Robustness of human vs. AI measurements under progressive image degradation

Jevsikov, Jevgeni, Stowell, Catherine C., Ng, Tiffany, Unsworth, Beth, Zolgharni, Massoud ORCID logoORCID: https://orcid.org/0000-0003-0904-2904, Francis, Darrel P., Manisty, Charlotte H. and Shun-Shin, Matthew J. (2025) Robustness of human vs. AI measurements under progressive image degradation. In: International Conference on AI in Healthcare, 8-10 September, 2025, Cambridge, United Kingdom.

[thumbnail of 270-277.pdf] PDF
270-277.pdf - Accepted Version
Restricted to Repository staff only until 25 August 2026.

Download (2MB) | Request a copy

Abstract

Echocardiography is widely used in cardiac imaging for real-time, non-invasive assessment of heart anatomy and function. However, image interpretation can be challenging due to inherent noise, particularly speckle. This study compares artificial intelligence (AI) and human experts in interpreting parasternal long-axis (PLAX) echocardiographic images under increasing noise levels. While both AI and human performance declined with image quality, AI consistently outperformed humans, indicating its potential to enhance clinical decision-making in challenging imaging scenarios.

Item Type: Conference or Workshop Item (Paper)
ISBN: 9783032006561
Identifier: 10.1007/978-3-032-00656-1_19
Page Range: pp. 261-268
Identifier: 10.1007/978-3-032-00656-1_19
Keywords: Echocardiography; computer vision; deep Learning
Subjects: Computing > Intelligent systems
Date Deposited: 03 Oct 2025 10:04
Last Modified: 03 Oct 2025 15:00
URI: https://repository.uwl.ac.uk/id/eprint/14142
Sustainable Development Goals: Goal 3: Good Health and Well-Being

Actions (login required)

View Item View Item

Menu