Aim: To review how machine learning (ML) is applied to imaging biomarkers in neuro-oncology, in particular for diagnosis, prognosis, and treatment response monitoring. Materials and Methods: The PubMed and MEDLINE databases were searched for articles published before September 2018 using relevant search terms. The search strategy focused on articles applying ML to high-grade glioma biomarkers for treatment response monitoring, prognosis, and prediction. Results: Magnetic resonance imaging (MRI) is typically used throughout the patient pathway because routine structural imaging provides detailed anatomical and pathological information and advanced techniques provide additional physiological detail. Using carefully chosen image features, ML is frequently used to allow accurate classification in a variety of scenarios. Rather than being chosen by human selection, ML also enables image features to be identified by an algorithm. Much research is applied to determining molecular profiles, histological tumour grade, and prognosis using MRI images acquired at the time that patients first present with a brain tumour. Differentiating a treatment response from a post-treatment-related effect using imaging is clinically important and also an area of active study (described here in one of two Special Issue publications dedicated to the application of ML in glioma imaging). Conclusion: Although pioneering, most of the evidence is of a low level, having been obtained retrospectively and in single centres. Studies applying ML to build neuro-oncology monitoring biomarker models have yet to show an overall advantage over those using traditional statistical methods. Development and validation of ML models applied to neuro-oncology require large, well-annotated datasets, and therefore multidisciplinary and multi-centre collaborations are necessary.
Bakas, Spyridon, Reyes, Mauricio, Jakab, Andras, Bauer, Stefan, Rempfler, Markus, Crimi, Alessandro, Shinohara, Russell Takeshi, Berger, Christoph, Ha, Sung Min, Rozycki, Martin, Prastawa, Marcel, Alberts, Esther, Lipkova, Jana, Freymann, John, Kirby, Justin, Bilello, Michel, Fathallah-Shaykh, Hassan, Wiest, Roland, Kirschke, Jan, Wiestler, Benedikt, Colen, Rivka, Kotrotsou, Aikaterini, Lamontagne, Pamela, Marcus, Daniel, Milchenko, Mikhail, Nazeri, Arash, Weber, Marc-Andre, Mahajan, Abhishek, Baid, Ujjwal, Kwon, Dongjin, Agarwal, Manu, Alam, Mahbubul, Albiol, Alberto, Albiol, Antonio, Alex, Varghese, Tran, Tuan Anh, Arbel, Tal, Avery, Aaron, B., Pranjal, Banerjee, Subhashis, Batchelder, Thomas, Batmanghelich, Kayhan, Battistella, Enzo, Bendszus, Martin, Benson, Eze, Bernal, Jose, Biros, George, Cabezas, Mariano, Chandra, Siddhartha, Chang, Yi-Ju, Chazalon, Joseph, Chen, Shengcong, Chen, Wei, Chen, Jefferson, Cheng, Kun, Christoph, Meinel, Chylla, Roger, Clérigues, Albert, Costa, Anthony, Cui, Xiaomeng, Dai, Zhenzhen, Dai, Lutao, Deutsch, Eric, Ding, Changxing, Dong, Chao, Dudzik, Wojciech, Estienne, Théo, Shin, Hyung Eun, Everson, Richard, Fabrizio, Jonathan, Fang, Longwei, Feng, Xue, Fidon, Lucas, Fridman, Naomi, Fu, Huan, Fuentes, David, Gering, David G, Gao, Yaozong, Gates, Evan, Gholami, Amir, Gong, Mingming, González-Villá, Sandra, Pauloski, J. Gregory, Guan, Yuanfang, Guo, Sheng, Gupta, Sudeep, Thakur, Meenakshi H, Maier-Hein, Klaus H., Han, Woo-Sup, He, Huiguang, Hernández-Sabaté, Aura, Herrmann, Evelyn, Himthani, Naveen, Hsu, Winston, Hsu, Cheyu, Hu, Xiaojun, Hu, Xiaobin, Hu, Yan, Hu, Yifan, Hua, Rui, Huang, Teng-Yi, Huang, Weilin, Huo, Quan, HV, Vivek, Isensee, Fabian, Islam, Mobarakol, Albiol, Francisco J., Wang, Chiatse J., Jambawalikar, Sachin, Jose, V Jeya Maria, Jian, Weijian, Jin, Peter, Jungo, Alain, Nuechterlein, Nicholas K, Kao, Po-Yu, Kermi, Adel, Keutzer, Kurt, Khened, Mahendra, Kickingereder, Philipp, King, Nik, Knapp, Haley, Knecht, Urspeter, Kohli, Lisa, Kong, Deren, Kong, Xiangmao, Koppers, Simon, Kori, Avinash, Krishnamurthi, Ganapathy, Kumar, Piyush, Kushibar, Kaisar, Lachinov, Dmitrii, Lee, Joon, Lee, Chengen, Lee, Yuehchou, Lefkovits, Szidonia, Lefkovits, Laszlo, Li, Tengfei, Li, Hongwei, Li, Wenqi, Li, Hongyang, Li, Xiaochuan, Lin, Zheng-Shen, Lin, Fengming, Liu, Chang, Liu, Boqiang, Liu, Xiang, Liu, Mingyuan, Liu, Ju, Lladó, Xavier, Luo, Lin, Iftekharuddin, Khan