Abstract
Purpose
Toward an efficient clinical management of hepatocellular carcinoma (HCC), we propose a classification framework dedicated to tumor necrosis rate estimation from dynamic contrast-enhanced CT scans. Based on machine learning, it requires weak interaction efforts to segment healthy, active and necrotic liver tissues.
Methods
Our contributions are two-fold. First, we apply random forest (RF) on supervoxels using multi-phase supervoxel-based features that discriminate tissues based on their dynamic in response to contrast agent injection. Second, we extend this technique in a hierarchical multi-scale fashion to deal with multiple spatial extents and appearance heterogeneity. It translates in an adaptive data sampling scheme combining RF and hierarchical multi-scale tree resulting from recursive supervoxel decomposition. By concatenating multi-phase features across the hierarchical multi-scale tree to describe leaf supervoxels, we enable RF to automatically infer the most informative scales without defining any explicit rules on how to combine them.
Results
Assessment on clinical data confirms the benefits of multi-phase information embedded in a multi-scale supervoxel representation for HCC tumor segmentation.
Conclusion
Dedicated but not limited only to HCC management, both contributions reach further steps toward more accurate multi-label tissue classification.
Similar content being viewed by others
Notes
scikit-image implementation, http://scikit-image.org.
scikit-learn implementation, http://scikit-learn.org/.
References
Achanta R, Shaji A, Smith K, Lucchi A, Fua P, Susstrunk S (2012) SLIC superpixels compared to state-of-the-art superpixel methods. IEEE Trans Pattern Anal Mach Intell 34:2274–2282
Akselrod-Ballin A, Galun M, Gomori JM, Filippi M, Valsasina P, Basri R, Brandt A (2009) Automatic segmentation and classification of multiple sclerosis in multichannel mri. IEEE Trans Biomed Eng 56(10):2461–2469
Avants BB, Epstein CL, Grossman M, Gee JC (2008) Symmetric diffeomorphic image registration with cross-correlation: evaluating automated labeling of elderly and neurodegenerative brain. Med Image Anal 12(1):26–41
Bauer S, Nolte LP, Reyes M (2011) Fully automatic segmentation of brain tumor images using support vector machine classification in combination with hierarchical conditional random field regularization. In: Medical image computing and computer-assisted intervention, pp 354–361
Beg MF, Miller MI, Trouvé A, Younes L (2005) Computing large deformation metric mappings via geodesic flows of diffeomorphisms. Int J Comput Vis 61:139–157
Breiman L (2001) Random forests. Mach Learn 45(1):5–32
Cantu M, Piardi T, Sommacale D, Ellero B, Woehl-Jaegle ML, Audet M, Ntourakis D, Wolf P, Pessaux P (2013) Pathologic response to non-surgical locoregional therapies as potential selection criteria for liver transplantation for hepatocellular carcinoma. Med Sci Monit Basic Res 18:273–284
Conze PH, Rousseau F, Noblet V, Heitz F, Memeo R, Pessaux P (2015) Semi-automatic liver tumor segmentation in dynamic contrast-enhanced CT scans using random forests and supervoxels. Mach Learn Med Imaging 9352:212–219
Conze PH, Noblet V, Rousseau F, Heitz F, Memeo R, Pessaux P (2016) Random forests on hierarchical multi-scale supervoxels for liver tumor segmentation in dynamic contrast-enhanced CT scans. In: IEEE international symposium on biomedical imaging, pp 416–419
Criminisi A, Shotton J, Konukoglu E (2012) Decision forests: a unified framework for classification, regression, density estimation, manifold learning and semi-supervised learning. Found Trends Comput Graph Vis 7(2–3):81–227
Fang R, Zabih R, Raj A, Chen T (2012) Segmentation of liver tumor using efficient global optimal tree metrics graph cuts. In: Abdominal imaging. Computational and clinical applications, pp 51–59
Forner A, Llovet JM, Bruix J (2012) Hepatocellular carcinoma. Lancet 379:1245–1255
Geremia E, Clatz O, Menze BH, Konukoglu E, Criminisi A, Ayache N (2011) Spatial decision forests for MS lesion segmentation in multi-channel magnetic resonance images. NeuroImage 57(2):378–390
Geremia E, Menze BH, Ayache N (2013) Spatially adaptive random forests. In: IEEE international symposium on biomedical imaging, pp 1344–1347
Ho MH, Yu CY, Chung KP, Chen TW, Chu HC, Lin CK, Hsieh CB (2011) Locoregional therapy-induced tumor necrosis as a predictor of recurrence after liver transplant in patients with HCC. Ann Surg Oncol 18(13):3632–3639
Irving B, Cifor A, Papie BW, Franklin J, Anderson EM, Brady M, Schnabel JA (2014) Automated colorectal tumour segmentation in DCE-MRI using supervoxel neighbourhood contrast characteristics. Med Image Comput Comput Assist Interv 8673:609–616
