Enhancing human perception of the imperceptible - 28/08/2025

AI-driven imaging expands possibilities in surgery

Doctors frequently experience a restricted view of the surgical field, particularly during endoscopic procedures. A novel technique developed by Prof. Dr. Lena Maier-Hein's team at the DKFZ in Heidelberg integrates spectral imaging with AI-driven data analysis. This innovation facilitates accurate tissue differentiation and delivers real-time insights into organ function.

If the brain no longer responds properly to insulin (insulin resistance), this can lead to overweight, diabetes, and Alzheimer's disease. Researchers at the DZD in Potsdam and Tübingen have discovered small chemical modifications to genetic material (epigenetic changes*) in the blood that indicate how well the brain responds to insulin. These markers could help to detect insulin resistance in the brain – by means of a simple blood test.

Designing protein binders from scratch has long been a daunting challenge within the field of computational biology. Researchers have now developed an innovative, training-free pipeline that uses the fundamental principle of shape complementarity to design site-specific protein binders, which are then optimised to fit precisely onto chosen target sites. The researchers tested this on proteins linked to cancer.

A team of researchers from HITS and MPIP have developed a model that learns how to generate proteins whose structures are highly flexible, even with patterns that are uncommon in natural proteins. Their work, presented at the International Conference on Machine Learning (ICML), marks a step towards the goal of designing new proteins for applications in biotechnology, therapeutics and environmental research.

AI-supported image recognition accelerates identification of zebrafish mutants - 16/07/2025

EmbryoNet AI automatically identifies developmental disorders

In complex organisms, embryonic development is tightly regulated by intricate signalling pathways. When these pathways are disrupted, they can lead to characteristic developmental defects that are not easy to detect with the naked eye. Developmental biologist Prof. Dr. Patrick Müller from the University of Konstanz has developed EmbryoNet, an AI-powered software tool that uses image analysis to reliably identify such developmental disorders.

Tübingen achieves remarkable success and has good chance of maintaining its University of Excellence title – Top research in three areas to be sustained from other sources of support.

A building with 7400 m² of research space for robotics and artificial intelligence is being built on the Max Planck Campus in Büsnau. Researchers from the Max Planck Institute for Intelligent Systems (MPI-IS) and the University of Stuttgart will be working there with partners from science and industry as well as start-ups from 2027.

Baden-Württemberg AI Alliance - 08/04/2025

Achieving widespread use of AI

Artificial intelligence presents both opportunities and risks. The Baden-Württemberg AI Alliance is dedicated to fostering collaboration among AI stakeholders to ensure that AI solutions deliver tangible benefits to citizens and businesses across the state. The alliance will also place greater emphasis on advancing AI applications within the healthcare sector.

A recent article by Prof. Dr. Martin Haimerl and Prof. Dr. Christoph Reich of Furtwangen University shows that machine learning (ML) in medicine is often evaluated without a comprehensive risk assessment. The authors investigated the extent to which current scientific papers include risk-based metrics in the evaluation of AI models for medical devices.

The MedTech company KARL STORZ announces the acquisition of the innovative software manufacturer Innersight Labs Ltd. (ISL) headquartered in London. In addition to state-of-the-art endoscopes, high-end medical devices, and integrated solutions for the operating room, KARL STORZ is also continuing to expand in innovative software solutions. Its large customer base, including top-class physicians worldwide, can now look forward to 3D models.

Vicinity Bio: Optimisation of cancer diagnostics - 27/11/2024

Comprehensive histological diagnostics through high-dimensional imaging and artificial intelligence

Microscopic examination of tissue samples is essential, particularly in tumour diagnostics. The Tübingen-based company Vicinity Bio leverages cutting-edge imaging technologies combined with machine learning to generate comprehensive datasets of individual cells from tissue sections. This approach not only helps identify more targeted therapies but also enhances our understanding of cellular functions and processes within tissues and tumours.

The reproducible and precise production of complex organoid models to simulate human organ malfunctions is the focus of an interdisciplinary research project at Heidelberg University. A research team from the life and engineering sciences is looking to combine the engineering of molecular systems with machine learning and automated production methods.

Whether we are predisposed to particular diseases depends to a large extent on the countless variants in our genome. However in the case of genetic variants the influence on the presentation of certain pathological traits has been difficult to determine. Researchers have introduced an algorithm based on deep learning that can predict the effects of rare genetic variants.

Five transnational and cross-institutional doctoral networks at Heidelberg University are being funded as part of the “Marie Skłodowska-Curie Actions”. They work together on current scientific topics with high innovation potential. Ruperto Carola coordinates an MSCA Doctoral Network on artificial intelligence in physics, two networks in medicine, life sciences and engineering.

Scientists have long sought ways to simulate the neural networks in the brain with computers in order to understand how it works. Now, researchers have combined new measurements of the wiring diagrams of the fruit fly with artificial intelligence methods to build a neural network that can do what few thought possible: To predict the activity of individual neurons without making a single measurement in a living brain.

At the finale of the Cyber Valley Days, and Science Minister Petra Olschowski opened the ELLIS Institute Tübingen, heralding the next phase of the AI Innovation Campus. With the world's first ELLIS Institute, Cyber Valley is gaining further radiance. Olschowski also announced at the ceremony that the Karlsruhe Institute of Technology (KIT) is joining the Cyber Valley Community.

The Damietta Server broadens the accessibility to protein design research and its applications in various biotechnological and biomedical fields. Researchers at the Max Planck Institute for Biology Tübingen, the University of Tübingen, and the University Hospital Tübingen have developed a web-based toolkit to accelerate and simplify protein design without needing powerful computers or extensive protein design expertise on the user’s end.

Since 2019, the Research Training Group (RTG) "Intraoperative Multisensoric Tissue Differentiation in Oncology," a collaboration between the University of Stuttgart and the University of Tübingen, has been advancing research in medical technology. With the help of new sensor methods, the researchers want to help make surgical procedures on cancer patients safer and more effective.

An Emmy Noether junior research group at Heidelberg University is investigating how to gain new insights into fundamental biological mechanisms from large-scale molecular data sets. Led by Junior Professor Dr Britta Velten, it has started work at the Centre for Organismal Studies and the Interdisciplinary Center for Scientific Computing.

Making a personalised T cell therapy for cancer patients currently takes at least six months; scientists at the German Cancer Research Center (DKFZ) and the University Medical Center Mannheim have shown that the laborious first step of identifying tumor-reactive T cell receptors for patients can be replaced with a machine learning classifier that halves this time.

Team involving the University of Freiburg has developed a new analytical method for the basement membrane in human lungs.

Working with a new Emmy Noether Group, Dr. Pascal Schlosser is investigating how machine learning can aid in understanding the complex relationships between genes, proteins, and diseases.

The family-owned MedTech company KARL STORZ announces the acquisition of the innovative software manufacturer Innersight Labs Ltd. (ISL) headquartered in London. In addition to state-of-the-art endoscopes, high-end medical devices, and integrated solutions for the operating room, KARL STORZ is also continuing to expand in innovative software solutions.

Current diagnostic methods do not always reliably distinguish between chronic inflammation of the pancreas and pancreatic cancer. About one third of all diagnoses are inconclusive. Scientists from the German Cancer Research (DKFZ) and from Heidelberg University Hospital (UKHD) therefore searched for molecular markers that could specify this diagnosis.

The computational design of new proteins for biomedical or other applications involves long computing times on powerful servers. A joint team of researchers from the Max Planck Institute for Biology Tübingen and the University Hospital Tübingen has now developed and tested a new computational method to greatly speed up the necessary energy calculations. Their framework allows for a precise and efficient design of functional proteins.