M., Tsai, Yuhsiang M., Ma, Jun, Ma, Kai, Mackie, Thomas, Mahmoudi, Issam, Marcinkiewicz, Michal, McKinley, Richard, Mehta, Sachin, Mehta, Raghav, Meier, Raphael, Merhof, Dorit, Meyer, Craig, Mitra, Sushmita, Moiyadi, Aliasgar, Mrukwa, Grzegorz, Monteiro, Miguel A. B., Myronenko, Andriy, Carver, Eric N, Nalepa, Jakub, Ngo, Thuyen, Niu, Chen, Oermann, Eric, Oliveira, Arlindo, Oliver, Arnau, Ourselin, Sebastien, French, Andrew P., Pound, Michael P., Pridmore, Tony P., Serrano-Rubio, Juan Pablo, Paragios, Nikos, Paschke, Brad, Pei, Linmim, Peng, Suting, Pham, Bao, Piella, Gemma, Pillai, G. N., Piraud, Marie, Popli, Anmol, Prčkovska, Vesna, Puch, Santi, Puybareau, Élodie, Qiao, Xu, Suter, Yannick R, Scott, Matthew R., Rane, Swapnil, Rebsamen, Michael, Ren, Hongliang, Ren, Xuhua, Rezaei, Mina, Lorenzo, Pablo Ribalta, Rippel, Oliver, Robert, Charlotte, Choudhury, Ahana Roy, Jackson, Aaron S., Manjunath, B. S., Salem, Mostafa, Salvi, Joaquim, Sánchez, Irina, Schellingerhout, Dawid, Shboul, Zeina, Shen, Haipeng, Shen, Dinggang, Shenoy, Varun, Shi, Feng, Shu, Hai, Snyder, James, Han, Il Song, Soni, Mehul, Stawiaski, Jean, Subramanian, Shashank, Sun, Li, Sun, Roger, Sun, Jiawei, Sun, Kay, Sun, Yu, Sun, Guoxia, Sun, Shuang, Park, Moo Sung, Szilagyi, Laszlo, Talbar, Sanjay, Tao, Dacheng, Tao, Dacheng, Khadir, Mohamed Tarek, Thakur, Siddhesh, Tochon, Guillaume, Tran, Tuan, Tseng, Kuan-Lun, Turlapov, Vadim, Tustison, Nicholas, Shankar, B. Uma, Vakalopoulou, Maria, Valverde, Sergi, Vanguri, Rami, Vasiliev, Evgeny, Vercauteren, Tom, Vidyaratne, Lasitha, Vivekanandan, Ajeet, Wang, Guotai, Wang, Qian, Wang, Weichung, Wen, Ning, Wen, Xin, Weninger, Leon, Wick, Wolfgang, Wu, Shaocheng, Wu, Qiang, Xia, Yong, Xu, Yanwu, Xu, Xiaowen, Xu, Peiyuan, Yang, Tsai-Ling, Yang, Xiaoping, Yang, Hao-Yu, Yang, Junlin, Yang, Haojin, Yao, Hongdou, Young-Moxon, Brett, Yue, Xiangyu, Zhang, Songtao, Zhang, Angela, Zhang, Kun, Zhang, Xuejie, Zhang, Lichi, Zhang, Xiaoyue, Zhao, Sicheng, Zhao, Yu, Zheng, Yefeng, Zhong, Liming, Zhou, Chenhong, Zhou, Xiaobing, Zhu, Hongtu, Zong, Weiwei, Kalpathy-Cramer, Jayashree, Farahani, Keyvan, Davatzikos, Christos, van Leemput, Koen, Menze, Bjoern
Gliomas are the most common primary brain malignancies, with different degrees of aggressiveness, variable prognosis and various heterogeneous histologic sub-regions, i.e., peritumoral edematous/invaded tissue, necrotic core, active and non-enhancing core. This intrinsic heterogeneity is also portrayed in their radio-phenotype, as their sub-regions are depicted by varying intensity profiles disseminated across multi-parametric magnetic resonance imaging (mpMRI) scans, reflecting varying biological properties. Their heterogeneous shape, extent, and location are some of the factors that make these tumors difficult to resect, and in some cases inoperable. The amount of resected tumor is a factor also considered in longitudinal scans, when evaluating the apparent tumor for potential diagnosis of progression. Furthermore, there is mounting evidence that accurate segmentation of the various tumor sub-regions can offer the basis for quantitative image analysis towards prediction of patient overall survival. This study assesses the state-of-the-art machine learning (ML) methods used for brain tumor image analysis in mpMRI scans, during the last seven instances of the International Brain Tumor Segmentation (BraTS) challenge, i.e. 2012-2018. Specifically, we focus on i) evaluating segmentations of the various glioma sub-regions in pre-operative mpMRI scans, ii) assessing potential tumor progression by virtue of longitudinal growth of tumor sub-regions, beyond use of the RECIST criteria, and iii) predicting the overall survival from pre-operative mpMRI scans of patients that undergone gross total resection. Finally, we investigate the challenge of identifying the best ML algorithms for each of these tasks, considering that apart from being diverse on each instance of the challenge, the multi-institutional mpMRI BraTS dataset has also been a continuously evolving/growing dataset.
