Fraunhofer Institute for Integrated Circuits IISWhat if medicine didn’t just heal, but changed lives? That is one of the visions of Science Year 2026. To turn this vision into reality, applied research at Fraunhofer IIS brings together innovative medical sensor technologies and AI‑driven data analysis.
Research for the medicine of tomorrow
Research for the medicine of tomorrow
March 5, 2026 | Fraunhofer IIS contributions to the Science Year 2026
In the Digital Health and Analytics department, researchers explore how preventive measures can be designed to help people fall ill less frequently and with less severe consequences. In addition, experts use medical sensor technologies and artificial intelligence to effectively support diagnostics and therapies and to improve healthcare and patient care. The first part of the series in the online magazine for the Science Year Medicine of the Future uses the examples of integrated sensors in everyday objects and intelligent data analysis in digital pathology to illustrate how innovations developed at Fraunhofer IIS are paving the way for digital medicine – whether directly for users or as valuable tools for research.
Continuous health monitoring through integrated sensors in everyday life
The reliable acquisition of medical data forms the basis for decision-making throughout the entire healthcare process. Early detection of cardiac arrhythmias, for example, is a crucial factor in effective treatment and in preventing long-term complications. Current diagnostic approaches either allow only short monitoring periods or require the implantation of a device. Because cardiac events often occur sporadically and without symptoms, they frequently go unnoticed and may escape detection even during a 24‑hour ECG, making timely diagnosis more difficult. Two innovations address this gap: a comfortable textile-based ECG and a medically validated toilet seat allow patients to monitor their health data regularly and continuously in everyday life.
CardioTEXTIL is a mobile multichannel ECG that supports the early detection of cardiac arrhythmias.
Designed to wear – not to stick: CardioTEXTIL
CardioTEXTIL, a mobile multichannel ECG for the early detection of arrhythmias, addresses this challenge. Dry electrodes integrated into the textile provide medically validated long‑term recordings over days, weeks, or even months. Unlike conventional long‑term ECG systems, CardioTEXTIL requires no cables or adhesive electrodes and can be put on and taken off with ease. Thanks to its washable textile components, it can be worn comfortably under everyday clothing and even during sleep, enabling the early detection of arrhythmias and other cardiovascular diseases.
ECG measurement in the little sanctuary of the home
In addition, Fraunhofer researchers, in collaboration with Hamberger Medical, have developed another way to closely monitor heart health and detect irregularities at an early stage: a medically validated toilet seat with six-channel ECG measurement.
But what advantages does integrating medical sensor technology into everyday objects offer compared with commercial wearables such as smartwatches or smart rings? Low-threshold long‑term monitoring via a toilet seat – or through integrated motion sensors in a car seat – opens up close-meshed data collection to a much broader target group. After all, everyone uses the bathroom several times a day. In this way, these solutions decentralize medicine and improve patient care.
For the resting ECG, both thighs must be in contact with the lateral electrodes. The fingers are placed on the two front sensors. An app guides the user through the 30 second measurement using visual and audio instructions.
Intelligent data analysis in digital pathology: turning data into knowledge
Digital technologies are opening up new possibilities in research and diagnostics. Modern pathology laboratories digitize tissue samples into gigapixel-scale images and analyze them on screen using AI-based methods. Intelligent image analysis makes it possible to detect pathological changes early and reliably and to differentiate between subtypes – enabling precise, individualized treatment recommendations for patients.
Cell quantification of breast tissue using the MIKAIA® IHC Cell Detection App. On the right, the tissue section is shown with cell annotations; on the left, without markings. The app identifies, classifies, and quantifies cells across different categories. Users can export the results as a CSV file following a “image to Excel” principle.
Digital pathology for improved patient care
One hundred percent of cancer diagnoses are ultimately made or confirmed by pathology. However, a care gap has been emerging for years: demographic change is driving growing demand for diagnostics and faster drug development for age‑related diseases, while pathology faces a simultaneous shortage of skilled professionals. AI‑based image analysis of digitized samples promises to accelerate research and automate diagnostic workflows, thereby improving patient care. At the same time, the enormous heterogeneity of samples – across different organs, staining methods, scanners, clinical questions, as well as indications and tests – makes the development of such systems highly cost‑intensive, particularly as both medical and technical expertise are required.
MIKAIA®: image analysis for digital pathology and spatial biology
This is where MIKAIA® comes in – the image analysis software developed by Fraunhofer IIS for digital pathology and spatial biology. The software is tailored to the needs and specific applications of life‑science researchers, without requiring in‑depth technical expertise. The MIKAIA® App Center currently offers around 20 image-analysis apps that cover the entire analysis pipeline – from tissue segmentation and tumor localization to cell classification and neighborhood analyses. These apps can be flexibly combined and applied automatically to hundreds of digitized samples, generating quantitative data.
Empowering life-science research
MIKAIA® helps researchers and pathologists in particular gain new insights more quickly and efficiently. If no suitable AI model is available for a specific research question, users can train their own models using the interactive AI Author, without any programming or AI expertise. Thanks to the underlying few-shot learning approach, only a small number of training annotations per new tissue type are usually required for the AI to learn the desired distinctions.
Digital solutions for research and users: Our innovative technologies – whether integrated sensor systems embedded in everyday objects or AI‑based data analysis – are steadily driving the adoption of precision diagnostics and personalized medicine, to the benefit of patients.