Arshia Javidan

143 posts

Arshia Javidan

Arshia Javidan

@APJavidan

PGY-5 Vascular Surgery @UofTVascular | MD/MSc @uoftmedicine & @ihpmeuoft 2021 | @MacBHSc 2017 | Machine learning, MedEd, systems change, and deadlifts

Toronto, Ontario Katılım Kasım 2018
964 Takip Edilen577 Takipçiler
Arshia Javidan
Arshia Javidan@APJavidan·
Excited to announce the project that @TiamFeridooni and I have been working on for the past year: an augmented LLM that combines the capabilities of open access models with publicly available vascular surgery resources for improved performance. Our first paper with more to come!
vasc ai@vasc_ai

Our most recent study published in @JVascSurgVI demonstrating the profiency of vasc.ai in responding to VESAP-5 multiple choice qustionare. 🔍 Study Highlights: chat.vasc.ai performed with a 93.8% accuracy in answering complex, domain-specific questions, surpassing general AI models such as GPT-4o (77.7%). This model minimizes the risk of misinformation, often seen in general-purpose AIs, by focusing on logical accuracy, making it a reliable tool for vascular surgery training, medical education, and patient care. You can read the full study here: jvsvi.org/article/S2949-… For academic vascular surgeons, chat.vasc.ai presents a significant advancement in how specialized knowledge is curated and applied. Its potential to support clinical assessments, enhance educational resources, and integrate rapidly evolving research makes it a valuable tool in both academic and clinical settings. #VascularSurgery #SurgicalEducation #RAG #AI #LLM

English
0
0
9
916
Roland Xu
Roland Xu@chloroprocaine·
Leaving @Sunnybrook today as an Anesthesia Fellow and returning tomorrow as a Staff Anesthesiologist. Thank you to everyone who have helped me get to this point in life, it truly takes a village. To new beginnings!
Roland Xu tweet media
English
6
1
89
4.5K
Arshia Javidan
Arshia Javidan@APJavidan·
Exploring the Fragility Index (FI) in Vascular Surgery Trials: A Thread. The FI assesses the robustness of statistically significant findings in randomized controlled trials (RCTs), focusing on the minimum number of event conversions required to change the outcome's significance.
English
2
5
18
1.9K