Chantal Pellegrini

Chantal Pellegrini, Ege Özsoy, Benjamin Busam et al.

Conversational AI tools that can generate and discuss clinically correct radiology reports for a given medical image have the potential to transform radiology. Such a human-in-the-loop radiology assistant could facilitate a collaborative diagnostic process, thus saving time and improving the quality of reports. Towards this goal, we introduce RaDialog, the first thoroughly evaluated and publicly available large vision-language model for radiology report generation and interactive dialog. RaDialog effectively integrates visual image features and structured pathology findings with a large language model (LLM) while simultaneously adapting it to a specialized domain using parameter-efficient fine-tuning. To keep the conversational abilities of the underlying LLM, we propose a comprehensive, semi-automatically labeled, image-grounded instruct dataset for chest X-ray radiology tasks. By training with this dataset, our method achieves state-of-the-art clinical correctness in report generation and shows impressive abilities in interactive tasks such as correcting reports and answering questions, serving as a foundational step toward clinical dialog systems. Our code is available on github: https://github.com/ChantalMP/RaDialog.

Chantal Pellegrini, Matthias Keicher, Ege Özsoy et al.

Radiology reporting is a crucial part of the communication between radiologists and other medical professionals, but it can be time-consuming and error-prone. One approach to alleviate this is structured reporting, which saves time and enables a more accurate evaluation than free-text reports. However, there is limited research on automating structured reporting, and no public benchmark is available for evaluating and comparing different methods. To close this gap, we introduce Rad-ReStruct, a new benchmark dataset that provides fine-grained, hierarchically ordered annotations in the form of structured reports for X-Ray images. We model the structured reporting task as hierarchical visual question answering (VQA) and propose hi-VQA, a novel method that considers prior context in the form of previously asked questions and answers for populating a structured radiology report. Our experiments show that hi-VQA achieves competitive performance to the state-of-the-art on the medical VQA benchmark VQARad while performing best among methods without domain-specific vision-language pretraining and provides a strong baseline on Rad-ReStruct. Our work represents a significant step towards the automated population of structured radiology reports and provides a valuable first benchmark for future research in this area. Our dataset and code is available at https://github.com/ChantalMP/Rad-ReStruct.

Chantal Pellegrini, Matthias Keicher, Ege Özsoy et al.

Automated diagnosis prediction from medical images is a valuable resource to support clinical decision-making. However, such systems usually need to be trained on large amounts of annotated data, which often is scarce in the medical domain. Zero-shot methods address this challenge by allowing a flexible adaption to new settings with different clinical findings without relying on labeled data. Further, to integrate automated diagnosis in the clinical workflow, methods should be transparent and explainable, increasing medical professionals' trust and facilitating correctness verification. In this work, we introduce Xplainer, a novel framework for explainable zero-shot diagnosis in the clinical setting. Xplainer adapts the classification-by-description approach of contrastive vision-language models to the multi-label medical diagnosis task. Specifically, instead of directly predicting a diagnosis, we prompt the model to classify the existence of descriptive observations, which a radiologist would look for on an X-Ray scan, and use the descriptor probabilities to estimate the likelihood of a diagnosis. Our model is explainable by design, as the final diagnosis prediction is directly based on the prediction of the underlying descriptors. We evaluate Xplainer on two chest X-ray datasets, CheXpert and ChestX-ray14, and demonstrate its effectiveness in improving the performance and explainability of zero-shot diagnosis. Our results suggest that Xplainer provides a more detailed understanding of the decision-making process and can be a valuable tool for clinical diagnosis.

