Save the Date: Research Symposium 2024 Mark this day in your calendars! On 31 October CAIM will be holding its research symposium at Auditorium Langhans. We will hear the project reports of our funded projects 2022-2024 and get to know our new CAIM Fellows. Also, look forward to our invited keynote by Daniel Rückert, TUM! More info and registration...
"Our field is very open to AI." Ekin Ermis is a Senior physician at the Department of Radiation Oncology of the Inselspital, Bern University Hospital, where she validates AI-based automation for the radiotherapy pipeline developed by the University of Bern. Over the last seven years, she has dedicated research and interdisciplinary dialogue to improving workflows and assuring sooner and high-quality treatment for patients with brain tumors. Read the interview...
5th Tele Emergency Medicine Congress Virtual Congress with presentations on future hospital structures and patient care, the role of AI in medical education and decision making as well as assuring balanced data sets. Topics this year are "Future Patient Care and AI". Learn more and register...
Overcoming rotation challenges in image segmentation Most segmentation networks used in medical image analysis rely on standard convolutional kernels which require expensive rotational data augmentation in order to learn 3D rotation equivariance. The Support Center for Advanced Neuroimaging has proposed a segmentation network with SO(3)-steerable convolutions that is robust to data poses not seen during training and improves parameter sharing, leading to smaller network sizes and better sample efficiency. Learn more...
"We use AI in everyday clinical routine." Simon Steppacher specializes in hip joint preservation in adolescents and young adults at the Department of Orthopaedic Surgery and Traumatology of the Inselspital. As his patients’ lives yet lie ahead of them, he uses artificial intelligence for detailed insights into pre-surgery imaging for the best outcomes. Interview on his research...
Five CAIM Fellows chosen CAIM is championing the integration of AI into healthcare by supporting young researchers and their projects through fellowships. From a competitive pool of 19 applications received in the fall of 2023, five promising projects have been selected in a multi-step process including external and internal review, pitches, and final evaluations to award from both clinical and technological viewpoints. Projects span anaesthesiology, cancer care, eye care, and cardiology and will each receive up to CHF 100,000 over the next two years, beginning in June. More about the five projects...
How do we feel tomorrow? Doctors could soon be asking artificial intelligence this question. Because data can be used to predict disease progression more accurately. This could save money and, more importantly, precious time. In an article on the future of healthcare with AI, Insa Schiffmann also interviewed CAIM Director Raphael Sznitman and Claus Beisbart, Co-Lead of the CAIM Ethics Lab. Go to the article (in German)...
AI and Doctor: Best Buddies? Only a few radiologists, if any, are still afraid that AI will steal their jobs. But does this mean that AI has to become a "best buddy"? In an interview with VISUS, radiologist Piotr Radojewski outlines how AI is currently being applied in his field, what its present limitations are and what the future could look like. Read the interview (in German)...
How Bernese AI is transforming medicine Hundreds of researchers are working on AI for medicine in Bern's laboratories. "Our centre aims to bring cutting-edge medical research into practice as quickly as possible," says Raphael Sznitman in an interview with der Bund about CAIM. The article not only mentions the Master's program AI in Medicine, but also two AI spin-offs from the University of Bern for ophthalmology. Read the article here (ABO)...
New AI Microlearning Platform for Clinicians In an educational collaborative between the University of Bern, the Inselspital, sitem-insel and Bayer, Bern is offering the first self-learning platform for medical professionals on how to integrate AI into clinical practice and clinicians into AI. The new online course features 5 modules of video units on AI fundamentals, development and validation, AI clinical application, AI integration as well as governance and ethics. Learn more and register here...
Gender gap that costs lives Medicine is still partially blind, drugs and treatments do not help women as well as men. Women are still underrepresented in clinical trials. As a result, there is a lack of data on how they react to a certain medication. The consequences can be serious side effects. Women also have a higher risk than men of not receiving optimal treatment for cancer. In an interview with Hauptstadt, pharmacist Enriqueta Vallejo-Yagüe and oncologist Berna Özdemir explain how they want to change this. Read the article (in German)...
Can artificial intelligence be unbiased? On 7 March, Gülser Corat, former Director for Gender Equality at UNESCO, joined us for a a symposium on biases at the University of Bern. She argued that the current design of artificial intelligence often promotes bias and explained how that could change. Read the interview...
CAIM Policy Talks Three-part series of lunch-time talks on the transition of digital technology into political discourse: 21 March: From Innovation to Policy 18 April: Digitalization and Health Policy Processes 30 May: Narrative, Science and Policy Support Learn more and register here...
