Building TrustworthyDigital Twins for Patient-Centred Medicine
Follow us on LinkedIn

Berliner Hochschule für Technik/ Berlin University of Applied Science

BHT (Berlin University of Applied Sciences and Technology) is one of Germany’s leading universities of applied sciences, combining academic excellence with strong industry collaboration. Active in national and European research projects, BHT focuses on digital transformation, engineering, sustainability, and applied innovation.

The Data Science +X Research Center, directed by Alexander Löser, is BHT’s interdisciplinary AI hub, bringing together 13 professors and more than 35 doctoral researchers from machine learning, mathematics, health, and engineering. The center develops trustworthy and explainable AI solutions for digital health, Industry 4.0, and biotechnology, with extensive experience in EU-funded collaborative projects.

DATEXIS (Data Science and Text-based Information Systems) is BHT’s research group for modular and explainable language models. The group contributes to national initiatives such as Soofi, one of Europe’s flagship language models, and conducts research on modular AI architectures, explainability, and self-improving learning systems, particularly for medical applications.

Role within TWIN-X

BHT (Berliner Hochschule für Technik) is a key AI methodology partner in TWIN-X, contributing advanced research on clinical reasoning, foundation models, and trustworthy Generative AI. Its primary leadership role is in WP2, where BHT develops reasoning models that infer missing clinical facts from structured and unstructured medical data, including disease staging, risk assessment, and evidence-based clinical inference. BHT also contributes enhancing language understanding through LLMs and semi-automated prompting strategies for clinical text processing. Across the project, BHT leads activities related to clinical generative reasoning and test-time compute. Its expertise in foundation models and reasoning architectures strengthens TWIN-X’s ability to deliver transparent and clinically meaningful decision-support tools.

Main contacts

Photo of Prof. Dr. Alexander Löser
Prof. Dr. Alexander Löser
Photo of Prof. Dr. Amy Siu
Prof. Dr. Amy Siu