Publication date: Available online 1 February 2018
Source:Academic Radiology
Author(s): Aimilia Gastounioti, Andrew Oustimov, Meng-Kang Hsieh, Lauren Pantalone, Emily F. Conant, Despina Kontos
Rationale and ObjectivesWe evaluate utilizing convolutional neural networks (CNNs) to optimally fuse parenchymal complexity measurements generated by texture analysis into discriminative meta-features relevant for breast cancer risk prediction.Materials and MethodsWith Institutional Review Board approval and Health Insurance Portability and Accountability Act compliance, we retrospectively analyzed "For Processing" contralateral digital mammograms (GE Healthcare 2000D/DS) from 106 women with unilateral invasive breast cancer and 318 age-matched controls. We coupled established texture features (histogram, co-occurrence, run-length, structural), extracted using a previously validated lattice-based strategy, with a multichannel CNN into a hybrid framework in which a multitude of texture feature maps are reduced to meta-features predicting the case or control status. We evaluated the framework in a randomized split-sample setting, using the area under the curve (AUC) of the receiver operating characteristic (ROC) to assess case-control discriminatory capacity. We also compared the framework to CNNs directly fed with mammographic images, as well as to conventional texture analysis, where texture feature maps are summarized via simple statistical measures that are then used as inputs to a logistic regression model.ResultsStrong case-control discriminatory capacity was demonstrated on the basis of the meta-features generated by the hybrid framework (AUC = 0.90), outperforming both CNNs applied directly to raw image data (AUC = 0.63, P < .05) and conventional texture analysis (AUC = 0.79, P < .05).ConclusionsOur results suggest that informative interactions between patterns exist in texture feature maps derived from mammographic images, which can be extracted and summarized via a multichannel CNN architecture toward leveraging the associations of textural measurements to breast cancer risk.
from Imaging via alkiviadis.1961 on Inoreader http://ift.tt/2GFul1r
Εγγραφή σε:
Σχόλια ανάρτησης (Atom)
Δημοφιλείς αναρτήσεις
-
Publication date: Available online 4 January 2018 Source: European Journal of Radiology Author(s): Peiyao Zhang, Jing Wang, Qin Xu, Zhen...
-
Publication date: March 2017 Source: Free Radical Biology and Medicine, Volume 104 from #AlexandrosSfakianakis via Alexandros G.Sfak...
-
Dtsch med Wochenschr DOI: 10.1055/s-0043-100054 Hintergrund und Fragestellung Ein etablierter Weg, die optimale Behandlung von Tumorpatien...
-
Background Hyperthyroidism is associated with increased thrombotic risk. As contact system activation through formation of neutrophil extrac...
-
Deepak Thapa, Vanita Ahuja, Deepanshu Dhiman Indian Journal of Anaesthesia 2017 61(12):1012-1014 from #AlexandrosSfakianakis via Alexa...
-
Abstract Limited memory size is considered as a major bottleneck in data centers for intelligent urban computing. It is shown that there e...
-
ecancer is supporting #BowelCancerAwarenessMonth https://t.co/opXxCAAxzE from #AlexandrosSfakianakis via Alexandros G.Sfakianakis on Inore...
-
Linked Article: Maintz et al. Br J Dermatol 2017; 176:481–487 . from #AlexandrosSfakianakis via Alexandros G.Sfakianakis on Inoreader h...
-
Angelina Jolie’s openness about her diagnosis has undoubtedly raised public awareness of the importance of inherited cancers. Although Desai...
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου