Download This Paper
Preprints with The Lancet is a collaboration between The Lancet Group of journals and SSRN to facilitate the open sharing of preprints for early engagement, community comment, and collaboration. Preprints available here are not Lancet publications or necessarily under review with a Lancet journal. These preprints are early-stage research papers that have not been peer-reviewed. The usual SSRN checks and a Lancet-specific check for appropriateness and transparency have been applied. The findings should not be used for clinical or public health decision-making or presented without highlighting these facts. For more information, please see the FAQs.
Automated Deep Learning-Assisted Early Detection of Radiation-Induced Temporal Lobe Injury on Mri:A Multicenter Retrospective Analysis
25 Pages Posted: 1 Aug 2024
More...Abstract
Background: Early detection of radiation-induced temporal lobe injury (RTLI) on MRI is crucial for nasopharyngeal carcinoma (NPC) patients as it can help prevent injury progression and delay function impairment. This multicenter study aims to evaluate the benefits of an automated deep learning-based tool (RTLI-DM) for early detection of RTLI on MRI.
Methods: A total of 396 RTLI and 3181 non-RTLI patients were retrospectively included from a cohort of 6483 NPC patients who underwent radiotherapy and MRI follow-ups at four hospitals from January 2010 to June 2021. We developed and validated an automated RTLI-DM using the initial abnormal MRI of RTLI patients and final MRI of non-RTLI patients. The RTLI-DM consist of a bilateral temporal lobes segmentation model based on Unet++ and a diagnostic model based on a modified DenseNet-121 network. The multi-reader study was conducted to compare the performance of four radiologists (two trainees, two experienced readers) in interpreting images without and with RTLI-DM assistance. Sensitivity, specificity, false positives (FPs) and false negatives (FNs) were compared between two reading sessions.
Findings: Assisted reading with RTLI-DM significantly improve the sensitivity (60·5% [95% CI: 53·6-67·0%] vs. 83·5% [95% CI: 77·7-88·0%]; 79·5% [95% CI:73·3-84·5%] vs. 91·5% [95% CI: 86·7-94·7%]) and FNs (0·395 vs. 0·165; 0·205 vs. 0·085) for both trainees and experienced readers, although there was a slight decrease in specificity (99·0% [95% CI: 96·2-100%] vs. 93·0% [95% CI: 88·5-95·9%]) and increase in FPs (0·01 vs. 0·07) for trainees (P < 0·05). The averaged time for assisted reading reduced by 29 seconds, compared with unassisted reading.
Interpretation: For early detection of RTLI on MRI, the RTLI-DM assistant significantly improved the diagnostic performance of radiologists while reducing the reading time.
Funding: This study was supported by the National Natural Science Foundation of China (Grant Nos. 82071894 and 91959117), National High-end Foreign Experts Recruitment Plan (Grant Nos. G2022161002L).
Declaration of Interest: The authors have nothing to disclose.
Ethical Approval: The institutional review board of four hospitals approved this retrospective study, and the requirement for informed consent was waived.
Keywords: radiation-induced temporal lobe injury, early detection, deep learning, MRI
Suggested Citation: Suggested Citation