Chantal Pellegrini, Nassir Navab, Anees Kazi

Pre-training has shown success in different areas of machine learning, such as Computer Vision, Natural Language Processing (NLP), and medical imaging. However, it has not been fully explored for clinical data analysis. An immense amount of clinical records are recorded, but still, data and labels can be scarce for data collected in small hospitals or dealing with rare diseases. In such scenarios, pre-training on a larger set of unlabelled clinical data could improve performance. In this paper, we propose novel unsupervised pre-training techniques designed for heterogeneous, multi-modal clinical data for patient outcome prediction inspired by masked language modeling (MLM), by leveraging graph deep learning over population graphs. To this end, we further propose a graph-transformer-based network, designed to handle heterogeneous clinical data. By combining masking-based pre-training with a transformer-based network, we translate the success of masking-based pre-training in other domains to heterogeneous clinical data. We show the benefit of our pre-training method in a self-supervised and a transfer learning setting, utilizing three medical datasets TADPOLE, MIMIC-III, and a Sepsis Prediction Dataset. We find that our proposed pre-training methods help in modeling the data at a patient and population level and improve performance in different fine-tuning tasks on all datasets.

Ege Özsoy, Tobias Czempiel, Felix Holm et al.

Modern surgeries are performed in complex and dynamic settings, including ever-changing interactions between medical staff, patients, and equipment. The holistic modeling of the operating room (OR) is, therefore, a challenging but essential task, with the potential to optimize the performance of surgical teams and aid in developing new surgical technologies to improve patient outcomes. The holistic representation of surgical scenes as semantic scene graphs (SGG), where entities are represented as nodes and relations between them as edges, is a promising direction for fine-grained semantic OR understanding. We propose, for the first time, the use of temporal information for more accurate and consistent holistic OR modeling. Specifically, we introduce memory scene graphs, where the scene graphs of previous time steps act as the temporal representation guiding the current prediction. We design an end-to-end architecture that intelligently fuses the temporal information of our lightweight memory scene graphs with the visual information from point clouds and images. We evaluate our method on the 4D-OR dataset and demonstrate that integrating temporality leads to more accurate and consistent results achieving an +5% increase and a new SOTA of 0.88 in macro F1. This work opens the path for representing the entire surgery history with memory scene graphs and improves the holistic understanding in the OR. Introducing scene graphs as memory representations can offer a valuable tool for many temporal understanding tasks.

Ege Özsoy, Felix Holm, Mahdi Saleh et al.

Scene Graph Generation (SGG) is a visual understanding task, aiming to describe a scene as a graph of entities and their relationships with each other. Existing works rely on location labels in form of bounding boxes or segmentation masks, increasing annotation costs and limiting dataset expansion. Recognizing that many applications do not require location data, we break this dependency and introduce location-free scene graph generation (LF-SGG). This new task aims at predicting instances of entities, as well as their relationships, without the explicit calculation of their spatial localization. To objectively evaluate the task, the predicted and ground truth scene graphs need to be compared. We solve this NP-hard problem through an efficient branching algorithm. Additionally, we design the first LF-SGG method, Pix2SG, using autoregressive sequence modeling. We demonstrate the effectiveness of our method on three scene graph generation datasets as well as two downstream tasks, image retrieval and visual question answering, and show that our approach is competitive to existing methods while not relying on location cues.

Chantal Pellegrini, Anees Kazi, Nassir Navab

Pre-training has shown success in different areas of machine learning, such as Computer Vision (CV), Natural Language Processing (NLP) and medical imaging. However, it has not been fully explored for clinical data analysis. Even though an immense amount of Electronic Health Record (EHR) data is recorded, data and labels can be scarce if the data is collected in small hospitals or deals with rare diseases. In such scenarios, pre-training on a larger set of EHR data could improve the model performance. In this paper, we apply unsupervised pre-training to heterogeneous, multi-modal EHR data for patient outcome prediction. To model this data, we leverage graph deep learning over population graphs. We first design a network architecture based on graph transformer designed to handle various input feature types occurring in EHR data, like continuous, discrete, and time-series features, allowing better multi-modal data fusion. Further, we design pre-training methods based on masked imputation to pre-train our network before fine-tuning on different end tasks. Pre-training is done in a fully unsupervised fashion, which lays the groundwork for pre-training on large public datasets with different tasks and similar modalities in the future. We test our method on two medical datasets of patient records, TADPOLE and MIMIC-III, including imaging and non-imaging features and different prediction tasks. We find that our proposed graph based pre-training method helps in modeling the data at a population level and further improves performance on the fine tuning tasks in terms of AUC on average by 4.15% for MIMIC and 7.64% for TADPOLE.

