Deep learning based automatic detection algorithm for acute intracranial haemorrhage: a pivotal randomized clinical trial

Acute intracranial haemorrhage (AIH) is a potentially life-threatening emergency that requires prompt and accurate assessment and management. This study aims to develop and validate an artificial intelligence (AI) algorithm for diagnosing AIH using brain-computed tomography (CT) images. A retrospective, multi-reader, pivotal, crossover, randomised study was performed to validate the performance of an AI algorithm trained using 104,666 slices from 3010 patients. Brain CT images (12,663 slices from 296 patients) were evaluated by nine reviewers in three subgroups (non-radiologist physicians, n = 3; board-certified radiologists, n = 3; neuroradiologists, n = 3) with and without AI assistance. Brain CT interpretation with AI assistance yielded significantly higher diagnostic accuracy than without AI assistance (0.9703 vs. 0.9471, p < 0.0001, patient-wise). Non-radiologist physicians showed the greatest improvement with AI assistance; board-certified radiologists also improved significantly, while neuroradiologists showed a non-significant trend toward improvement. AI assistance improved diagnostic performance for AIH detection on brain CT, with the most notable impact among non-radiologist physicians.

Tae Jin Yun, Jin Wook Choi, Miran Han, Woo Sang Jung, Seung Hong Choi, Roh-Eul Yoo, In Pyeong Hwang