MI at RSNA 2025

The Radiological Society of North America’s (RSNA) Annual Meeting, RSNA 2025: Imaging the Individual, starts this week from November 30 – December 4, and numerous MI faculty and trainees will be in attendance presenting exhibits, posters, delivering talks and presentations and representing the Department well.

See below RSNA presentations from U of T’s Department of Medical Imaging:

Posters/Exhibits/Abstracts

Submission by: Dr. Caroline Rutten, Assistant Professor
Poster: LOOK DOWN! WHEN PEDIATRIC BODY IMAGING HOLDS THE KEY TO NEUROLOGICAL DIAGNOSIS
Team from The Hospital for Sick Children
Authors: Caroline Rutten, Asst. Professor; Kevin K.F. Fung; Govind Chavhan, Professor; Elka Miller, Professor; Helen Branson, Assoc. Professor; Birgit Ertl-Wagner, Professor & Radiologist-in-Chief, SickKids; Vivek Pai, Asst. Professor 

Poster: SCALP SCRATCHES AND SKULL SURPRISES: WHAT’S HIDING BENEATH?
Team from The Hospital for Sick Children
Authors: Caroline Rutten, Asst. Professor; Daniel Schlam; Soyini Jackson; Anthea Lafreniere; Samantha Gerrie, Asst. Professor; Elka Miller, Professor; Vivek Pai, Asst. Professor

Submission by: Dr. Oscar Navarro, Professor & Pediatric Radiology Program Director
Educational Exhibit: Imaging of abdominal and pelvic tumors in the neonate and young infant
Authors: Diana Veiga Canuto, PhD MD; Roberto Llorens Salvador; Oscar M. Navarro, MD, Professor & Program Director, Pediatrics

Summary: This exhibit reviews the diagnostic approach to abdominal and pelvic tumors in the neonate and young infant, which present unique diagnostic challenges, may show peculiar behavior and prognosis, and can be associated with other congenital anomalies.

Submission by: Dr. Michael Patlas, MI Chair & Professor
Abstract: ACUTE ABDOMINAL PAIN IN ONCOLOGY PATIENTS: DIAGNOSTIC CHALLENGES AND IMAGING INSIGHTS
Authors: Jia Cheng Yao, MD; Iain D.C. Kirkpatrick, MD FRCPC; Vincent M. Mellnick, MD; Philippe A. Soyer, MD PhD; Kate Hanneman, MD MPH, Assoc. Professor & VC, Research; Michael N. Patlas, MD FRCPC, Professor & Chair

Abstract: OPTIMIZING CLINICAL DECISION-MAKING IN PACREATIC CANCER: THE ROLE OF GPT-40 AND DEEPSEEK V3 LARGE LANGUAGE MODELS
Authors: Ankush Jajodia, MD MBBS, Asst. Professor & VC, EDI; Kartik Gupta, MD; Mila Ferri Latinovich; Michael N. Patlas, MD FRCPC, Professor & Chair; Khaled Y. Elbanna, FRCR

Abstract: SEX-SPECIFIC ASSOCATIATIONS OF SHORT-TERM ENVIRONMENTAL EXPOSURES WITH UTILIZATION OF MEDICAL IMAGING IN THE EMERGENCY DEPARTMENT
Authors: Shane O'Driscoll, MBBCh FFR(RCSI); Scott Delaney; Rachel Nethery; Chloe DesRoche, MD MSc; Julien Aguet, MD, Asst. Professor; Birgit B. Ertl-Wagner, MD PhD, Professor & Radiologist-in-Chief, SickKids; Anish Kirpalani, MD FRCPC, Asst. Professor & Radiologist-in-Chief, St. Michael's Hospital; Heidi C. Schmidt, MD, Professor & Radiologist-in-Chief, JDMI; Hayley Panet, Ania Z. Kielar, MD FRCPC, Asst. Professor; Michael N. Patlas, MD FRCPC, Professor & Chair; Joseph Choi; Kate Hanneman, MD MPH, Assoc. Professor & VC, Research; Felipe Castillo Aravena, MD

Abstract: VALUE ADDITION OF SUBSPECIALTY TRAINED RADIOLOGISTS IN MULTIDISCIPLINARY TUMOR BORADS FOR PACREATIC ADENOCARCINOMA: A META-ANALYSIS & SYSTEMATIC REVIEW
Authors: Ankush Jajodia, MD MBBS, Asst. Professor & VC, EDI; Adam Caulfield, MD; Mila Ferri Latinovich; Kartik Gupta, MD; Khaled Y. Elbanna, FRCR; Michael N. Patlas, MD FRCPC, Professor & Chair 

