Dataset Information

Deep learning radiomics-based prediction of distant metastasis in patients with locally advanced rectal cancer after neoadjuvant chemoradiotherapy: A multicentre study.

ABSTRACT:

Background

Accurate predictions of distant metastasis (DM) in locally advanced rectal cancer (LARC) patients receiving neoadjuvant chemoradiotherapy (nCRT) are helpful in developing appropriate treatment plans. This study aimed to perform DM prediction through deep learning radiomics.

Methods

We retrospectively sampled 235 patients receiving nCRT with the minimum 36 months' postoperative follow-up from three hospitals. Through transfer learning, a deep learning radiomic signature (DLRS) based on multiparametric magnetic resonance imaging (MRI) was constructed. A nomogram was established integrating deep MRI information and clinicopathologic factors for better prediction. Harrell's concordance index (C-index) and time-dependent receiver operating characteristic (ROC) were used as performance metrics. Furthermore, the risk of DM in patients with different response to nCRT was evaluated with the nomogram.

Findings

DLRS performed well in DM prediction, with a C-index of 0·747 and an area under curve (AUC) at three years of 0·894 in the validation cohort. The performance of nomogram was better, with a C-index of 0·775. In addition, the nomogram could stratify patients with different responses to nCRT into high- and low-risk groups of DM (P < 0·05).

Interpretation

MRI-based deep learning radiomics had potential in predicting the DM of LARC patients receiving nCRT and could help evaluate the risk of DM in patients who have different responses to nCRT.

Funding

The funding bodies that contributed to this study are listed in the Acknowledgements section.

SUBMITTER: Liu X

PROVIDER: S-EPMC8237293 | biostudies-literature | 2021 Jul

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep learning radiomics-based prediction of distant metastasis in patients with locally advanced rectal cancer after neoadjuvant chemoradiotherapy: A multicentre study.

Liu Xiangyu X Zhang Dafu D Liu Zhenyu Z Li Zhenhui Z Xie Peiyi P Sun Kai K Wei Wei W Dai Weixing W Tang Zhenchao Z Ding Yingying Y Cai Guoxiang G Tong Tong T Meng Xiaochun X Tian Jie J

EBioMedicine 20210620

<h4>Background</h4>Accurate predictions of distant metastasis (DM) in locally advanced rectal cancer (LARC) patients receiving neoadjuvant chemoradiotherapy (nCRT) are helpful in developing appropriate treatment plans. This study aimed to perform DM prediction through deep learning radiomics.<h4>Methods</h4>We retrospectively sampled 235 patients receiving nCRT with the minimum 36 months' postoperative follow-up from three hospitals. Through transfer learning, a deep learning radiomic signature ...[more]

PMID: 34157487

Similar Datasets

Project description:Highlights • Pre-treatment MRI images were used to construct a prediction model for pathologic T downstaging in locally advanced rectal cancer.• The prediction model showed good performance in training cohort (AUC 0.842), internal testing cohort (AUC 0.809) and prospective cohort (AUC 0.727).• High-probability group (score > 81.82) had potential benefits from sufficient consolidation chemotherapy. Background and purpose Predicting tumour response would be useful for selecting patients with locally advanced rectal cancer (LARC) for organ preservation strategies. We aimed to develop and validate a prediction model for T downstaging (ypT0-2) in LARC patients after neoadjuvant chemoradiotherapy and to identify those who may benefit from consolidation chemotherapy. Materials and methods cT3-4 LARC patients at three tertiary medical centers from January 2012 to January 2019 were retrospectively included, while a prospective cohort was recruited from June 2021 to March 2022. Eight filter (principal component analysis, least absolute shrinkage and selection operator, partial least-squares discriminant analysis, random forest)-classifier (support vector machine, logistic regression) models were established to select radiomic features. A nomogram combining radiomics and significant clinical features was developed and validated by calibration curve and decision curve analysis. Interaction test was conducted to investigate the consolidation chemotherapy benefits. Results A total of 634 patients were included (426 in training cohort, 174 in testing cohort and 34 in prospective cohort). A radiomic prediction model using partial least-squares discriminant analysis and a support vector machine showed the best performance (AUC: 0.832 [training]; 0.763 [testing]). A nomogram combining radiomics and clinical features showed significantly better prognostic performance (AUC: 0.842 [training]; 0.809 [testing]) than the radiomic model. The model was also tested in the prospective cohort with AUC 0.727. High-probability group (score > 81.82) may have potential benefits from ≥ 4 cycles consolidation chemotherapy (OR: 4.173, 95 % CI: 0.953–18.276, p = 0.058, pinteraction = 0.021). Conclusion We identified and validated a model based on multicenter pre-treatment radiomics to predict ypT0-2 in cT3-4 LARC patients, which may facilitate individualised treatment decision-making for organ-preservation strategies and consolidation chemotherapy.

Project description:PurposeRadiomics provides quantitative tissue heterogeneity profiling and is an exciting approach to developing imaging biomarkers in the context of precision medicine. Normal-appearing parenchymal tissues surrounding primary tumors can harbor microscopic disease that leads to increased risk of distant metastasis (DM). This study assesses whether computed-tomography (CT) imaging features of such peritumoral tissues can predict DM in locally advanced non-small cell lung cancer (NSCLC).Material and methods200 NSCLC patients of histological adenocarcinoma were included in this study. The investigated lung tissues were tumor rim, defined to be 3mm of tumor and parenchymal tissue on either side of the tumor border and the exterior region extended from 3 to 9mm outside of the tumor. Fifteen stable radiomic features were extracted and evaluated from each of these regions on pre-treatment CT images. For comparison, features from expert-delineated tumor contours were similarly prepared. The patient cohort was separated into training and validation datasets for prognostic power evaluation. Both univariable and multivariable analyses were performed for each region using concordance index (CI).ResultsUnivariable analysis reveals that six out of fifteen tumor rim features were significantly prognostic of DM (p-value < 0.05), as were ten features from the visible tumor, and only one of the exterior features was. Multivariablely, a rim radiomic signature achieved the highest prognostic performance in the independent validation sub-cohort (CI = 0.64, p-value = 2.4×10-5) significantly over a multivariable clinical model (CI = 0.53), a visible tumor radiomics model (CI = 0.59), or an exterior tissue model (CI = 0.55). Furthermore, patient stratification by the combined rim signature and clinical predictor led to a significant improvement on the clinical predictor alone and also outperformed stratification using the combined tumor signature and clinical predictor.ConclusionsWe identified peritumoral rim radiomic features significantly associated with DM. This study demonstrated that peritumoral imaging characteristics may provide additional valuable information over the visible tumor features for patient risk stratification due to cancer metastasis.

Dataset Information

Deep learning radiomics-based prediction of distant metastasis in patients with locally advanced rectal cancer after neoadjuvant chemoradiotherapy: A multicentre study.

Background

Methods

Findings

Interpretation

Funding

Publications

Deep learning radiomics-based prediction of distant metastasis in patients with locally advanced rectal cancer after neoadjuvant chemoradiotherapy: A multicentre study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets