1475
clinical study

Deep learning algorithm in detecting intracranial hemorrhages on emergency computed tomographies

Materials & Methods

In a retrospective multi-center cohort study, consecutive emergency non-contrast HCT scans were analyzed by a commercially available ICH detection software (Aidoc, Tel Aviv, Israel). Discrepancies between AI analysis and initial radiology report (RR) were reviewed by a blinded neuroradiologist to determine the number of additional ICHs detected and evaluate reasons leading to errors.

Results

4,946 HCT (May 2020 to Sept. 2020) from 18 hospitals were included in the analysis. 205 reports (4.1%) were classified as hemorrhages by both radiology report and AI. Out of a total of 162 (3.3%) discrepant reports, 62 were confirmed as hemorrhages by the reference neuroradiologist. 33 ICHs were identified exclusively via RRs. The AI algorithm detected an additional 29 instances of ICH, missed 12.4% of ICH and overcalled 1.9%; RRs missed 10.9% of ICHs and overcalled 0.2%. Many of the ICHs missed by the AI algorithm were located in the subarachnoid space (42.4%) and under the calvaria (48.5%). 85% of ICHs missed by RRs occurred outside of regular working-hours. Calcifications (39.3%), beam-hardening artifacts (18%), tumors (15.7%), and blood vessels (7.9%) were the most common reasons for AI overcalls. ICH size, image quality, and primary examiner experience were not found to be significantly associated with likelihood of incorrect AI results.

Conclusions

Complementing human expertise with AI resulted in a 12.2% increase in ICH detection. The AI algorithm overcalled 1.9% HCT.  

Read Study

Aidoc Ask the Experts Webinar

Popular Posts

Article
What Should Hospital Decision Makers Ask AI Vendors?
Article
Infographic: Cultivating the "Giving Tree" of AI
Article
What Does an AI-Powered PERT Workflow Look Like?
Article
Top 3 Takeaways from RSNA 2023

Stay Connected on AI Trends and Insights

Follow us on social

From Our Blog

View all

© Copyright 2023 Aidoc