{"id":26304,"date":"2025-10-21T12:28:36","date_gmt":"2025-10-21T10:28:36","guid":{"rendered":"https:\/\/sano.science\/?post_type=seminars&p=26304"},"modified":"2026-01-14T21:24:10","modified_gmt":"2026-01-14T20:24:10","slug":"translating-ai-into-clinical-practice-lessons-from-the-childrens-hospital-of-philadelphia","status":"publish","type":"seminars","link":"https:\/\/sano.science\/seminars\/translating-ai-into-clinical-practice-lessons-from-the-childrens-hospital-of-philadelphia\/","title":{"rendered":"172. Translating AI into Clinical Practice: Lessons from the Children\u2019s Hospital of Philadelphia"},"content":{"rendered":"\n

Abstract:<\/h2>\n\n\n\n

Artificial intelligence is advancing rapidly in radiology, yet clinical translation remains constrained by factors that go beyond algorithm accuracy. In this seminar, I will share insights from my Visiting Scholar internship at the Children\u2019s Hospital of Philadelphia (CHOP) in the ImagineQuant Research Group, combining clinical immersion in neuroradiology with hands-on translational AI research.
I will discuss why infrastructure matters for AI adoption, highlighting our three-instance PACS prototype for governed, real-time AI integration, and connect these outcomes to my recent research direction towards Uncertainty-Aware Explainable AI for brain metastasis assessment.
This seminar will cover results from our analysis of the BraTS-METS Lighthouse 2025 Challenge data that establish STAPLE-based probabilistic reference standards for brain tumor segmentation and quantify inter- and intra-rater variability. I will also introduce our \u201cDisagreement Profile\u201d framework that converts multi-metric segmentation disagreement into interpretable profiles that guide iterative, weighted consensus.<\/p>\n\n\n\n

<\/div>\n\n\n\n

About the author:<\/h2>\n\n\n\n

Monika Pytlarz<\/a> is a PhD student working on AI analysis of tumors, with a focus on gliomas. Her research combines radiology, histopathology, and genomics to improve tumor classification and patient risk stratification, despite the biological heterogeneity of gliomas. The project\u2019s computational image analysis of immunostaining provides insights into the tumor immune microenvironment, supporting the development of more precise therapeutic strategies. She also investigates cross-modal image translation, including the synthesis of histology from MRI scans.<\/p>\n\n\n\n

Monika obtained a BSc in Electroradiology from Collegium Medicum of the Jagiellonian University and an MSc in Bioinformatics from JU, and brings clinical experience from hospital diagnostic imaging departments. She explores translating AI models into clinical practice via computer-aided diagnosis tools and PACS integration.<\/p>\n\n\n\n

rsipvision.com<\/a><\/p>\n\n\n\n

<\/div>\n\n\n\n
\n