"The loss of humanity would be a real problem" The use of artificial intelligence appears to be a solution in times of care shortages. But how should it be judged from an ethical perspective? An interview with the philosopher of science Prof. Claus Beisbart from the Institute of Philosophy at the University of Bern. Read the interview (in German, French or Italian) in Spitex magazine...
Avoiding Bias in AI: Combating Inequality in Data, Gender and Socioeconomics Does AI reproduce bias in training data? Can machine learning overcome gender bias in medicine and provide equal access to care? With former UNESCO Equality Director Saniye Gülser as a keynote speaker we are even questioning if data-driven machine learning is still the right paradigm to operate with. Don't miss the next Spring Seminar of our initiative Diversity for AI in Medicine on 7 March 2024! Learn more...
Interpretable ML system for colorectal cancer diagnosis With engineers and clinicians from Portugal, the Digital Pathology group of the Institute for Tissue Medicine and Pathology, University of Bern, has developed a scalable deep learning system to diagnose colorectal cancer from pathology images. The system learns from weak labels, leverages a small subset of fully annotated samples, and the prototype includes explainable predictions and active learning. Developed and tested with one of the largest colorectal samples datasets, it showed high to very high accuracy. More on the study...
CAIM Ethics Talk "AI, autonomy, and mental health" What role should AI play in our lives and in healthcare? Is it a mere "tool" for healthcare professionals or patients? Could it act as an on-par colleague? Or should AI be allowed to develop freely and autonomously - one day maybe surpassing human ability? Claus Beisbart, Michael Kaess and Mihai Petrovici discuss these exciting points with us in our third Ethics Talk. Learn more and register...
"I want to understand how the brain learns and makes predictions." Katharina Wilmes is fascinated by the question of how humans learn in an uncertain world. To find out more, she creates neuroscientific models, combining mathematics, neural theory, and computational neural network simulations with experimental testing of hypotheses linking sensory experience and learning. Read the interview...
AI@UniBE - Research in Dialogue Get inspired by current research with AI at UniBE and shape the dialogue on responsible AI! Event organized by the Uni Bern initiative "Artificial Intelligence in Education, Organisation and Research". Program and registration...
MLP mixers in lung image segmentation The AI in Health and Nutrition lab at the ARTORG Center conducted performance tests on a computer-aided diagnosis (CAD) system for lung fibrosis, employing multi-layer-perceptron-mixers (MLP-mixers) for segmenting lung and airway anatomy. Additionally, the system identifies idiopathic lung disease patterns in a second step. Currently, MLP-mixers segmentation demonstrates performance comparable to nnU-Net on chest CT images. Learn more…
Pfizer prize for AI-based coma recovery prediction Florence Aellen, Athina Tzovara, and their colleagues from Cognitive Computational Neuroscience, University of Bern and Inselspital, have been awarded one of the coveted Pfizer Research Prizes in Neuroscience and Neurology. They challenged the traditional approach for predicting coma outcome which leaves a percentage of patients without clear prognosis, and used deep learning algorithms applied to coma patients’ EEG responses to sound stimuli to assess the integrity of neural functions in coma and chances of recovery. Learn more about the project...
Multiplexed imaging for next-generation pathology Fifth part of the "Trailblazers" series. Leeat Keren iis an assistant professor in the department of Molecular Cell Biology, Weizmann Institute of Science. She completed her postdoc with Dr. Michael Angelo in Stanford University in 2020. In this online talk she will introduce AI for spatial omics and the interpretation of cellular expression patterns within their native context in the tissue for next-generation pathology. Register here for the online talk...
CAIM Ethics Talk "Transparent AI for acute medicine & diagnostics" Acute medicine lives of fast decisions which are essential for the wellbeing and outcome of patients. How can AI help here and avoid misdiagnosis rather than provoking it? What is the role of clinical metabolomics and how explainable can and must AI technology be to serve in acute care? Learn more about our second talk and register...
CAIM Ethics Talk "Ethics and responsibility in AI development" How would responsible AI look like? How to attribute accountability for AI applications in healthcare? And how will AI for neuroscience be influenced by ethical considerations? Find out at our CAIM Ethics Talk on 7 December 2023 at noon on the ARTORG Center top floor, Murtenstrasse 50! More about the panel and registration...
CAIM featured in NZZ Format The NZZ Format documentary “KI in der Medizin – Sind Algorithmen bald die besseren Ärzte?” visited CAIM as pioneering interdiisciplinary research, teaching and translation platform for AI technologies for healthcare. Here computer scientists, biomedical engineers and doctors are working together to develop new AI technologies.💡 The documentary explains how AI is already used in a clinical practice today, highlights the opportunities and ethical challenges and provides an outlook into the future. Enjoy insights into our work here...
