Dataset Information

Predictive accuracy of machine learning for radiation-induced temporal lobe injury in nasopharyngeal carcinoma patients: a systematic review and meta-analysis.

ABSTRACT:

Background

Radiotherapy is a common treatment for nasopharyngeal carcinoma (NPC) but can cause radiation-induced temporal lobe injury (RTLI), resulting in irreversible damage. Predicting RTLI at the early stage may help with that issue by personalized adjustment of radiation dose based on the predicted risk. Machine learning (ML) models have recently been used to predict RTLI but their predictive accuracy remains unclear because the reported concordance index (C-index) varied widely from around 0.31 to 0.97. Therefore, a meta-analysis was needed.

Methods

The PubMed, Web of Science, Embase, and Cochrane Library databases were searched from inception to November 2022. Studies that fully develop one or more ML risk models of RTLI after radiotherapy for NPC were included. The Prediction model Risk Of Bias Assessment Tool (PROBAST) was used to assess the risk of bias in the included research. The primary outcome of this review was the C-index, specificity (Spe), and sensitivity (Sen).

Results

The meta-analysis included 14 studies with 15,573 NPC patients reporting a total of 72 prediction models. Overall, 94.44% of models were found to have a high risk of bias. Radiomics was included in 57 models, dosimetric predictors in 28, and clinical data in 27. The pooled C-index for ML models predicting RTLI was 0.77 [95% confidence interval (CI): 0.75-0.79] in the training set and 0.78 (95% CI: 0.75-0.81) in the validation set. The pooled Sen was 0.75 (95% CI: 0.69-0.80) in the training set and 0.70 (95% CI: 0.66-0.73) in the validation set and the pooled Spe was 0.78 (95% CI: 0.73-0.82) in the training set and 0.79 (95% CI: 0.75-0.82) in the validation set. Models with radiomics and clinical data achieved the most excellent discriminative performance, with a pooled C-index of 0.895.

Conclusions

ML models can accurately predict RTLI at an early stage, allowing for timely interventions to prevent further damage. The kind of ML methods and the selection of predictors may influence the predictive accuracy.

SUBMITTER: Li Y

PROVIDER: S-EPMC10583015 | biostudies-literature | 2023 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predictive accuracy of machine learning for radiation-induced temporal lobe injury in nasopharyngeal carcinoma patients: a systematic review and meta-analysis.

Li Yiling Y Gong Fengyuan F Guo Yangyang Y Ng Wai Tong WT Mejia Michael Benedict A MBA Nei Wen-Long WL Wang Cuicui C Jin Zhanguo Z

Translational cancer research 20230825 9

<h4>Background</h4>Radiotherapy is a common treatment for nasopharyngeal carcinoma (NPC) but can cause radiation-induced temporal lobe injury (RTLI), resulting in irreversible damage. Predicting RTLI at the early stage may help with that issue by personalized adjustment of radiation dose based on the predicted risk. Machine learning (ML) models have recently been used to predict RTLI but their predictive accuracy remains unclear because the reported concordance index (C-index) varied widely from ...[more]

PMID: 37859745

Similar Datasets

Project description:BackgroundRadiotherapy is the mainstay of treatment for nasopharyngeal carcinoma. Radiation-induced temporal lobe injury (TLI) can regress or resolve in the early phase, but it is irreversible at a later stage. However, no study has proposed a risk-based follow-up schedule for its early detection. Planning evaluation is difficult when dose-volume histogram (DVH) parameters are similar and optimization is terminated.MethodsThis multicenter retrospective study included 6065 patients between 2014 and 2018. A 3D ResNet-based deep learning model was developed in training and validation cohorts and independently tested using concordance index in internal and external test cohorts. Accordingly, the patients were stratified into risk groups, and the model-predicted risks were used to develop risk-based follow-up schedules. The schedule was compared with the Radiation Therapy Oncology Group (RTOG) recommendation (every 3 months during the first 2 years and every 6 months in 3-5 years). Additionally, the model was used to evaluate plans with similar DVH parameters.FindingsOur model achieved concordance indexes of 0.831, 0.818, and 0.804, respectively, which outperformed conventional prediction models (all P < 0.001). The temporal lobes in all the cohorts were stratified into three groups with discrepant TLI-free survival. Personalized follow-up schedules developed for each risk group could detect TLI 1.9 months earlier than the RTOG recommendation. According to a higher median predicted 3-year TLI-free survival (99.25% vs. 99.15%, P < 0.001), the model identified a better plan than previous models.InterpretationThe deep learning model predicted TLI more precisely. The model-determined risk-based follow-up schedule detected the TLI earlier. The planning evaluation was refined because the model identified a better plan with a lower risk of TLI.FundingThe Sun Yat-sen University Clinical Research 5010 Program (2015020), Guangdong Basic and Applied Basic Research Foundation (2022A1515110356), Medical Scientific Research Foundation of Guangdong Province (A2022367), and Guangzhou Science and Technology Program (2023A04J1788).