Submission by: Pascal Tyrrell, PhD, Director, Data Science
Abstract: DEBLURRING ULTRASOUND IMAGES TO IMPROVE CLASSIFICATION OF THICKENED SYNOVIUM
Authors: Youyang Guo, Noushin Jafarpisheh, PhD; Pascal Tyrrell, PhD, Director, Data Science

Summary: This study demonstrates that pre-training a model using a deblurring masked auto-encoder (DeblurMIM) significantly improves the automated detection of thickened synovium on knee musculoskeletal ultrasound (MSK-US), achieving 79.4% accuracy compared to a 71.5% accuracy baseline CNN model. The DeblurMIM approach enhanced diagnostic sensitivity and maintained robustness even under increased image blur, offering a practical strategy to improve MSK-US interpretation.

Abstract: MEDSEGGEN: SEMANTIC SYNTHESIS FOR MEDICAL IMAGING SEGMENTATION DATA GENERATION
Authors: Xin Lei Lin, Soroush Mehraban, Daniel Saragih, Varun Sahni; Pascal Tyrrell, PhD, Director, Data Science

Summary: This study introduces MedSegGen, a two-stage semantic synthesis framework using Latent Diffusion Models (LDMs) to generate diverse, anatomically realistic image-mask pairs for Gastrointestinal (GI) endoscopy image segmentation. The framework utilizes a novel mask interpolation module and contour-aware loss to overcome limited labeled data, successfully augmenting datasets to improve the training and generalization of medical image segmentation models.

Abstract: RADFORMER: A SELF-SUPERVISED FOUNDATION MODEL TO ADDRESS INCOMPLETENESS IN RADIOMICS FEATURES
Authors: Khashayar Namdar, PhD MEng; Saeidehsadat Mirjalili; Pascal Tyrrell, PhD, Director, Data Science; Leo Anthony Celi, PhD

Summary: This research introduces RadFormer, a self-supervised transformer-based radiomics foundation model designed to improve the robustness and reproducibility of radiomics pipelines by generating generalizable feature embeddings. When tested on prostate cancer classification with incomplete feature sets, RadFormer-Augmented features achieved an AUC of 0.729, successfully recovering the performance lost due to missing higher-order features and demonstrating its utility as a substitute for handcrafted radiomic features.

Submission by: Dr. Shobhit Mathur, Asst. Professor
Presenter: Kyobin Hwang, Medical Student, U of T
Supervisor: Dr. Shobhit Mathur, Assistant Professor
Abstract: Evaluation of Second-Generation Reconstruction Algorithms to Improve Gray-White Matter Differentiation on Non-Contrast Head CT Studies
Authors: Ian Anderson; Asutosh Sahu; Akhil Nair; Kyobin Hwang; Brian Nett; Joel Kosowan; Shobhit Mathur, Asst. Professor

Summary: The study compared the performance of second-generation (EB1, EB2, EB3) vs first-generation (EC3) CT reconstruction methods for enhancing gray-white matter differentiation on non-contrast head CT for acute stroke. We found that EB3 reconstruction method performs best and improves gray-white matter differentiation on non-contrast CT head compared to EC3 without significant change in noise, texture, and beam-hardening artifacts.

Submission by: Dr. April Khademi, Associate Professor
Abstract/oral presentation: A pipeline for extraction of novel FLAIR biomarkers for diagnosis of Alzheimer’s disease and vascular dementia
Authors: Matthew C. So MD; Pejman Jabehdar Maralani MD, Assoc. Professor; Sandra E. Black MD; April Khademi PhD, PEng, Asst. Professor

Summary: A fast, highly interpretable neurodegenerative biomarker extraction pipeline is described using routinely acquired FLAIR sequences. This may be used to improve diagnosis in unclear clinical cases (such as differentiation from pseudodementia), to quantify and track the degree of parenchymal degeneration, and to prognosticate the course of disease.