David Bani-Harouni, Chantal Pellegrini, Julian Lüers et al.

Safe deployment of Large Vision-Language Models (LVLMs) in radiology report generation requires not only accurate predictions but also clinically interpretable indicators of when outputs should be thoroughly reviewed, enabling selective radiologist verification and reducing the risk of hallucinated findings influencing clinical decisions. One intuitive approach to this is verbalized confidence, where the model explicitly states its certainty. However, current state-of-the-art language models are often overconfident, and research on calibration in multimodal settings such as radiology report generation is limited. To address this gap, we introduce ConRad (Confidence Calibration for Radiology Reports), a reinforcement learning framework for fine-tuning medical LVLMs to produce calibrated verbalized confidence estimates alongside radiology reports. We study two settings: a single report-level confidence score and a sentence-level variant assigning a confidence to each claim. Both are trained using the GRPO algorithm with reward functions based on the logarithmic scoring rule, which incentivizes truthful self-assessment by penalizing miscalibration and guarantees optimal calibration under reward maximization. Experimentally, ConRad substantially improves calibration and outperforms competing methods. In a clinical evaluation we show that ConRad's report level scores are well aligned with clinicians' judgment. By highlighting full reports or low-confidence statements for targeted review, ConRad can support safer clinical integration of AI-assistance for report generation.

Ege Özsoy, Chantal Pellegrini, Tobias Czempiel et al.

Operating rooms (ORs) are complex, high-stakes environments requiring precise understanding of interactions among medical staff, tools, and equipment for enhancing surgical assistance, situational awareness, and patient safety. Current datasets fall short in scale, realism and do not capture the multimodal nature of OR scenes, limiting progress in OR modeling. To this end, we introduce MM-OR, a realistic and large-scale multimodal spatiotemporal OR dataset, and the first dataset to enable multimodal scene graph generation. MM-OR captures comprehensive OR scenes containing RGB-D data, detail views, audio, speech transcripts, robotic logs, and tracking data and is annotated with panoptic segmentations, semantic scene graphs, and downstream task labels. Further, we propose MM2SG, the first multimodal large vision-language model for scene graph generation, and through extensive experiments, demonstrate its ability to effectively leverage multimodal inputs. Together, MM-OR and MM2SG establish a new benchmark for holistic OR understanding, and open the path towards multimodal scene analysis in complex, high-stakes environments. Our code, and data is available at https://github.com/egeozsoy/MM-OR.

Ege Özsoy, Chantal Pellegrini, Matthias Keicher et al.

Every day, countless surgeries are performed worldwide, each within the distinct settings of operating rooms (ORs) that vary not only in their setups but also in the personnel, tools, and equipment used. This inherent diversity poses a substantial challenge for achieving a holistic understanding of the OR, as it requires models to generalize beyond their initial training datasets. To reduce this gap, we introduce ORacle, an advanced vision-language model designed for holistic OR domain modeling, which incorporates multi-view and temporal capabilities and can leverage external knowledge during inference, enabling it to adapt to previously unseen surgical scenarios. This capability is further enhanced by our novel data augmentation framework, which significantly diversifies the training dataset, ensuring ORacle's proficiency in applying the provided knowledge effectively. In rigorous testing, in scene graph generation, and downstream tasks on the 4D-OR dataset, ORacle not only demonstrates state-of-the-art performance but does so requiring less data than existing models. Furthermore, its adaptability is displayed through its ability to interpret unseen views, actions, and appearances of tools and equipment. This demonstrates ORacle's potential to significantly enhance the scalability and affordability of OR domain modeling and opens a pathway for future advancements in surgical data science. We will release our code and data upon acceptance.

