Hemostasis Today

June, 2026
June 2026
M T W T F S S
1234567
891011121314
15161718192021
22232425262728
2930  
Mahsa Asadi Anar: The Next Challenge for AI in Stroke Imaging
Jun 18, 2026, 12:34

Mahsa Asadi Anar: The Next Challenge for AI in Stroke Imaging

Mahsa Asadi Anar, Founder of Mediverse Research Center, shared on LinkedIn about a recent article she and her colleagues co-authored, published in BioMedical Engineering OnLine, adding:

”Stroke imaging is one of the areas where AI can make a real clinical difference, but only if the models are tested carefully.

In our new paper we meta-analyzed 101 studies on machine learning models for acute ischemic stroke lesion segmentation.

The results were encouraging.

Deep learning models, especially U-Net-based architectures, showed strong performance, with pooled estimates of Dice 0.84, AUC 0.91, accuracy 0.89, sensitivity 0.85, and specificity 0.93.

But the main message is not just that AI performs well.

The field now needs stronger external validation, clearer reporting, and real testing in clinical workflows before these tools can be trusted at scale.

Love the outcomes Mediverse had this past month!

Cheers to more.”

Title: Machine learning models of segmentation in acute ischemic stroke: a systematic review and meta-analysis

Authors: Sadaf Salehi, Sahar Birzhandi, Zeinab Torbati Aghdam, Raha Amirvala, Hediyeh Dasoomi, Zayd M.Yaghi, Mohammad Saeed Soleimani Meigoli, Zahra Moghari Hesari, Kiana Taghipoor, Kiana saffarian, Mohsen Shahba, Alireza Ghaedamini, Sasan Ghazanafar Ahari, Melika Arab Bafrani, Matin Akhbari, Ali Babapour, Amirhossein Rigi, Samira Peiravi, Mahsa Asadi Anar

Mahsa Asadi Anar: The Next Challenge for AI in Stroke Imaging

Stay updated on all scientific advances with Hemostasis Today.