EUVIMEDEUVIMED entdecken
Vergleichsstudie

Mind the Gap: Human and AI Uncertainties in Cardiac MRI Segmentation

Qiao Lin iD, Xin Chen, Chao Chen, Nikesh Jathanna, Peter P Swoboda, Shahnaz Jamil-Copley, Jonathan M Garibaldi

Cardiovascular Engineering and Technology · 2026-06 · 17 (3) · 268-280

DOI: 10.1007/s13239-026-00832-2 ↗

Kurz gefasst

Vergleichsstudie in der Zeitschrift Cardiovascular Engineering and Technology, online erschienen am 06.04.2026. Verfasst von Qiao Lin und 6 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science, University of Nottingham, Ningbo, China. qiao.lin@nottingham.edu.cn. Hauptthemen laut MEDLINE: Image Interpretation, Computer-Assisted, Magnetic Resonance Imaging, Artificial Intelligence, Heart. Die Arbeit nennt 17 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

PURPOSE: This study conducts quantitative and qualitative analyses to investigate the relationship between human-annotated and AI-derived uncertainty in cardiac MRI segmentation, aiming to enhance the reliability of AI-based cardiac MRI segmentation models and foster better human-AI collaboration. METHODS: The CMRI dataset used in the experiments consists of 483 scans, each with two types of labels: an annotated segmentation mask and an uncertainty score per CMRI slice, both provided by clinicians. First, the AI-derived uncertainty estimated by the fuzzy-based algorithm is utilized to indicate the quality of segmentation. Multiple levels of uncertainties are derived from the method, including class-wise, slice-wise, subject-wise, etc. Subsequently, they are compared to the human-annotated uncertainty scores. Finally, qualitative analyses are conducted with clinicians to investigate all uncertainty measures potentially impacting real clinical applications. RESULTS: Experimental results show a strong inverse correlation between AI-derived uncertainty and Dice score, a standard metric for segmentation quality, indicating that lower uncertainty predicts higher segmentation quality. Additionally, it is found that human-annotated uncertainty coincides with AI-derived uncertainty for some anatomical structures (e.g., papillary muscle). However, high human-annotated uncertainty does not necessarily correlate with low AI segmentation quality, and there is no obvious association between human-annotated uncertainty and the size of the segmented structure. Concluded from qualitative analysis with clinicians, humans are better at utilizing prior knowledge (e.g. cardiac structural and contextual information) for uncertainty scoring, while the current AI method lacks this capability and is mainly data-driven for decision-making and uncertainty estimation. CONCLUSION: AI-derived uncertainty could be utilized as quality control for CMRI segmentation. Humans utilize structural and contextual information to formulate uncertainty, while AI models currently lack this capability.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Image Interpretation, Computer-AssistedMagnetic Resonance ImagingArtificial IntelligenceHeartHumansUncertaintyPredictive Value of TestsReproducibility of ResultsDatabases, FactualDatasets as TopicFuzzy Logic

Stichwörter der Autor:innen

AI-derived uncertaintyClinical settingsHuman-annotated uncertaintyQuality controlSemantic segmentation

Autor:innen und Institutionen

  • School of Computer Science, University of Nottingham, Ningbo, China. qiao.lin@nottingham.edu.cn.
  • School of Computer Science, University of Nottingham, Nottingham, NG81BB, UK.
  • Nottingham University Hospitals NHS Trust, Nottingham, NG51PB, UK.
  • Leeds Institute of Cardiovascular and Metabolic Medicine, University of Leeds, Leeds, LS29JT, UK.
  • School of Computer Science, University of Nottingham, Ningbo, China.

ORCID-Profile: Qiao Lin

Zitieren

APA

Lin, Q., Chen, X., Chen, C., Jathanna, N., Swoboda, P. P., Jamil-Copley, S., & Garibaldi, J. M. (2026). Mind the Gap: Human and AI Uncertainties in Cardiac MRI Segmentation. Cardiovascular Engineering and Technology, 17(3), 268-280. https://doi.org/10.1007/s13239-026-00832-2

Vancouver

Lin Q, Chen X, Chen C, Jathanna N, Swoboda PP, Jamil-Copley S, et al. Mind the Gap: Human and AI Uncertainties in Cardiac MRI Segmentation. Cardiovascular Engineering and Technology. 2026;17(3):268-280. doi:10.1007/s13239-026-00832-2

BibTeX
@article{Lin2026,
  author = {Lin, Qiao and Chen, Xin and Chen, Chao and Jathanna, Nikesh and Swoboda, Peter P and Jamil-Copley, Shahnaz and Garibaldi, Jonathan M},
  title = {Mind the Gap: Human and AI Uncertainties in Cardiac MRI Segmentation},
  journal = {Cardiovascular Engineering and Technology},
  year = {2026},
  volume = {17},
  number = {3},
  pages = {268--280},
  doi = {10.1007/s13239-026-00832-2}
}
Kostenlos starten →