From fragile pipelines to reliable systems: error-correctable cascades in medical AI
4 affiliations
- 1Department of Pathology, State Key Laboratory of Oncology in South China, Guangdong Provincial Clinical Research Center for Cancer, Sun Yat- sen University Cancer Center, Guangzhou, 510060 China
- 2Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China
- 3Department of Cellular and Molecular Diagnostics Center, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, China
- 4School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, 510006 China
Abstract
Background: Artificial intelligence (AI) has achieved expert-level performance in medical imaging, but its clinical translation requires a reliable and interpretable system design. Decision-level hierarchical cascaded pipelines are widely used in medical AI, yet their benefits and limitations require further investigation.
Methods: We compared decision-level hierarchical cascaded and parallel architectures in medical imaging and revealed a trade-off between error vulnerability and diagnostic transparency. Based on these observations, we developed an error-aware framework integrating automated error detection, adaptive correction, and confidence-guided human-AI collaboration. The framework was evaluated on over 10,000 patient cases, including 6,103 colorectal pathology and 7,164 cervical cytology samples, with additional validation in endoscopic and ophthalmological imaging.
Results: Decision-level hierarchical cascaded systems, although susceptible to error amplification, were more aligned with hierarchical clinical reasoning and enabled more transparent decision pathways. The proposed framework mitigated error propagation through automated detection and correction. Combined with confidence-guided human-AI collaboration, it achieved absolute accuracy gains of 16.12 and 10.11 percentage points in colorectal pathology and cervical cytology, respectively, compared with parallel systems, with consistent gains across other imaging tasks.
Conclusions: Our study highlights the trade-off between error propagation and interpretability in hierarchical cascaded AI systems and demonstrates that error-aware design can improve their robustness and clinical applicability, contributing to the development of more reliable and human-centered medical AI.
Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s12916-026-05120-3.
Related
- Generative AI, cyberchondria, and digital health inequality: a public health challenge for Somalia and other fragile health systems
- Assisting patients with reliable information gathering: A guide to building efficient QA systems for medical use
- Defining a subgroup within autism leads to potential enrichment criteria in investigational therapeutic development: Comparing resting state gamma band power to a fragile X syndrome reference sample
- Fragile X Syndrome Diagnosis Patterns in a Multistate, Population-Based Cohort
- Spontaneous Coronary Artery Dissection in Fragile X Syndrome
BibTeX
@article{zheng2026from,
title = {{From fragile pipelines to reliable systems: error-correctable cascades in medical AI}},
author = {Ke Zheng and Xueyi Zheng and Chunhua Qu and Jue Wang and Xiaobo Wen and Xinke Zhang and Sha Fu and Ruixuan Wang and Dan Xie and Muyan Cai},
journal = {BMC Medicine},
year = {2026},
month = aug,
doi = {10.1186/s12916-026-05120-3},
url = {https://doi.org/10.1186/s12916-026-05120-3},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - From fragile pipelines to reliable systems: error-correctable cascades in medical AI
AU - Ke Zheng
AU - Xueyi Zheng
AU - Chunhua Qu
AU - Jue Wang
AU - Xiaobo Wen
AU - Xinke Zhang
AU - Sha Fu
AU - Ruixuan Wang
AU - Dan Xie
AU - Muyan Cai
PY - 2026
DA - 2026/08/06
JO - BMC Medicine
DO - 10.1186/s12916-026-05120-3
UR - https://doi.org/10.1186/s12916-026-05120-3
AB - Background: Artificial intelligence (AI) has achieved expert-level performance in medical imaging, but its clinical translation requires a reliable and interpretable system design. Decision-level hierarchical cascaded pipelines are widely used in medical AI, yet their benefits and limitations require further investigation. Methods: We compared decision-level hierarchical cascaded and parallel architectures in medical imaging and revealed a trade-off between error vulnerability and diagnostic transparency. Based on these observations, we developed an error-aware framework integrating automated error detection, adaptive correction, and confidence-guided human-AI collaboration. The framework was evaluated on over 10,000 patient cases, including 6,103 colorectal pathology and 7,164 cervical cytology samples, with additional validation in endoscopic and ophthalmological imaging. Results: Decision-level hierarchical cascaded systems, although susceptible to error amplification, were more aligned with hierarchical clinical reasoning and enabled more transparent decision pathways. The proposed framework mitigated error propagation through automated detection and correction. Combined with confidence-guided human-AI collaboration, it achieved absolute accuracy gains of 16.12 and 10.11 percentage points in colorectal pathology and cervical cytology, respectively, compared with parallel systems, with consistent gains across other imaging tasks. Conclusions: Our study highlights the trade-off between error propagation and interpretability in hierarchical cascaded AI systems and demonstrates that error-aware design can improve their robustness and clinical applicability, contributing to the development of more reliable and human-centered medical AI. Supplementary Information: The online version contains supplementary material available at https://doi.org/10.1186/s12916-026-05120-3.
SN - 1741-7015
ER -