Mengqiu Cao1, Shiteng Suo1,2, Xiao Zhang3, Xiaoqing Wang1, Jianrong Xu1, Wei Yang4, Yan Zhou1. 1. Department of Radiology, Renji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai 200127, China. 2. Biomedical Instrument Institute, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai 200030, China. 3. Zhuhai Precision Medical Center, Zhuhai People's Hospital (Zhuhai Hospital Affiliated with Jinan University), Zhuhai 519000, China. 4. Guangdong Provincial Key Laboratory of Medical Image Processing, School of Biomedical Engineering, Southern Medical University, Guangzhou 510515, China.
Abstract
PURPOSE: Preoperative prediction of isocitrate dehydrogenase 1 (IDH1) mutation in lower-grade gliomas (LGGs) is crucial for clinical decision-making. This study aimed to examine the predictive value of a machine learning approach using qualitative and quantitative MRI features to identify the IDH1 mutation in LGGs. MATERIALS AND METHODS: A total of 102 LGG patients were allocated to training (n = 67) and validation (n = 35) cohorts and were subject to Visually Accessible Rembrandt Images (VASARI) feature extraction (23 features) from conventional multimodal MRI and radiomics feature extraction (56 features) from apparent diffusion coefficient maps. Feature selection was conducted using the maximum Relevance Minimum Redundancy method and 0.632+ bootstrap method. A machine learning model to predict IDH1 mutation was then established using a random forest classifier. The predictive performance was evaluated using receiver operating characteristic (ROC) curves. RESULTS: After feature selection, the top 5 VASARI features were enhancement quality, deep white matter invasion, tumor location, proportion of necrosis, and T1/FLAIR ratio, and the top 10 radiomics features included 3 histogram features, 3 gray-level run-length matrix features, and 3 gray-level size zone matrix features and one shape feature. Using the optimal VASARI or radiomics feature sets for IDH1 prediction, the trained model achieved an area under the ROC curve (AUC) of 0.779 ± 0.001 or 0.849 ± 0.008 on the validation cohort, respectively. The fusion model that integrated outputs of both optimal VASARI and radiomics models improved the AUC to 0.879. CONCLUSION: The proposed machine learning approach using VASARI and radiomics features can predict IDH1 mutation in LGGs.
PURPOSE: Preoperative prediction of isocitrate dehydrogenase 1 (IDH1) mutation in lower-grade gliomas (LGGs) is crucial for clinical decision-making. This study aimed to examine the predictive value of a machine learning approach using qualitative and quantitative MRI features to identify the IDH1 mutation in LGGs. MATERIALS AND METHODS: A total of 102 LGG patients were allocated to training (n = 67) and validation (n = 35) cohorts and were subject to Visually Accessible Rembrandt Images (VASARI) feature extraction (23 features) from conventional multimodal MRI and radiomics feature extraction (56 features) from apparent diffusion coefficient maps. Feature selection was conducted using the maximum Relevance Minimum Redundancy method and 0.632+ bootstrap method. A machine learning model to predict IDH1 mutation was then established using a random forest classifier. The predictive performance was evaluated using receiver operating characteristic (ROC) curves. RESULTS: After feature selection, the top 5 VASARI features were enhancement quality, deep white matter invasion, tumor location, proportion of necrosis, and T1/FLAIR ratio, and the top 10 radiomics features included 3 histogram features, 3 gray-level run-length matrix features, and 3 gray-level size zone matrix features and one shape feature. Using the optimal VASARI or radiomics feature sets for IDH1 prediction, the trained model achieved an area under the ROC curve (AUC) of 0.779 ± 0.001 or 0.849 ± 0.008 on the validation cohort, respectively. The fusion model that integrated outputs of both optimal VASARI and radiomics models improved the AUC to 0.879. CONCLUSION: The proposed machine learning approach using VASARI and radiomics features can predict IDH1 mutation in LGGs.
Authors: Albert Lai; Samir Kharbanda; Whitney B Pope; Anh Tran; Orestes E Solis; Franklin Peale; William F Forrest; Kanan Pujara; Jose A Carrillo; Ajay Pandita; Benjamin M Ellingson; Chauncey W Bowers; Robert H Soriano; Nils O Schmidt; Sankar Mohan; William H Yong; Somasekar Seshagiri; Zora Modrusan; Zhaoshi Jiang; Kenneth D Aldape; Paul S Mischel; Linda M Liau; Cameron J Escovedo; Weidong Chen; Phioanh Leia Nghiemphu; C David James; Michael D Prados; Manfred Westphal; Katrin Lamszus; Timothy Cloughesy; Heidi S Phillips Journal: J Clin Oncol Date: 2011-10-24 Impact factor: 44.544
Authors: J A Brunberg; T L Chenevert; P E McKeever; D A Ross; L R Junck; K M Muraszko; R Dauser; J G Pipe; A T Betley Journal: AJNR Am J Neuroradiol Date: 1995-02 Impact factor: 3.825
Authors: Céline De Looze; Alan Beausang; Jane Cryan; Teresa Loftus; Patrick G Buckley; Michael Farrell; Seamus Looby; Richard Reilly; Francesca Brett; Hugh Kearney Journal: J Neurooncol Date: 2018-05-16 Impact factor: 4.130
Authors: Philipp Kickingereder; Felix Sahm; Alexander Radbruch; Wolfgang Wick; Sabine Heiland; Andreas von Deimling; Martin Bendszus; Benedikt Wiestler Journal: Sci Rep Date: 2015-11-05 Impact factor: 4.379