Trailblazers "AI in Medicine" series New series in fall & spring, organized by the Institute for Tissue Medicine and Pathology as part of the initiative "Diversity for AI in Medicine" (DAIM). The seven talks by speakers from Belgium, Canada, Israel, Singapore, Switzerland and UK can be booked separately or as the whole series. Fall 2023 features "Women in Digital Pathology", spring 2024 "Women in Spatial Omics". All further information and registration...
Saliency feature learning for MRI Introducing saliency feature learning to multi-sequence MRI could improve the segmentation quality of deep learning systems. This is the finding of a recent study of the Medical Image Analysis lab in collaboration with the Support Center for Advanced Neuroimaging (SCAN) and the Department of Radiation Oncology of the Inselspital, Bern University Hospital. More on the study…
AI for medicine; what is the medicine for AI? Second part of the "Trailblazers" series. Maria Gabrani is Research Staff Member at IBM Zurich, Director IBM-Research Europe and responsible for 5 EMEA labs and 2 global IBM Research Strategies, namely, Accelerated Discovery, and Security. In her talk she will ask how AI impacts the delivery and discovery in healthcare & life sciences, what its limitations are and what its role is to better serve the needs of patients and caregivers. Register here for the online talk...
Personalized lab data readings across hospitals Researchers from the BioRef consortium of the University Hospitals of Bern (Inselspital) and Lausanne (CHUV), the University Children's Hospital Zurich and Swiss Paraplegic Research have established a multicentric IT framework together with Tune Insight and SPHN to obtain patient group-specific reference intervals from routine laboratory data. This infrastructure, BioRef TI4Health, is the first of its kind featuring a federated network where patient data remains at each participating hospital, yet allowing a joint evaluation in real-time, reproducible, and secured by homomorphic encryption. More on the project...
Monitoring nutrition with a single image In a refinement of its earlier system for automated nutrient assessment via smartphone, the AI in Health and Nutrition laboratory in collaboration with the School of Health Professions, Bern University of Applied Sciences, has developed an easy-to-use single-image-input pipeline to output nutrient composition within the goFOODTM application. In a retrospective real-world study, the new system reduced user burden and showed promising results. Know more…
Multiclass AI model rates Covid severity To improve the initial patient assessment and disease characterization from lung CT scans, the Medical Image Analysis lab of the ARTORG Center, Uni Bern, and the Department of Radiology of the Inselspital, Bern University Hospital have developed a 2-stage multiclass lung lesion model trained to classify disease severity based on the WHO Clinical Progression Scale. The models performance exceeded that of a single-class model as well as the radiologists’ assessment. More about the study…
Gaming for Science Mike Falkner spends up to 10 hours a day behind a screen. What can't possibly be healthy, actually improves memory and fine motor skills. But first things first... Learn more...
CAIM Fellowships Call 2023 The CAIM Research Fund is opening a call for CAIM Fellowships. The project funding to support early-career researchers with promising career prospects in translational AI in medicine research will support projects for up to 2 years with up to 100’000 CHF. Deadline for applications is 25 September 2023. More information and application documents...
AI-assisted brain tumor measurements Automated tumor segmentation for glioblastoma is promising, yet little is known about longitudinal accuracy of automated measurements to assess treatment response. The Medical Image Analysis lab at the ARTORG Center for Biomedical Engineering Research at the University of Bern compared assessment consistency by two AI-based tumor segmentation tools against expert ratings over time, identifying a need for further improvement. Learn more...
AI in Digital Pathology: Opportunities and Challenges First part of the "Trailblazers" series. Anne Martel, Professor in Medical Biophysics at the University of Toronto, will outline some of the unique challenges of working with these extremely large whole slide images in computational pathology and discuss some of the approaches that her lab has developed to overcome the problems of sparse annotations and weak, noisy labels. Register here for the online talk...
Machine learning limited in IBD subclassification The Inselspital and ARTORG Center, Uni Bern, have tested a supervised and unsupervised machine learning approach using antibodies in lab samples to distinguish Crohn’s disease (CD) and ulcerative colitis (UC) as subtypes of inflammatory bowel disease (IBD). The approach showed good sensitivity to correctly assign CD and UC, but could not aid in finer distinction of subtypes such as IBD-unclassified due to the lack of specificity of the analyzed markers. Learn more...
