Fraunhofer-Gesellschaft
Fraunhofer-Gesellschaft
""
© Getty Images

© Getty Images

Data – Intelligent Analysis of Large Amounts of Data

Gaining Knowledge from Data Without Losing Trust

Today, health data is often fragmented, difficult to access or incompatible with each other. Its enormous potential for improved diagnoses and therapies remains untapped. Fraunhofer combines medical data from clinics, research and care in secure data spaces and uses AI methods such as digital twins and in-silico models. Sensitive data remains protected and under patients’ control.

Project examples (selection)

 

© Fraunhofer ITEM / Ralf Mohr

Toxicology: Next-Generation Risk Assessment

Researchers are developing predictive models that enable the safety of substances to be assessed at an early stage. In vitro and in silico methods as well as modern modeling are making testing and evaluation of substances safer, more efficient and more sustainable. 

 

Framework Program for Synthetic Data Validation in Healthcare

The development of AI-supported applications that are secure, reproducible and scalable requires the availability of sufficient real data. SYNTHIA is developing the basis for generating realistic, multimodal synthetic data and is making it available via an open collaborative platform, including validated data sets, workflows and evaluation data.

Structuring Medical Data

Most medical information is present in unstructured PDF format. A software service aims to automatically structure medical data using large language models (LLMs) to reduce the workload on specialists, minimize errors and make data usable for care and research. 

AI App Boosts Women's Health during Menopause

The PETRA project is developing an AI-powered app that provides women with personalized information, recommendations and digital therapy support during perimenopause. This strengthens women's health literacy, alleviates specific symptoms and provides new insights for gender-equitable medical research.

 

© Fraunhofer IMTE / O. Malzahn

Test Platforms for Safe Automation in Medical Technology

Computer-simulated physiological patient models can realistically reproduce highly complex bodily mechanisms in digital form. These models quantify expert medical knowledge and combine it with data from hospitals or the literature to enable personalized analyses through patient-specific adaptations. In addition to digital simulation, physical models are being developed, such as a lung simulator as a physical interface for ventilators. 

AI Platform for Personalized Early Detection of Dementia

Researchers are developing a freely available, interoperable AI platform that combines biomarkers, samples from the home environment and health data from over 4,000 individuals to screen for individual dementia risk at an early stage. 

Clinical Trial Management System

The Digital Trial manager (DTM) is an all-in-one solution facilitating GCP-compliant study management for CROs, digital health application providers (DiGAs), medical device manufacturers, hospitals and research institutes. It is individually configurable and enables effective cross-location networking.

 

© Fraunhofer MEVIS

Smart Data Curation for AI and Image Analysis

CuraMate is a tool for the efficient curation of medical image data and supports annotating, segmenting and quantifying image features. It enables seamless collaboration on studies and continuously improves AI models through integrated training and correction loops. 

Efficient Doctor's Letters for Modern Healthcare

The AI-supported doctor's letter generator acts as a co-writer, assisting in the structured creation of discharge letters, suggesting intelligent text passages and integrating securely and interoperably in existing hospital systems. This gives medical personnel more time for their patients.

Digital Ecosystem for Rural Healthcare

Access to modern diagnostics is often limited in rural areas. The Neighborhood Diagnostics project provides a digital ecosystem that uses an app to bundle data from wearables and home monitoring devices, establishing a sound basis for medical decisions. 

Contactless Monitoring of Vital Functions

AI-supported contactless sensor technology is designed for reliable remote measurement of vital parameters such as respiratory rate, pulse and body temperature. This provides a potential early warning system for changes in the condition of hospital patients. 

 

© Ina Fassbender

Smart Hospital

In a showroom, we demonstrate AI-based applications for diagnosis and data analysis, including automated documentation, which relieve the burden on hospitals and accelerate the path to smart hospitals. 

 

© Fraunhofer IIS

Software for Preclinical and Clinical Research

Manual analysis of pathological image data is time-consuming and subjective and is often not scalable for modern research questions such as biomarker detection. An image analysis multi-tool for digital pathology and spatial biology makes clinical and preclinical research more efficient and faster.

Innovative Technologies for Brain Regeneration

Fraunhofer is researching new ways to repair damaged brain tissue. The method uses epigenetic tools to convert brain cells into functional neurons. This is a promising approach for diseases such as strokes or neurodegenerative diseases, for which there are currently only limited treatment options. 

Improved Procedures for Brain Tumors

Intelligent ultrasound aspirators are designed to make neurosurgical interventions for brain tumors more precise and safer. Sensory feedback and AI-supported tissue analysis improve the differentiation between healthy and tumorous tissue and promote function-preserving operations. 

AI Reduces Fatigue in Cancer Patients

Fatigue one of the many side effects of cancer and has been difficult to measure until now. The FAITH project combines AI models, smartwatch data and an app to track individual fatigue levels and provide personalized recommendations for exercise and behavior patterns. The goals: greater self-determination and better metrics for therapy decisions.

Care information through medical data intelligence

The aim of the IMMUVision project is to create a digital platform for the care of patients with immune-mediated inflammatory systemic diseases. A digital twin compares patient-specific health data such as diagnoses, disease progression and therapies with data from population studies and guidelines.

European Health Data Space for Improved Diagnoses and Treatments

A European data space is designed to securely link health data from hospitals, research institutions and healthcare providers. The goal is faster analysis, earlier diagnosis and personalized treatment while retaining data sovereignty.

Further information

 

The Medical Data Space – a secure data space for medicine

Security in the transmission and storage of sensitive data is becoming increasingly important. That is why Fraunhofer is working with partners in the healthcare sector to develop the Medical Data Space. Linking different types of medical data on a single IT platform can bring major benefits for diagnosis and treatment.

  • more info (medical-data-space.fraunhofer.de)

Artificial intelligence in medicine

Applying and developing artificial intelligence (AI) for the medical sector is a key focus of the Fraunhofer-Gesellschaft. With its expertise, Fraunhofer creates innovative solutions that support various stakeholders in the medical sector and significantly improve the quality of healthcare services.