跳到正文
原文
Nature 健康护理主题·· 4 天前精选AI 评分66

AI辅助超声无创检测慢性肾脏病中重度间质纤维化与肾小管萎缩:系统综述与Meta分析

AI-assisted ultrasonography for noninvasive detection of moderate-to-severe interstitial fibrosis and tubular atrophy in chronic kidney disease: a systematic review and meta-analysis

AI 导读

本Meta分析纳入14项研究、3,776名参与者,评估AI辅助超声检测慢性肾脏病患者≥25%中重度间质纤维化与肾小管萎缩的诊断准确性。汇总敏感性为0.821,特异性为0.828,AUROC为0.880(95% CI:0.815–0.924)。SWE与深度学习模型数值表现较高,但亚组分析属探索性结论,需进一步大规模前瞻性验证。

推荐理由

该Meta分析汇总14项研究的诊断性能数据,AI辅助超声对肾纤维化检测呈现平衡的敏感性与特异性。

正文 · 原文

Scientific Reports (2026) Cite this article

We’re sharing this article early to provide faster access to peer-reviewed, accepted research. It is citable and carries a permanent DOI. This version is subject to further edits and will be replaced automatically by the final Version of Record. All legal disclaimers apply.

Abstract

Interstitial fibrosis and tubular atrophy (IFTA) is a common pathway of irreversible kidney damage. Although renal biopsy remains the standard for IFTA assessment, its invasive nature limits repeated monitoring. Ultrasonography (US) assisted by artificial intelligence (AI) has emerged as a potential noninvasive alternative, but its diagnostic performance for IFTA remains to be systematically evaluated. PubMed, Scopus, the Cochrane Library, Wanfang Database, and China National Knowledge Infrastructure were searched through October 2025. Studies evaluating AI-assisted B-mode ultrasonography, shear wave elastography (SWE), or combined US + SWE for detecting moderate-to-severe IFTA (≥ 25%), with renal biopsy as the reference standard, were included. The primary outcomes were AUROC, sensitivity, and specificity. Sensitivity and specificity were jointly synthesized using a bivariate random-effects model, with estimation of the summary operating point and corresponding 95% confidence and prediction regions. AUROC was separately pooled using a random-effects model. Diagnostic accuracy and other diagnostic measures were evaluated as secondary outcomes. Fourteen studies including 3,776 participants were eligible, of whom 1,244 contributed to pooled diagnostic performance estimates. The bivariate random-effects model yielded a summary sensitivity of 0.821 and summary specificity of 0.828. The pooled AUROC was 0.880 (95% CI, 0.815–0.924). SWE and deep learning models showed numerically higher performance, but subgroup findings were exploratory because of limited study numbers and wide confidence intervals. AI-assisted ultrasonography serves as a reliable, noninvasive diagnostic tool for identifying moderate-to-severe IFTA, demonstrating balanced sensitivity and specificity. While SWE and DL architectures exhibited favorable numeric trends, these subgroup observations remain exploratory and warrant further large-scale prospective validation.

Funding

This work was financially supported of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan. Grant no. DP2-TMU-114-HIS-01

Author information

Author notes

  1. Wei-Jou Yin and Yung-Hsin Lan contributed equally to this work.

Authors and Affiliations

  1. Department of Medical Education, School of Medicine, College of Medicine, Taipei Medical University, Taipei City, Taiwan

    Wei-Jou Yin & Yung-Hsin Lan

  2. Taipei Veterans General Hospital, Taipei City, Taiwan

    Wei-Jou Yin

  3. National Taiwan University Hospital, Taipei, Taiwan

    Yung-Hsin Lan

  4. Division of nephrology, Department of Internal Medicine, School of Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan

    Chih-Chin Kao & Yen-Chung Lin

  5. Division of nephrology, Department of internal medicine, Taipei Medical University Hospital, Taipei, Taiwan

    Chih-Chin Kao & Yen-Chung Lin

  6. Division of general medicine, School of medicine, College of medicine, Taipei Medical University, Taipei, Taiwan

    Yen-Chung Lin

Authors

  1. Wei-Jou Yin
  2. Yung-Hsin Lan
  3. Chih-Chin Kao
  4. Yen-Chung Lin

Corresponding author

Correspondence to Yen-Chung Lin.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yin, WJ., Lan, YH., Kao, CC. et al. AI-assisted ultrasonography for noninvasive detection of moderate-to-severe interstitial fibrosis and tubular atrophy in chronic kidney disease: a systematic review and meta-analysis. Sci Rep (2026). https://doi.org/10.1038/s41598-026-80053-8

Download citation

  • Received: 04 March 2026

  • Accepted: 04 November 2026

  • Published: 16 November 2026

  • DOI: https://doi.org/10.1038/s41598-026-80053-8

Keywords

来源:Nature 健康护理主题 · nature.com