Deep Learning Dose Prediction for Radiotherapy Radiotherapy treatment requires accurate contouring of tumor volumes and organs at risk. In a collaboration between the ARTORG Center for Biomedical Engineering Research, Uni Bern, and the Department of Radiation Therapy, Inselspital, a deep learning architecture for automatic segmentation and 3D dose prediction for glioblastoma patients has shown good sensitivity and robustness. Learn more...
"Deep Learning can contribute to improving surgical outcomes." Negin Ghamsarian is a postdoctoral researcher at the ARTORG Center for Biomedical Engineering Research. She believes that self-supervised and semi-supervised deep learning can overcome current constraints in the feasibility of AI techniques for surgical video analysis to better predict postoperative complications and offer more precise surgical interventions. Read the interview...
"We want to make eye testing and monitoring more widely available." As part of the Venture Leaders program 2023, PeriVision CEO Patrick Kessel allows insights into the young company´s technology, market potential as well as his personal approach towards entrepreneurship. Read the VentureLab interview...
PeriVision closes first Seed Round PeriVision, a spin-off from the University of Bern, is utilizing the power of AI, virtual reality, and cloud computing to redefine eye testing and care. This spring, the company successfully closed the first tranche of a seed round, raising CHF 0.5-1.0M to further accelerate research and development, and prepare for the launch of their inaugural product in Q1 2024. Learn more...
AI assisted epilepsy diagnosis In a clinical validation, the Support Center for Advanced Neuroimaging (SCAN) of the Inselspital has compared the accuracy of combining an AI assistant developed at SCAN with a human expert reading of MRI imaging with the performance of a neuroradiologist reading to diagnose hippocampal sclerosis. The team found that the "human + machine" model improved the mean accuracy of the raters from 77.5% (MRI only) to 86.3% (MRI + AI assistant). Learn more...
DAIM Lunch Talk Lisa Falco CAIM is thrilled to host Lisa Falco, Lead Consultant for AI & Data at Zühlke, for a DAIM Lunch Talk on "Responsible (Generative) AI". Lisa has 15+ years of industry experience in machine learning for healthcare and has led the development of two AI based medical device products, that were brought to market under regulatory requirements. She is the author of “Go Figure! The astonishing science of the female body”, an advocate for women’s health and active in the FemTech community. More info & registration...
Statistical learning in lab data Laboratory information systems are struggling to manage and analyze the amount of complex and entangled data generated in laboratory medicine today. Statistical learning, a generalized framework from machine learning and AI, is predestined for processing this “Big Data” and holds the potential to revolutionize the field. To ensure quality, transparency, and public acceptance, human experts should carefully validate AI-based systems, including patient-privacy protection. Learn more...
Tumor budding T-cell graphs Minimally invasive types of colorectal cancer (pT1) may only necessitate removing the polyps instead of requiring full colon resection to stop cancer progression. To assist pathologists, the Institute for Tissue Medicine and Pathology of the University of Bern has evaluated a Graph Neural Network based approach for risk stratification based on the interaction of tumor buds and T-cells as well as other histopathological risk factors with about 20% higher specifitiy than current guidelines. Learn more...
First anniversary DAIM In May 2023, our initiative "Diversity for AI in Medicine" became one year old! We are extremely happy to witness the commencement of such a vibrant community of AI researchers for medicine here in Bern. Special thanks to all our supporters and the many speakers that have inspired us this past year! Discover DAIM’s first year here...
AI in interstitial lung disease Over 200 diagnoses exist for the umbrella term "interstitial lung diseases" (ILDs). These are currently determined by an expert board based on CT imaging, patient data, pulmonary function tests, and histology. AI methods could help aid in accurate ILD diagnosis and predict prognosis and progression in a holistic system, as a review paper by the ARTORG Center and Inselspital comparing current approaches shows. More on the review...
Un centre consacré à la recherche sur l'IA en médecine In a special issue dedicated to AI in medicine, the Société Vaudoise de Médecine presents our research center. The short introduction to CAIM include the essential role of ethics and our strive for multi-perspective work environments through the inclusion of diverse teams and medical as well as engineering professions. It also gives an overview of our Master's program AI in Medicine and the five research projects CAIM is currently funding. Read the article (in French)...
"AI supports conscious nutrition" Artificial intelligence for the management of diabetes or obesity is gaining importance in everyday life and in medicine. Prof. Dr. Stavroula Mougiakakou, Professor of Biomedical Engineering at the ARTORG Center of the University of Bern, gives an insight into the development of food apps and shows how much humanity is still in the algorithms of AI. Read the interview in gastroMAG (German) ...