Background and Purpose: Biopsy is the main determinants of glioma clinical management, but require invasive sampling that fail to detect relevant features because of tumor heterogeneity. The purpose of this study was to evaluate the accuracy of a voxel-wise, multiparametric MRI radiomic method to predict features and develop a minimally invasive method to objectively assess neoplasms. Methods: Multiparametric MRI were registered to T1-weighted gadolinium contrast-enhanced data using a 12 degree-of-freedom affine model. The retrospectively collected MRI data included T1-weighted, T1-weighted gadolinium contrast-enhanced, T2-weighted, fluid attenuated inversion recovery, and multi-b-value diffusion-weighted acquired at 1.5T or 3.0T. Clinical experts provided voxel-wise annotations for five disease states on a subset of patients to establish a training feature vector of 611,930 observations. Then, a k-nearest-neighbor (k-NN) classifier was trained using a 25% hold-out design. The trained k-NN model was applied to 13,018,171 observations from seventeen histologically confirmed glioma patients. Linear regression tested overall survival (OS) relationship to predicted disease compositions (PDC) and diagnostic age (alpha = 0.05). Canonical discriminant analysis tested if PDC and diagnostic age could differentiate clinical, genetic, and microscopic factors (alpha = 0.05). Results: The model predicted voxel annotation class with a Dice similarity coefficient of 94.34% +/- 2.98. Linear combinations of PDCs and diagnostic age predicted OS (p = 0.008), grade (p = 0.014), and endothelia proliferation (p = 0.003); but fell short predicting gene mutations for TP53BP1 and IDH1. Conclusions: This voxel-wise, multi-parametric MRI radiomic strategy holds potential as a non-invasive decision-making aid for clinicians managing patients with glioma.
In this paper we present a method for simultaneously segmenting brain tumors and an extensive set of organs-at-risk for radiation therapy planning of glioblastomas. The method combines a contrast-adaptive generative model for whole-brain segmentation with a new spatial regularization model of tumor shape using convolutional restricted Boltzmann machines. We demonstrate experimentally that the method is able to adapt to image acquisitions that differ substantially from any available training data, ensuring its applicability across treatment sites; that its tumor segmentation accuracy is comparable to that of the current state of the art; and that it captures most organs-at-risk sufficiently well for radiation therapy planning purposes. The proposed method may be a valuable step towards automating the delineation of brain tumors and organs-at-risk in glioblastoma patients undergoing radiation therapy.
To improve the robustness of deep learning–based glioblastoma segmentation in a clinical setting with sparsified datasets. In this retrospective study, preoperative T1-weighted, T2-weighted, T2-weighted fluid-attenuated inversion recovery, and postcontrast T1-weighted MRI from 117 patients (median age, 64 years; interquartile range [IQR], 55–73 years; 76 men) included within the Multimodal Brain Tumor Image Segmentation (BraTS) dataset plus a clinical dataset (2012–2013) with similar imaging modalities of 634 patients (median age, 59 years; IQR, 49–69 years; 382 men) with glioblastoma from six hospitals were used. Expert tumor delineations on the postcontrast images were available, but for various clinical datasets, one or more sequences were missing. The convolutional neural network, DeepMedic, was trained on combinations of complete and incomplete data with and without site-specific data. Sparsified training was introduced, which randomly simulated missing sequences during training. The effects of sparsified training and center-specific training were tested using Wilcoxon signed rank tests for paired measurements. A model trained exclusively on BraTS data reached a median Dice score of 0.81 for segmentation on BraTS test data but only 0.49 on the clinical data.