EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Artificial Intelligence for Liver Cancer Diagnosis: Integrating Image Analysis, Algorithms, and Clinical Applications

Reza Atazadegan Jahromi, Amirhassan Abootalebzadeh, Maryam Abbasi, Ali Sharafkhah, Akbar Sarvari

InfoScience Trends · 2025

DOI: 10.61882/ist.202502.10.04 ↗

Kurz gefasst

Fachartikel in der Zeitschrift InfoScience Trends, 2025. Verfasst von Reza Atazadegan Jahromi und 4 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Artificial intelligence (AI) is increasingly applied to liver cancer diagnosis, where multiphase CT, dynamic MRI, and ultrasound/CEUS provide complex but essential imaging data. These modalities present challenges of integration, standardization, and interpretability, making them a critical testbed for AI research. This systematic review evaluated how AI algorithms are used in diagnostic imaging of focal liver lesions, examining algorithm families, integration strategies, diagnostic performance, methodological quality, and areas of maturity and limitation. A comprehensive search of PubMed, Ovid MEDLINE, IEEE Xplore, Scopus, and Google Scholar (2010–2025) was conducted. Studies applying AI or machine learning to radiologic diagnosis of focal liver lesions were included. Two reviewers independently screened titles, abstracts, and full texts, with discrepancies resolved by consensus. Risk of bias was assessed using QUADAS-2, and AI-specific reporting quality was appraised with CLAIM. Data were extracted on imaging protocols, algorithm design, validation strategies, and performance metrics. From 2,180 records, 32 studies met inclusion criteria: CT (n≈13), MRI (n≈13), and US/CEUS (n≈6). Common strategies included multi-phase/sequence fusion, cascaded detection–classification, clinical data fusion, and LI-RADS–aware feature modeling. Externally validated multiphase CT with clinical integration and large multiclass MRI pipelines achieved the strongest and most consistent performance. Ultrasound/CEUS pipelines demonstrated promise but were often single-center. Risk of bias clustered in patient selection and index test domains, with frequent lack of external validation, incomplete reporting, and limited calibration or uncertainty analysis. AI has demonstrated promising diagnostic capability in liver cancer imaging, particularly with phase-aware fusion and clinically integrated pipelines. However, limited external validation, methodological heterogeneity, and uneven reporting constrain clinical translation. Prospective, multi-center, standardized studies with transparency and robustness testing are needed to advance toward reliable deployment.

Zitieren

APA

Jahromi, R. A., Abootalebzadeh, A., Abbasi, M., Sharafkhah, A., & Sarvari, A. (2025). Artificial Intelligence for Liver Cancer Diagnosis: Integrating Image Analysis, Algorithms, and Clinical Applications. InfoScience Trends. https://doi.org/10.61882/ist.202502.10.04

Vancouver

Jahromi RA, Abootalebzadeh A, Abbasi M, Sharafkhah A, Sarvari A. Artificial Intelligence for Liver Cancer Diagnosis: Integrating Image Analysis, Algorithms, and Clinical Applications. InfoScience Trends. 2025. doi:10.61882/ist.202502.10.04

BibTeX
@article{Jahromi2025,
  author = {Jahromi, Reza Atazadegan and Abootalebzadeh, Amirhassan and Abbasi, Maryam and Sharafkhah, Ali and Sarvari, Akbar},
  title = {Artificial Intelligence for Liver Cancer Diagnosis: Integrating Image Analysis, Algorithms, and Clinical Applications},
  journal = {InfoScience Trends},
  year = {2025},
  doi = {10.61882/ist.202502.10.04}
}
Kostenlos starten →