Conference on Artificial Intelligence in Medical Imaging

Agents can Segment Too: Assessing Out-of-the-Box Performance of Multimodal Tool-Equipped LLMs for Tumor Segmentation 

• Sameed Khan, Cleveland Clinic Lerner College of Medicine 
• Daniyal Nadeem; Mahmoud Omar, MD; Charit Tippareddy, MD 

 

Benchmarking Foundation Vision-Language Models for Gastrointestinal Contrast Extravasation Detection on CT 

• Michael Fei, Creighton University School of Medicine 
• Christopher Bridge, DPhil; Dania Daye, MD, PhD 

 

ConTEXTual Net CT: A 3D Visual Grounding Model for Abdominal CT 

• Samuel D. Church, MS, University of Wisconsin – Madison 
• Joshua D. Warner, MD, PhD; Danyal Maqbool, MS; Junjie Hu, PhD; Meghan G. Lubner, MD; Tyler J. Bradshaw, PhD 

 

From Narrative Radiology Reports to Vision-Language Model Training Labels: Transparent Physician Validation Using Evidence-Grounded Visualization 

• Mohammed Ayman Habib, PhD Student, University of Utah 
• Isaac Anthony; Osei Brempong, PhD Student; Derrick Wong; Morteza Fayazi, PhD; Seyyedkazem Hashemizadehkolowri, PhD; John Roberts, PhD; Tyler Richards, MD; Maryam Soltanolkotabi, MD 

 

Inducing False Positive Diagnoses in a Radiology Vision Language Model: Comparative Vulnerability of User Prompts, System Instructions, and Image Embedded Labels 

• Ebubechukwu D. Enwerem, University of Pennsylvania 
• Kristian Quevada, MD; Satvik Tripathi; Tessa Cook, MD, PhD, CIIP, FSIIM 

 

SPARC-Rad: A Multimodal Benchmark Dataset and Evaluation Pipeline for Spatial and Anatomical Reasoning in Radiology Vision-Language Models 

• Satvik Tripathi, University of Pennsylvania 
• Dania Daye, MD, PhD; Tessa Cook, MD, PhD, CIIP, FSIIM