ORIGINAL RESEARCH article
Front. Phys.
Sec. Medical Physics and Imaging
Temporal trends and reliability of computed tomography scanner failures: evidence from a real-world time-series analysis
Provisionally accepted
Xiaoxiao Luan*
Sujuan Yu
Shaohua YinZhenlin LiuDongXu WangYing XiaoRuoyao PanBoqi JiaFeng Xu
Yun Tian
Rong Li- Peking University Third Hospital, Haidian, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Objective: We aimed to characterize temporal and component-specific patterns of CT malfunctions and identifying optimal monitoring intervals are critical to strengthening preventive maintenance and informed equipment management. Methods: This study collected failure data from three CT scanners (uCT790, uCT860, uCT960+) at a tertiary hospital, collecting from installation through December 2024. Time-series analysis was used to characterized annual and monthly failure counts, and Pie charts visualized component-level contributions. Reliability performance was quantified using mean time between failures (MTBF), mean time to repair (MTTR), mean time between maintenance (MTBM), mean repair time (MTR), mean maintenance time (MMT), and mean logistics delay time (MLDT). A multi-granularity time-series Poisson-Prophet model (monthly, bi-weekly, weekly) was performed to evaluate predictive accuracy using mean absolute error (MAE). Results: From 2019 to 2024, failure trajectories differed across CT scanners. The uCT790 showed a rise to five failures in 2023 followed by a decline to two in 2024. The uCT860 showed a steady increase, reaching six failures in 2024. The uCT960+ increased annually to seven failures in 2024. Failures were concentrated in the scanning table, power supply, and X-ray tube. Reliability analysis showed that the uCT790 achieved the highest MTBF (8554.03 hours), the uCT860 the shortest MTTR (6.60 hours), and the uCT960+ the lowest MLDT (3.47 hours). Across forecasting granularities, predictive accuracy improved with finer granularity: monthly MAE 0.33–1.50, biweekly 0–0.33, and weekly 0–0.17. Conclusions: Through integrated reliability assessment and time-series forecasting, this study delineated the distinct failure patterns of three CT scanners and identified bi-weekly forecasting as the most cost-effective temporal resolution, balancing predictive accuracy with resource efficiency.
Keywords: computed tomography, Hardware failure, Predictivemaintenance, Reliability, Time-series forecasting
Received: 19 Dec 2025; Accepted: 26 Jan 2026.
Copyright: © 2026 Luan, Yu, Yin, Liu, Wang, Xiao, Pan, Jia, Xu, Tian and Li. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Xiaoxiao Luan
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.