Disentangled Representation Learning for Privacy-Preserving Case-Based Explanations

• Tipo: Artigo em Livro de Atas de Conferência Internacional
• Data: 2023

Autores

Montenegro, H (Autor) Outra; Silva, W (Autor) Outra; Jaime S Cardoso (Autor) FEUP

Ata de Conferência Internacional

Título: MEDICAL APPLICATIONS WITH DISENTANGLEMENTS, MAD 2022

Páginas: 33-45

1st Workshop on Medical Applications with Disentanglements (MAD) at the MICCAI Conference

Singapore, SINGAPORE, SEP 22, 2022

Indexação

ISI Web of Knowledge - 0 Citações

Scopus

Outras Informações

ID Authenticus: P-00X-XKK

DOI: 10.1007/978-3-031-25046-0_4

Abstract (EN): The lack of interpretability of Deep Learning models hinders their deployment in clinical contexts. Case-based explanations can be used to justify these models' decisions and improve their trustworthiness. However, providing medical cases as explanations may threaten the privacy of patients. We propose a generative adversarial network to disentangle identity and medical features from images. Using this network, we can alter the identity of an image to anonymize it while preserving relevant explanatory features. As a proof of concept, we apply the proposed model to biometric and medical datasets, demonstrating its capacity to anonymize medical images while preserving explanatory evidence and a reasonable level of intelligibility. Finally, we demonstrate that the model is inherently capable of generating counterfactual explanations.

Idioma: Inglês

Tipo (Avaliação Docente): Científica

Nº de páginas: 13