Chantal Pellegrini, Adrian Delchev, Ege Özsoy et al.

Structured radiology reporting promises faster, more consistent communication than free text, but automation remains difficult as models must make many fine-grained, discrete decisions about rare findings and attributes from limited structured supervision. In contrast, free-text reports are produced at scale in routine care and implicitly encode fine-grained, image-linked information through detailed descriptions. To leverage this unstructured knowledge, we propose ProtoSR, an approach for injecting free-text information into structured report population. First, we introduce an automatic extraction pipeline that uses an instruction-tuned LLM to mine 80k+ MIMIC-CXR studies and build a multimodal knowledge base aligned with a structured reporting template, representing each answer option with a visual prototype. Using this knowledge base, ProtoSR is trained to retrieve prototypes relevant for the current image-question pair and augment the model predictions through a prototype-conditioned residual, providing a data-driven second opinion that selectively corrects predictions. On the Rad-ReStruct benchmark, ProtoSR achieves state-of-the-art results, with the largest improvements on detailed attribute questions, demonstrating the value of integrating free-text derived signal for fine-grained image understanding.

Paul Stangel, David Bani-Harouni, Chantal Pellegrini et al.

A safe and trustworthy use of Large Language Models (LLMs) requires an accurate expression of confidence in their answers. We propose a novel Reinforcement Learning approach that allows to directly fine-tune LLMs to express calibrated confidence estimates alongside their answers to factual questions. Our method optimizes a reward based on the logarithmic scoring rule, explicitly penalizing both over- and under-confidence. This encourages the model to align its confidence estimates with the actual predictive accuracy. The optimal policy under our reward design would result in perfectly calibrated confidence expressions. Unlike prior approaches that decouple confidence estimation from response generation, our method integrates confidence calibration seamlessly into the generative process of the LLM. Empirically, we demonstrate that models trained with our approach exhibit substantially improved calibration and generalize to unseen tasks without further fine-tuning, suggesting the emergence of general confidence awareness. We provide our training and evaluation code in the supplementary and will make it publicly available upon acceptance.

David Bani-Harouni, Chantal Pellegrini, Ege Özsoy et al.

Clinical decision-making is a dynamic, interactive, and cyclic process where doctors have to repeatedly decide on which clinical action to perform and consider newly uncovered information for diagnosis and treatment. Large Language Models (LLMs) have the potential to support clinicians in this process, however, most applications of LLMs in clinical decision support suffer from one of two limitations: Either they assume the unrealistic scenario of immediate availability of all patient information and do not model the interactive and iterative investigation process, or they restrict themselves to the limited "out-of-the-box" capabilities of large pre-trained models without performing task-specific training. In contrast to this, we propose to model clinical decision-making for diagnosis with a hypothesis-driven uncertainty-aware language agent, LA-CDM, that converges towards a diagnosis via repeatedly requesting and interpreting relevant tests. Using a hybrid training paradigm combining supervised and reinforcement learning, we train LA-CDM with three objectives targeting critical aspects of clinical decision-making: accurate hypothesis generation, hypothesis uncertainty estimation, and efficient decision-making. We evaluate our methodology on MIMIC-CDM, a real-world dataset covering four abdominal diseases containing various clinical tests and show the benefit of explicitly training clinical decision-making for increasing diagnostic performance and efficiency.

Ege Özsoy, Arda Mamur, Felix Tristram et al.

