Swarm learning for decentralized artificial intelligence in cancer histopathology

Abstract

Artificial Intelligence (AI) can extract clinically actionable information from medical image data. In cancer histopathology, AI can be used to predict the presence of molecular alterations directly from routine histopathology slides. However, training robust AI systems requires large datasets whose collection faces practical, ethical and legal obstacles. These obstacles could be overcome with swarm learning (SL) where partners jointly train AI models, while avoiding data transfer and monopolistic data governance. Here, for the first time, we demonstrate the successful use of SL in large, multicentric datasets of gigapixel histopathology images comprising over 5000 patients. We show that AI models trained using Swarm Learning can predict BRAF mutational status and microsatellite instability (MSI) directly from hematoxylin and eosin (H&E)-stained pathology slides of colorectal cancer (CRC). We trained AI models on three patient cohorts from Northern Ireland, Germany and the United States of America and validated the prediction performance in two independent datasets from the United Kingdom using SL-based AI models. Our data show that SL enables us to train AI models which outperform most locally trained models and perform on par with models which are centrally trained on the merged datasets. In addition, we show that SL-based AI models are data efficient and maintain a robust performance even if only subsets of local datasets are used for training. In the future, SL can be used to train distributed AI models for any histopathology image analysis tasks, overcoming the need for data transfer and without requiring institutions to give up control of the final AI model.

An Investigation in Analyzing the Food Quality Well-Being for Lung Cancer Using Blockchain through CNN

“…Nevertheless, there may be few backlogs in AI while dealing with the healthcare like it requires human scrutiny, may oversee societal variables, and may lead to job loss. As per Le and Hsu, the application of the model with a 98.04% sensitivity level can be highly effective in reducing human errors by 99.56% [20]. Moreover, with an accuracy rate of 97.23%, researchers can e ectively investigate the use of blockchain facilities in lung cancer prediction for a sustainable healthcare service in the future.…”

Encrypted federated learning for secure decentralized collaboration in cancer image analysis

“…For the histopathological use-case, less than 5% of time was used for decryption and encryption (which happens at edge) and homomorphic aggregation of the weights (which happens at the central server, Figure 3b). This difference is due to the different network architectures and different number of parameters: the histopathological use-case employs a fixed backbone feature extractor 25 and thus has fewer parameters to optimize. Encryption and decryption scales approximately linear with the number of weights to be updated, while neural network training complexity scales more than linearly in our setup.…”

“…For the histopathological data we collected digital whole slide images (WSI) of H&E-stained slides of human colorectal cancer (CRC) from five patient cohorts, three of which were used as training cohorts and two of which were used as test cohorts following the division of data in a previous study 25 . The training cohorts are representative of real-world clinical settings.…”

Swarm learning for decentralized healthcare

“…Such an application uses the same underlying techniques as dermato-logical applications as demonstrated in [1]. Swarm learning has also been successfully applied by other groups [8] in analyzing pathology images.…”