Artificial intelligence software for assessing brain ischemic penumbra/core infarction on computed tomography...

Area of Science:

• Neurology
• Medical Imaging
• Artificial Intelligence

Background:

• Artificial intelligence (AI) is increasingly used in medical diagnosis, necessitating real-world accuracy evaluations.
• Assessing the accuracy of AI in medical diagnosis is critical for patient safety and effective treatment.

Purpose of the Study:

• To evaluate the diagnostic accuracy of Shukun AI software in determining ischemic penumbra and core infarction.
• To assess AI performance in acute ischemic stroke patients with large vessel occlusion undergoing mechanical thrombectomy.

Main Methods:

• Consecutive acute stroke patients with large vessel occlusion undergoing mechanical thrombectomy (MT) were analyzed.
• Shukun AI assessed computed tomography angiography (CTA) and perfusion imaging.
• AI diagnoses were compared against expert neurointerventional diagnoses, with discrepancies considered inaccuracies.

Main Results:

• The study included 22 patients, with a 90.9% recanalization rate and 63.6% achieving favorable outcomes (mRS 0-2) at 3 months.
• Shukun AI's computed tomography (CT) perfusion diagnosis was found inaccurate in 3 patients, resulting in a 13.6% inaccuracy rate.
• Mechanical thrombectomy (MT) was performed on patients post-AI assessment.

Conclusions:

• The Shukun AI software demonstrated limitations in accurately assessing the ischemic penumbra.
• Accurate assessment of ischemic penumbra is vital for guiding mechanical thrombectomy decisions.
• Integrating diverse data, including clinical information and various imaging modalities (CT, CTA, MRI), is essential for optimal MT decision-making.