which is oftentimes not available. Federated learning has been recently introduced to train machine learning models in a privacy-preserved distributed fashion demanding annotated data at the clients, which is usually expensive and not available, especially in the medical field. To this end, we propose FedPerl, a semi-supervised federated learning method that utilizes peer learning from social sciences and ensemble averaging from committee machines to build communities and encourage its members to learn from each other such that they produce more accurate pseudo labels. We also propose the peer anonymization (PA) technique as a core component of FedPerl. PA preserves privacy and reduces the communication cost while maintaining the performance without additional complexity. We validated our method on 38,000 skin lesion images collected from 4 publicly available datasets. FedPerl achieves superior performance over the baselines and state-of-the-art SSFL by 15.8%, and 1.8% respectively. Further, FedPerl shows less sensitivity to noisy clients.

Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries

“…While the bulk of the papers we've reviewed so far focus purely on designing federated algorithms that can predict different aspects of cancer with high degrees of accuracy, a large sub-group of the papers in our review also aim at addressing challenges federated learning currently faces. For many papers, that challenge is either data heterogeneity [58][59][60][61][62][63][64][65], a common barrier in the medi-cal field where patients can be subject to different geographic and demographic conditions, or label deficiency [66,67], where it is not always guaranteed that clients' sites will have access to labeled data.…”

A Review of Medical Federated Learning: Applications in Oncology and Cancer Research

“…While the bulk of the papers we've reviewed so far focus purely on designing federated algorithms that can predict different aspects of cancer with high degrees of accuracy, a large sub-group of the papers in our review also aim at addressing challenges federated learning currently faces. For many papers, that challenge is either data heterogeneity [58][59][60][61][62][63][64][65], a common barrier in the medi-cal field where patients can be subject to different geographic and demographic conditions, or label deficiency [66,67], where it is not always guaranteed that clients' sites will have access to labeled data.…”

“…Their method is inspired by knowledge distillation [68], where they model disease relationships in each client by a relation matrix calculated from the local model output, then aggregate the relation matrices from all clients to form a global one that is used locally in each round to ensure that clients will have similar disease relationships. In [67], the authors proposed a semi-supervised Federated Learning method, FedPerl. The method was inspired by peer learning from educational psychology and ensemble averaging from committee machines and aims to gain extra knowledge by learning from similar clients i.e.…”

Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries

“…While the bulk of the papers we've reviewed so far focus purely on designing federated algorithms that can predict different aspects of cancer with high degrees of accuracy, a large sub-group of the papers in our review also aim at addressing challenges federated learning currently faces. For many papers, that challenge is either data heterogeneity [58][59][60][61][62][63][64][65], a common barrier in the medi-cal field where patients can be subject to different geographic and demographic conditions, or label deficiency [66,67], where it is not always guaranteed that clients' sites will have access to labeled data.…”