We are excited to announce our participation in the prestigious NeurIPS 2023 Congress, a landmark event in the field of artificial intelligence, held from December 10 to 16 at the Ernest N. Morial Convention Center in New Orleans. NeurIPS, renowned for its impact and standing alongside other AI giants like ICLR and ICML, offers a platform for researchers and practitioners to exchange groundbreaking ideas and advancements in AI.

This year, our partner, Universidad Politécnica de Madrid (UPM), represented by Alejandro Almodóvar , took center stage at the Deep Generative Models for Health Workshop on December 15. Mr. Almodovar presented a compelling poster summarizing a recently published paper, “Federated learning for causal inference using deep generative disentangled models”, that marks a significant stride in the application of deep generative models in healthcare, which is deeply related to the objectives of SYNTHEMA.

The workshop brought together a diverse group of academics, industry professionals, and healthcare experts, all united by a common goal: leveraging AI to revolutionize healthcare. The event provided an excellent opportunity for participants to engage with the latest research, methodologies, and applications of deep learning in health-related fields.

Our presence at NeurIPS 2023 underlines our dedication to staying at the forefront of AI research and development. It also highlights our commitment to fostering partnerships with leading academic institutions like UPM to drive forward the potential of AI in enhancing healthcare outcomes.

The feedback and insights gained from this event will undoubtedly propel our ongoing efforts to innovate and develop AI tools that are not only cutting-edge but also ethically responsible and beneficial to society. We look forward to continuing our journey in advancing healthcare through AI and contributing to a healthier, more informed world.

Poster presented by SYNTHEMA consortium member UPM at NeurIPS23

Publication for NeurIPS : Federated learning for causal inference using deep generative disentangled models

Authors: Alejandro Almodóvar, Juan Parras, Santiago Zazo.

Published in: Deep Generative Models for Health Workshop @ NeurIPS 2023

Abstract:  In the context of decentralized and privacy-constrained healthcare data settings, we introduce an innovative approach to estimate individual treatment effects (ITE) via federated learning. Emphasizing the critical importance of data privacy in healthcare, especially when drawing on data from various global hospitals, we address challenges arising from data scarcity and specific treatment assignment criteria influenced by the availability of the medication of interest. Our methodology uses federated learning applied to neural network-based generative causal inference models to bridge the gap between decentralized and centralized ITE estimation on a benchmark dataset

Read the full paper here.

Leave a Reply

Your email address will not be published. Required fields are marked *