Submission by: Dr. Mary-Louise Greer, Professor & Division Lead, Pediatric Imaging
Educational Exhibit: CANCER PREDISPOSITION SYNDROMES IN CHILDREN - WHO, WHEN AND HOW TO SCREEN?
Authors: Kevin Fung, Caroline Rutten, Asst. Professor; Vivek B. Pai, Asst. Professor; Yin Ting Chiu; Elaine YL Kan; Eman EM Marie; Anita Villani; David Malkin; Andrea S. Doria, Professor; Mary-Louise C. Greer, Professor & Division Lead, Pediatric Imaging

Summary: A multimodality review of surveillance imaging in childhood cancer predisposition syndromes

Submission by: Dr. Felipe Castillo Aravena, Cardiothoracic Imaging Fellow
Educational Exhibit: CAEE-54MODERN HEART TRANSPLANT REJECTION SCREENING AND SURVEILLANCE: INTEGRATED ROLE OF CARDIAC IMAGING
Authors: Felipe Castillo Aravena, MD, Cardiothoracic Imaging Fellow; Yas Moayedi; Juan Dueroposada; Kate Hanneman, MD MPH, Assoc. Professor & VC, Research

Educational Exhibit: CHEE-89LOOKING UP FOR A DIAGNOSIS: IMAGING AND DIFFERENTIAL DIAGNOSIS OF UPPER-ZONE PREDOMINANT LUNG DISEASES
Authors: Felipe Castillo Aravena, MD, Cardiothoracic Imaging Fellow; Kate Hanneman, MD MPH, Assoc. Professor & VC, Research; Felipe S. Torres, MD PhD, Asst. Professor; Micheal McInnis, MD, Asst. Professor; Felipe A. Sanchez Tijmes, MD, Asst. Professor; Daniel Vargas, MD; Elsie Nguyen, MD FRCPC, Professor & Division Lead, Cardiothoracic Imaging

Abstract Oral Presentation: M6-SSNPM01-7SEX-SPECIFIC ASSOCIATIONS OF SHORT-TERM ENVIRONMENTAL EXPOSURES WITH UTILIZATION OF MEDICAL IMAGING IN THE EMERGENCY DEPARTMENT
Authors: Shane O'Driscoll, MBBCh FFR(RCSI); Scott Delaney; Rachel Nethery; Chloe DesRoche, MD MSc; Julien Aguet, MD, Assistant Professor; & Radiologist-in-Chief, St. Michael's Hospital,Anish Kirpalani, MD FRCPC, Asst. Professor & Radiologist-in-Chief, St. Michael's Hospital; Heidi C. Schmidt, MD, Professor & Radiologist-in-Chief, JDMI; Hayley Panet; Ania Z. Kielar, MD FRCPC, Assoc. Professor; Michael N. Patlas, MD FRCPC, Professor & Chair; Joseph Choi, Kate Hanneman, MD MPH, Assoc. Professor & VC, Research; Felipe Castillo Aravena, MD, Cardiothoracic Imaging Fellow

Abstract Oral Presentation: TRAINEE RESEARCH PRIZE AWARDEE: R1-SSCA09-3SEX-SPECIFIC ASSOCIATIONS OF LONG-TERM AIR POLLUTION EXPOSURE WITH CORONARY ARTERY STENOSIS ON CARDIAC CT
Authors: Felipe Castillo Aravena, MD, Cardiothoracic Imaging Fellow; Chloe DesRoche, MD,MSc, Scott Delaney, Rachel Nethery, Paaladinesh Thavendiranathan, MD, Heather Ross, Kate Hanneman, MD MPH, Assoc. Professor & VC, Research

Submission by: Dr. Lara Gabrielle Lim, MD, Body Imaging Clinical Fellow, St. Michael's Hospital
Presentation: First Impressions Count: Imaging Order Affects Ultrasound Accuracy for Renal Stone Assessment, How test sequence shapes diagnostic accuracy
Authors: Kai-Ho Fok, MD MSc FRCSC; Hui Ming Lin, HBSc MRT(R); Errol Colak, MD FRCPC, Assoc. Professor & Odette Professorship in Artificial Intelligence for Medical Imaging; Deelan Patel, MD FRCPC; Michael Ordon, MD MSc FRCSC; Guan Huang, MD FRCPC, Asst. Professor

Summary: We investigated the impact of imaging order on the accuracy of ultrasound in renal stone evaluation.

Lectures/Courses

Speaker: Dr. Claudia Martinez-Rios, MI Asst. Professor & Radiologist, The Hospital for Sick Children, Department of Diagnostic and Interventional Radiology
Lecture: R1-CPD10 Practical Approach to Pediatric Thyroid Nodules (Yes or No to TIRADS)
In session: Small Parts, Big Stories: Painting the Picture of Pediatric Soft Tissue Imaging
December 4 at 8 a.m.