Project description:The natural course of radiation encephalopathy following nasopharyngeal carcinoma (NPC) radiotherapy remains poorly understood. The present study aimed to investigate the magnetic resonance imaging (MRI) characteristics and evolution of radiation encephalopathy. A series of 162 follow-up MRI examinations from 68 NPC patients with radiation encephalopathy in the temporal lobes were analyzed retrospectively. Each component of radiation encephalopathy was defined as follows: i) contrast enhanced lesions were enhanced lesions on contrast enhanced T1-weighted images (T1WI); ii) white matter lesions were lesions of homogeneous hyperintensity on T2-weighted images (T2WI) and hypointensity on T1WI; iii) cysts were round or oval well-defined lesions of hyperintensity on T2WI; iv) hemosiderin deposition were nodular or annular hypointense lesions with lower hypointense than normal white matter on both T1WI and T2WI; v) gray matter lesions were defined as disruption or erosion of hyperintensity in the cortex on T2WI. Contrast enhanced lesions, white matter lesions, gray matter lesions, cysts and hemosiderin deposition were detected in 105 (100.0%), 98 (93.3%), 94 (89.5%), 2 (1.7%) and 2 (1.7%) cases of the 105 initial diagnosed temporal lobe lesions. Contrast enhanced lesions were the most commonly observed, followed by white matter lesions, gray matter lesions, temporal lobe atrophy, cysts and hemosiderin deposition. In addition, 12 new lesions were identified during the follow-up, 4 of which presented as solid enhanced nodular lesions. Importantly, in 11 of the 117 (9.4%) affected temporal lobes, solid enhanced nodular lesions were observed to be the only initial abnormalities to occur. For those enhanced nodular lesions measuring <0.8 cm, no necrosis could be detected. On the contrary, all the contrast enhanced lesions measuring >2.0 cm exhibited a necrotic core. To the best of our knowledge, the present study revealed for the first time solid enhanced nodular lesions as the earliest MRI abnormalities of radiation encephalopathy following NPC radiotherapy.

Project description:BackgroundIt is common for patients diagnosed with medial temporal lobe epilepsy (TLE) to have extrahippocampal damage. However, it is unclear whether microstructural extrahippocampal abnormalities are consistent enough to enable classification using diffusion MRI imaging. Therefore, we implemented a support vector machine (SVM)-based method to predict TLE from three different imaging modalities: mean kurtosis (MK), mean diffusivity (MD), and fractional anisotropy (FA). While MD and FA can be calculated from traditional diffusion tensor imaging (DTI), MK requires diffusion kurtosis imaging (DKI).MethodsThirty-two TLE patients and 36 healthy controls underwent DKI imaging. To measure predictive capability, a fivefold cross-validation (CV) was repeated for 1000 iterations. An ensemble of SVM models, each with a different regularization value, was trained with the subject images in the training set, and had performance assessed on the test set. The different regularization values were determined using a Bayesian-based method.ResultsMean kurtosis achieved higher accuracy than both FA and MD on every iteration, and had far superior average accuracy: 0.82 (MK), 0.68 (FA), and 0.51 (MD). Finally, the MK voxels with the highest coefficients in the predictive models were distributed within the inferior medial aspect of the temporal lobes.ConclusionThese results corroborate our earlier publications which indicated that DKI shows more promise in identifying TLE-associated pathological features than DTI. Also, the locations of the contributory MK voxels were in areas with high fiber crossing and complex fiber anatomy. These traits result in non-Gaussian water diffusion, and hence render DTI less likely to detect abnormalities. If the location of consistent microstructural abnormalities can be better understood, then it may be possible in the future to identify the various phenotypes of TLE. This is important since treatment outcome varies dependent on type of TLE.

Dataset Information

Predictive accuracy of machine learning for radiation-induced temporal lobe injury in nasopharyngeal carcinoma patients: a systematic review and meta-analysis.

Background

Methods

Results

Conclusions

Publications

Predictive accuracy of machine learning for radiation-induced temporal lobe injury in nasopharyngeal carcinoma patients: a systematic review and meta-analysis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets