Swarm Learning as a privacy-preserving machine learning approach for disease classification

Abstract

<jats:title>Abstract</jats:title><jats:p>Identification of patients with life-threatening diseases including leukemias or infections such as tuberculosis and COVID-19 is an important goal of precision medicine. We recently illustrated that leukemia patients are identified by machine learning (ML) based on their blood transcriptomes. However, there is an increasing divide between what is technically possible and what is allowed because of privacy legislation. To facilitate integration of any omics data from any data owner world-wide without violating privacy laws, we here introduce Swarm Learning (SL), a decentralized machine learning approach uniting edge computing, blockchain-based peer-to-peer networking and coordination as well as privacy protection without the need for a central coordinator thereby going beyond federated learning. Using more than 14,000 blood transcriptomes derived from over 100 individual studies with non-uniform distribution of cases and controls and significant study biases, we illustrate the feasibility of SL to develop disease classifiers based on distributed data for COVID-19, tuberculosis or leukemias that outperform those developed at individual sites. Still, SL completely protects local privacy regulations by design. We propose this approach to noticeably accelerate the introduction of precision medicine.</jats:p>

Blockchain technologies to mitigate COVID-19 challenges: A scoping review

“…Another prominent application of blockchain—proposed or developed—in 4 studies was immunity passports [26] , [27] , [28] , [29] , which show that an individual is disease risk-free because he has already been infected with COVID-19 or was given a COVID-19 vaccine; therefore, he can travel without restrictions. Two studies employed blockchain for diagnosing COVID-19 based on blood samples [30] or computed tomography (CT) images [31] . The remaining applications of blockchain were used for telemedical laboratory services [32] , social distancing [33] , securely sharing patients’ data [34] , monitoring isolated people [35] , tracking the COVID-19 status as reported by trusted sources (number of new cases, deaths and recovered cases) [36] , tracking and controlling the delivery of a COVID-19 vaccine [37] , and identity verification, record attestation, and record sharing [38] .…”

“…While the blockchain technology was developed in 10 studies [ 22 , 23 , 25 , [28] , [29] , [30] , [31] , [32] , 35 , 36 ], its use was proposed in the remaining 9 studies. The type of access rights in using blockchain was public in 10 studies [ 20 , 22 , 24 , 25 , 27 , 29 , 33 , [36] , [37] , [38] ], private in 5 studies [ 23 , 26 , 30 , 31 , 34 ], hybrid in 3 studies [ 21 , 28 , 35 ], and both public as well as hybrid in 1 study [32] .…”

A tool for federated training of segmentation models on whole slide images

“…Approaches such as peer-to-peer federated learning 36 and swarm learning 37 offer data synchronization strategies that do not require a central coordina�ng (master) server. While the lack of centralized training coordina�on may be beneficial for some tasks, we argue that for federated medical image segmenta�on, it is likely that only one group will be responsible for model development.…”

Blockchains for COVID-19 Contact Tracing and Vaccine Support: A Systematic Review

“…In this section we briefly introduce some further blockchainbased applications for mitigating COVID-19 consequences. Some interesting proposals [74], [75] combine machine learning and blockchain to define a federated or swarm learning approach. [74] proposes a blockchain based federated learning framework to train and share a collaborative model.…”