Hangyu Zhang1, Xudong Zhu1, Bin Li1, Xiaomeng Dai2, Xuanwen Bao1, Qihan Fu1, Zhou Tong1, Lulu Liu1, Yi Zheng1, Peng Zhao1, Luan Ye2, Zhihong Chen2, Weijia Fang1, Lingxiang Ruan3, Xinyu Jin4. 1. The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China. 2. Institute of Information Science and Electronic Engineering, Zhejiang University, Yuquan Campus, Hangzhou, Zhejiang, China. 3. The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China. ruanlx@126.com. 4. Institute of Information Science and Electronic Engineering, Zhejiang University, Yuquan Campus, Hangzhou, Zhejiang, China. jinxy@zju.edu.cn.
Abstract
PURPOSE: The existing medical imaging tools have a detection accuracy of 97% for peritoneal metastasis(PM) bigger than 0.5 cm, but only 29% for that smaller than 0.5 cm, the early detection of PM is still a difficult problem. This study is aiming at constructing a deep convolution neural network classifier based on meta-learning to predict PM. METHOD: Peritoneal metastases are delineated on enhanced CT. The model is trained based on meta-learning, and features are extracted using multi-modal deep Convolutional Neural Network(CNN) with enhanced CT to classify PM. Besides, we evaluate the performance on the test dataset, and compare it with other PM prediction algorithm. RESULTS: The training datasets are consisted of 9574 images from 43 patients with PM and 67 patients without PM. The testing datasets are consisted of 1834 images from 21 testing patients. To increase the accuracy of the prediction, we combine the multi-modal inputs of plain scan phase, portal venous phase and arterial phase to build a meta-learning-based multi-modal PM predictor. The classifier shows an accuracy of 87.5% with Area Under Curve(AUC) of 0.877, sensitivity of 73.4%, specificity of 95.2% on the testing datasets. The performance is superior to routine PM classify based on logistic regression (AUC: 0.795), a deep learning method named ResNet3D (AUC: 0.827), and a domain generalization (DG) method named MADDG (AUC: 0.834). CONCLUSIONS: we proposed a novel training strategy based on meta-learning to improve the model's robustness to "unseen" samples. The experiments shows that our meta-learning-based multi-modal PM predicting classifier obtain more competitive results in synchronous PM prediction compared to existing algorithms and the model's improvements of generalization ability even with limited data.
PURPOSE: The existing medical imaging tools have a detection accuracy of 97% for peritoneal metastasis(PM) bigger than 0.5 cm, but only 29% for that smaller than 0.5 cm, the early detection of PM is still a difficult problem. This study is aiming at constructing a deep convolution neural network classifier based on meta-learning to predict PM. METHOD: Peritoneal metastases are delineated on enhanced CT. The model is trained based on meta-learning, and features are extracted using multi-modal deep Convolutional Neural Network(CNN) with enhanced CT to classify PM. Besides, we evaluate the performance on the test dataset, and compare it with other PM prediction algorithm. RESULTS: The training datasets are consisted of 9574 images from 43 patients with PM and 67 patients without PM. The testing datasets are consisted of 1834 images from 21 testing patients. To increase the accuracy of the prediction, we combine the multi-modal inputs of plain scan phase, portal venous phase and arterial phase to build a meta-learning-based multi-modal PM predictor. The classifier shows an accuracy of 87.5% with Area Under Curve(AUC) of 0.877, sensitivity of 73.4%, specificity of 95.2% on the testing datasets. The performance is superior to routine PM classify based on logistic regression (AUC: 0.795), a deep learning method named ResNet3D (AUC: 0.827), and a domain generalization (DG) method named MADDG (AUC: 0.834). CONCLUSIONS: we proposed a novel training strategy based on meta-learning to improve the model's robustness to "unseen" samples. The experiments shows that our meta-learning-based multi-modal PM predicting classifier obtain more competitive results in synchronous PM prediction compared to existing algorithms and the model's improvements of generalization ability even with limited data.
Authors: Jan Franko; Qian Shi; Charles D Goldman; Barbara A Pockaj; Garth D Nelson; Richard M Goldberg; Henry C Pitot; Axel Grothey; Steven R Alberts; Daniel J Sargent Journal: J Clin Oncol Date: 2011-12-12 Impact factor: 44.544
Authors: Irene Thomassen; Yvette R van Gestel; Bert van Ramshorst; Misha D Luyer; Koop Bosscha; Simon W Nienhuijs; Valery E Lemmens; Ignace H de Hingh Journal: Int J Cancer Date: 2013-08-05 Impact factor: 7.396
Authors: Valery E Lemmens; Yvonne L Klaver; Vic J Verwaal; Harm J Rutten; Jan Willem W Coebergh; Ignace H de Hingh Journal: Int J Cancer Date: 2010-10-13 Impact factor: 7.396