Summary: This educational session focuses on the sonographic evaluation for pediatric thyroid nodules with updates in management. By the end of the session, audience members will have an increased confidence in their ability to diagnose thyroid nodules in children using ultrasound, with an improved awareness of current standards of care and reporting.

Learning Objective: Update learners on pediatric thyroid nodule guidelines and provide imaging examples to differentiate benign from malignant.

Speaker: Dr. Kartik Jhaveri, MI Professor & Director, Abdominal MRI, Abdominal Division, JDMI
Refresher Course: Rectal Cancer Treatment Response Assessment
December 3 from 8 a.m. – 9 a.m.

Refresher Course: Technical and Diagnostic Challenges in the Evaluation of Treatment Naïve and Treated Prostate Cancer
December 4 from 9:30 a.m. – 10:30 a.m.

Speaker: Dr. Oscar Navarro, Professor & Pediatric Radiology Program Director
Lecture: Challenging cases of lumps and bumps
In session: Small Parts, Big Stories: Painting the Picture of Pediatric Soft Tissue Imaging
Summary: This talk describes ultrasound findings of unusual pediatric lumps and bumps and atypical presentations of common lumps and bumps in children and will discuss a management plan.

Speaker: Dr. Elka Miller, Professor
Lecture: Tiny Lives, Big Emergencies: Fetal and Neonatal Neuroimaging Emergencies
Session: M1-CNR10 – Pediatric Intracranial Emergencies

Speaker: Dr. Anastasia Oikonomou, Professor, Cardiothoracic Division, Sunnybrook Health Sciences Centre
Educational Session (W1-CCH05): Lung Cancer Imaging: Pearls, Pitfalls, and Recent Advances
Lecture: Radiomics and Imaging Biomarkers in Lung Cancer
Learning Objectives:

• Introduction to the clinical and imaging biomarkers
• To discuss the current role of radiomics imaging biomarkers in metastatic and locally advanced lung cancer
• To discuss the challenges in the clinical adoption of radiomics imaging biomarkers and future directions

Speaker: Dr. Mary-Louise Greer, Professor & Division Lead, Pediatric Imaging
Educational Course: Dos and Dont's of Whole Body MR: Evaluation of Oncologic and Congenital Diseases Versus Celebrity Screening
Session Number: T8-CMS07
Title: WB-MRI MRI in the Evaluation of Pediatric Cancer Patients

Summary: Review of syndromic and non-syndromic pediatric cancer whole-body MRI indications and techniques.

Special Mentions

Reviewer for RadioGraphics Pediatric Panel:
Dr. Mary Louise Greer, MI Professor & Division Lead, Pediatric Imaging 

2024 Rising Star Award Session
Dr. Satheesh Krishna Jeyaraj, Associate Professor & Director of Abdominal & Pelvic CT, JDMI, was awarded the Rising Star Award at last year’s RSNA Annual Meeting, which came with the opportunity to create a session for the Douglas W. MacEwan Rising Star Track at this year’s meeting, RSNA 2025: Imaging the Individual.

Dr. Jeyaraj’s session will be joined by MI Assistant Professor Dr. Rajesh Bhayana, among other guest speakers.

Session: T3-RCP40 Cool Tools: Gen AI/LLMs for Radiology
Date: December 2, 2025 at 9:30 a.m.
Description: Discover practical generative AI tools beyond ChatGPT for radiology.  This session explores creative uses, advanced research capabilities, coding  assistants, and ways to keep up with rapidly changing AI developments. Find resources referenced during the talk here.

Learning Objectives:
1) Understand the currently available GenAI/LLMs tools and their capabilities.
2) Learn practical ways these tools are being used for day-to-day radiology and in education.
3) Develop lifelong learning techniques for AI/LLMs.

Session Speakers:
Dr. Woojin Kim, MD
T3-RCP40C AI Made Fun: How Radiologists can Practically Leverage Generative AI
Presentation details

Dr. Rajesh Bhayana, MD FRCPC
T3-RCP40D Cool Tools: GenAI/LLMs for Radiology
Presentation details

Dr. Satheesh Krishna, MD
T3-RCP40ECool Tools: GenAI/LLMs for Radiology
Presentation details

Dr. Benjamin Kwan, MD FRCPC
T3-RCP40FCool Tools: Integrating AI for Radiology Medical Education