Can AI chatbots transform patient care? Generative AI such as ChatGPT are currently on everyone's lips. Also in healthcare, such systems have an enormous potential to revolutionize existing processes. An overview by the Computational Medicine Group, University Institute of Clinical Chemistry, Inselspital Bern, illustrates the opportunities and risks of this fascinating technology. Read the article in Netzwoche (in German)...
Tele Emergency Medicine Congress Forth Swiss congress under the focus "Metaverse in Acute Medicine: Entering digital worlds". New this year are Medical Extended Reality Workshops at the Verkehrshaus Luzern from 9.00-12.00h. The afternoon sessions from 13.30h are online. More information and registration...
How AI is changing medicine In an article on current implementation of artificial intelligence in healthcare today, Barbara Reye explores the disruptive potential of machine learning in medicine. Examples range from intensive care monitoring and false alarm reductions over early signs for cerebral hemorrhage, medical image analysis and diagnostic support, skin cancer, newborn health and a nutrition app developed at the University of Bern to help users stay on track with a health-promoting Mediterranean diet. Read the article...
"I believe that AI is the key to a new era of healthcare." Song Xue is a biomedical postdoctoral researcher specialized in deep learning at the AI for Translational Theranostics (AITT) group of the Department of Nuclear Medicine, Inselspital. Applying artificial intelligence to nuclear medicine, Song aims to reduce radiation in diagnostic PET imaging and to personalize dose prediction for radionuclide therapy. Read the interview...
Bern Interpretable AI Symposium One-day meeting to bring researchers together in the medical image interpretable AI community. Our hope and objective is to attempt to “open the black box”, share insights into challenges and breakthroughs in this field and to foster better interaction with each other. Program & registration...
"With deep learning you can explore rich patterns of neural activity." Florence Aellen is the deep learning specialist at the Cognitive Computational Neuroscience Lab of the Institute of Computer Science, University of Bern. For her PhD she works with a very interdisciplinary research team to unravel interrelations between electrical brain activity and states of consciousness. Herself originally coming from a background in mathematics and theoretical physics, she now puts her computational expertise in the service of clinical applications, learning a lot in the process. Read the interview...
AI Medtech Founders Ready to start your own AI business? Networking and insights sharing event on Women´s Day 2023 featuring four Medtech founders and hands-on tips on how to get support and funding and how to build your network. More on the event...
AI to detect lymph node metastases in colorectal cancer Screening of lymph node metastases in colorectal cancer (CRC) could be solved by an AI-assisted diagnostic solution. A group of pathologists and data scientists led by the Institute for Tissue Medicine and Pathology, University of Bern, has proposed a deep learning–based workflow for the evaluation of CRC lymph node metastases. The approach showed very good performance, providing the basis for a computer-assisted diagnostic tool for easy and efficient lymph node screening in CRC patients. Learn more...
Measuring blood glucose with the smartwatch Researchers at the Inselspital, Bern University Hospital, and the University of Bern developed a method that uses machine learning and smartwatch data to detect hypoglycemia. The approach was tested in a pilot study with 31 diabetes patients and could in the future be used to non invasively measure blood glucose levels as a valuable complement to existing methods for blood glucose monitoring. Read the study...
Bridging deep learning and clinical radiotherapy The ARTORG Medical Image Analysis group (MIA) and the Department of Radio-Oncology of the Inselspital have developed an open-source, deep learning framework-independent Python package feasible for processing DICOM RT Structure Sets, in collaboration with Varian a Siemens Healthineers Company. “PyRaDiSe” (Python (Py), radiotherapy (Ra), DICOM-based (Di), auto-segmentation (Se)) goes beyond current 2D reconstruction, potentially supporting acceptance by healthcare professionals. View the package…
Predicting heart attack from routine hospital data? The Department of Clinical Chemistry and the Insel Data Science Center at the Inselspital have conducted a proof-of-concept study to infer the diagnosis of myocardial ischemia from predictive analytes extracted from routine hospital lab data. Authors could demonstrate that an assembly of algorithms was able to create unbiased guidelines for laboratory diagnostics by using computational evidence in today’s era of healthcare digitalization. This is an important first step towards evidence-based laboratory diagnostics and, therefore, improved patient care. Learn more...
AI-assisted interpretation in cardiotocography Artificial intelligence is gaining interest in the field of cardiotocography interpretation in obstetrics, since it promises to remove existing biases and improve the well-known issues of inter- and intra-observer variability. A review publication by the Department of Obstetrics and Feto-maternal Medicine, Bern University Hospital, with CSEM hints at potentials of AI to help detect complications in the unborn child earlier. Read the review paper...