Operating rooms (ORs) demand precise coordination among surgeons, nurses, and equipment in a fast-paced, occlusion-heavy environment, necessitating advanced perception models to enhance safety and efficiency. Existing datasets either provide partial egocentric views or sparse exocentric multi-view context, but do not explore the comprehensive combination of both. We introduce EgoExOR, the first OR dataset and accompanying benchmark to fuse first-person and third-person perspectives. Spanning 94 minutes (84,553 frames at 15 FPS) of two emulated spine procedures, Ultrasound-Guided Needle Insertion and Minimally Invasive Spine Surgery, EgoExOR integrates egocentric data (RGB, gaze, hand tracking, audio) from wearable glasses, exocentric RGB and depth from RGB-D cameras, and ultrasound imagery. Its detailed scene graph annotations, covering 36 entities and 22 relations (568,235 triplets), enable robust modeling of clinical interactions, supporting tasks like action recognition and human-centric perception. We evaluate the surgical scene graph generation performance of two adapted state-of-the-art models and offer a new baseline that explicitly leverages EgoExOR's multimodal and multi-perspective signals. This new dataset and benchmark set a new foundation for OR perception, offering a rich, multimodal resource for next-generation clinical perception.

Chantal Pellegrini, Ege Özsoy, David Bani-Harouni et al.

Healthcare systems face significant challenges in managing and interpreting vast, heterogeneous patient data for personalized care. Existing approaches often focus on narrow use cases with a limited feature space, overlooking the complex, longitudinal interactions needed for a holistic understanding of patient health. In this work, we propose a novel approach to patient pathway modeling by transforming diverse electronic health record (EHR) data into a structured representation and designing a holistic pathway prediction model, EHR2Path, optimized to predict future health trajectories. Further, we introduce a novel summary mechanism that embeds long-term temporal context into topic-specific summary tokens, improving performance over text-only models, while being much more token-efficient. EHR2Path demonstrates strong performance in both next time-step prediction and longitudinal simulation, outperforming competitive baselines. It enables detailed simulations of patient trajectories, inherently targeting diverse evaluation tasks, such as forecasting vital signs, lab test results, or length-of-stay, opening a path towards predictive and personalized healthcare.

Ege Özsoy, Chantal Pellegrini, David Bani-Harouni et al.

Surgical procedures unfold in complex environments demanding coordination between surgical teams, tools, imaging and increasingly, intelligent robotic systems. Ensuring safety and efficiency in ORs of the future requires intelligent systems, like surgical robots, smart instruments and digital copilots, capable of understanding complex activities and hazards of surgeries. Yet, existing computational approaches, lack the breadth, and generalization needed for comprehensive OR understanding. We introduce ORQA, a multimodal foundation model unifying visual, auditory, and structured data for holistic surgical understanding. ORQA's question-answering framework empowers diverse tasks, serving as an intelligence core for a broad spectrum of surgical technologies. We benchmark ORQA against generalist vision-language models, including ChatGPT and Gemini, and show that while they struggle to perceive surgical scenes, ORQA delivers substantially stronger, consistent performance. Recognizing the extensive range of deployment settings across clinical practice, we design, and release a family of smaller ORQA models tailored to different computational requirements. This work establishes a foundation for the next wave of intelligent surgical solutions, enabling surgical teams and medical technology providers to create smarter and safer operating rooms.

Lukas Buess, Jan Geier, David Bani-Harouni et al.

Spoken communication plays a central role in clinical workflows. In radiology, for example, most reports are created through dictation. Yet, nearly all medical AI systems rely exclusively on written text. In this work, we address this gap by exploring the feasibility of learning visual-language representations directly from spoken radiology reports. Specifically, we synthesize a large-scale dataset (Speech-RATE) of spoken radiology reports and train SpeechCT-CLIP, a contrastive model that aligns speech and 3D CT volumes in a shared representation space. While naive speech-based models underperform compared to text-trained counterparts, we show that knowledge distillation from a pretrained text-image CLIP model effectively transfers semantic alignment capabilities from text to speech, substantially narrowing this gap. Experiments demonstrate improved zero-shot classification F1 from 0.623 to 0.705, recovering 88% of the performance difference, and strong retrieval results without requiring text at inference. These findings highlight speech as a practical alternative to text in multimodal pretraining and open the door to voice-driven diagnostic support tools in clinical practice.