Irving B, Franklin JM, Papie BW, Anderson EM, Sharma RA, Gleeson FV, Brady M, Schnabel JA (2016) Pieces-of-parts for supervoxel segmentation with global context: application to DCE-MRI tumour delineation. Med Image Anal 32:69–83
Lee J, Cai W, Singh A, Yoshida H (2010) Estimation of necrosis volumes in focal liver lesions based on multi-phase hepatic CT images. In: Virtual colonoscopy & abdominal imaging. Computational challenges & clinical opportunities, pp 60–67
Machairas V, Baldeweck T, Walter T, Decencière E (2016) New general features based on superpixels for image segmentation learning. In: IEEE international symposium on biomedical imaging, pp 1409–1413
Memeo R, de Blasi V, Cherkaoui Z, Dehlawi A, de Angelis N, Piardi T, Sommacale D, Marescaux J, Mutter D, Pessaux P (2016) New approaches in locoregional therapies for hepatocellular carcinoma. J Gastrointest Cancer 47:239–246
Montillo A, Shotton J, Winn J, Iglesias JE, Metaxas D, Criminisi A (2011) Entangled decision forests and their application for semantic segmentation of CT images. In: Information processing in medical imaging, pp 184–196
Peter L, Pauly O, Chatelain P, Mateus D, Navab N (2015) Scale-adaptive forest training via an efficient feature sampling scheme. In: Medical image computing and computer-assisted intervention, pp 637–644
Popovic A, de la Fuente M, Engelhardt M, Radermacher K (2007) Statistical validation metric for accuracy assessment in medical image segmentation. Int J Comput Assist Radiol Surg 2(3–4):169–181
Raj A, Juluru K (2009) Visualization and segmentation of liver tumors using dynamic contrast MRI. In: IEEE conference of engineering in medicine and biology, pp 6985–6989
Ronot M, Vilgrain V (2014) Hepatocellular carcinoma: diagnostic criteria by imaging techniques. Best Pract Res Clin Gastro-enterol 28(5):795–812
Ronot M, Bouattour M, Wassermann J, Bruno O, Dreyer C, Larroque B, Castera L, Vilgrain V, Belghiti J, Raymond E, Faivre S (2014) Alternative response criteria (Choi, EASL and mRECIST) versus RECIST1.1 in patients with advanced hepatocellular carcinoma treated with Sorafenib. Oncologist 19:394–402
Shim JH, Kim KM, Lee YJ, Ko GY, Yoon HK, Sung KB, Park KM, Lee SG, Lim YS, Lee HC, Chung YH, Lee YS, Suh DJ (2010) Complete necrosis after transarterial chemoembolization could predict prolonged survival in patients with recurrent intrahepatic HCC after curative resection. Ann Surg Oncol 17(3):869–877
Shimizu A, Narihira T, Furukawa D, Kobatake H, Nawano S, Shinozaki K (2008) Ensemble segmentation using Adaboost with application to liver lesion extraction from a CT volume. In: Workshop on 3D segmentation in the clinic
Tu Z, Bai X (2010) Auto-context and its application to high-level vision tasks and 3D brain image segmentation. IEEE Trans Pattern Anal Mach Intell 32(10):1744–1757
Warfield SK, Zou KH, Wells WM (2004) Simultaneous truth and performance level estimation (STAPLE): an algorithm for the validation of image segmentation. IEEE Trans Med Imaging 23(7):903–921
Wu W, Chen AY, Zhao L, Corso JJ (2014) Brain tumor detection and segmentation in a CRF framework with pixel-pairwise affinity and superpixel-level features. Int J Comput Assist Radiol Surg 9(2):241–253
Yi Z, Criminisi A, Shotton J, Blake A (2009) Discriminative, semantic segmentation of brain tissue in MR images. In: Medical image computing and computer-assisted intervention, pp 558–565
Zikic D, Glocker B, Konukoglu E, Criminisi A, Demiralp C, Shotton J, Thomas O, Das T, Jena R, Price S (2012) Decision forests for tissue-specific segmentation of high-grade gliomas in multi-channel MR. In: Medical image computing and computer-assisted intervention, pp 369–376
Acknowledgments
This work received the financial support from Fondation Arc, http://www.fondation-arc.org. Liver segmentation masks are provided by Visible Patient, www.visiblepatient.com.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
Patrick Pessaux is speaker honorarium from Integra. Pierre-Henri Conze, Vincent Noblet, François Rousseau, Fabrice Heitz, Vito de Blasi and Riccardo Memeo declare that they have no conflict of interest.
Ethical standard
For this type of study, formal consent is not required.
Informed consent
Informed consent was obtained from all individual participants included in the study.
Rights and permissions
About this article
Cite this article
Conze, PH., Noblet, V., Rousseau, F. et al. Scale-adaptive supervoxel-based random forests for liver tumor segmentation in dynamic contrast-enhanced CT scans. Int J CARS 12, 223–233 (2017). https://doi.org/10.1007/s11548-016-1493-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11548-016-1493-1