Vincenza Granata1, Roberta Fusco2, Federica De Muzio3, Carmen Cutolo4, Sergio Venanzio Setola1, Federica Dell'Aversana5, Francesca Grassi5, Andrea Belli6, Lucrezia Silvestro7, Alessandro Ottaiano7, Guglielmo Nasti7, Antonio Avallone7, Federica Flammia8, Vittorio Miele9,8, Fabiana Tatangelo10, Francesco Izzo6, Antonella Petrillo1. 1. Division of Radiology, Istituto Nazionale Tumori IRCCS Fondazione Pascale-IRCCS Di Napoli, Naples, Italy. 2. Medical Oncology Division, Igea SpA, Naples, Italy. r.fusco@igeamedical.com. 3. Department of Medicine and Health Sciences V. Tiberio, University of Molise, 86100, Campobasso, Italy. 4. Department of Medicine, Surgery and Dentistry, University of Salerno, 84084, Fisciano, Italy. 5. Division of Radiology, Università Degli Studi Della Campania Luigi Vanvitelli, Naples, Italy. 6. Division of Hepatobiliary Surgery, Istituto Nazionale Tumori IRCCS Fondazione Pascale-IRCCS Di Napoli, Naples, Italy. 7. Division of Abdominal Oncology, Istituto Nazionale Tumori IRCCS Fondazione Pascale, Naples, Italy. 8. Division of Radiology, Azienda Ospedaliera Universitaria Careggi, 50134, Florence, Italy. 9. Italian Society of Medical and Interventional Radiology (SIRM), SIRM Foundation, via della Signora 2, 20122, Milan, Italy. 10. Division of Pathology, Istituto Nazionale Tumori IRCCS Fondazione Pascale-IRCCS Di Napoli, 80131, Naples, Italy.
Abstract
PURPOSE: The purpose of this study is to evaluate the Radiomics and Machine Learning Analysis based on MRI in the assessment of Liver Mucinous Colorectal Metastases.Query METHODS: The cohort of patients included a training set (121 cases) and an external validation set (30 cases) with colorectal liver metastases with pathological proof and MRI study enrolled in this approved study retrospectively. About 851 radiomics features were extracted as median values by means of the PyRadiomics tool on volume on interest segmented manually by two expert radiologists. Univariate analysis, linear regression modelling and pattern recognition methods were used as statistical and classification procedures. RESULTS: The best results at univariate analysis were reached by the wavelet_LLH_glcm_JointEntropy extracted by T2W SPACE sequence with accuracy of 92%. Linear regression model increased the performance obtained respect to the univariate analysis. The best results were obtained by a linear regression model of 15 significant features extracted by the T2W SPACE sequence with accuracy of 94%, a sensitivity of 92% and a specificity of 95%. The best classifier among the tested pattern recognition approaches was k-nearest neighbours (KNN); however, KNN achieved lower precision than the best linear regression model. CONCLUSIONS: Radiomics metrics allow the mucinous subtype lesion characterization, in order to obtain a more personalized approach. We demonstrated that the best performance was obtained by T2-W extracted textural metrics.
PURPOSE: The purpose of this study is to evaluate the Radiomics and Machine Learning Analysis based on MRI in the assessment of Liver Mucinous Colorectal Metastases.Query METHODS: The cohort of patients included a training set (121 cases) and an external validation set (30 cases) with colorectal liver metastases with pathological proof and MRI study enrolled in this approved study retrospectively. About 851 radiomics features were extracted as median values by means of the PyRadiomics tool on volume on interest segmented manually by two expert radiologists. Univariate analysis, linear regression modelling and pattern recognition methods were used as statistical and classification procedures. RESULTS: The best results at univariate analysis were reached by the wavelet_LLH_glcm_JointEntropy extracted by T2W SPACE sequence with accuracy of 92%. Linear regression model increased the performance obtained respect to the univariate analysis. The best results were obtained by a linear regression model of 15 significant features extracted by the T2W SPACE sequence with accuracy of 94%, a sensitivity of 92% and a specificity of 95%. The best classifier among the tested pattern recognition approaches was k-nearest neighbours (KNN); however, KNN achieved lower precision than the best linear regression model. CONCLUSIONS: Radiomics metrics allow the mucinous subtype lesion characterization, in order to obtain a more personalized approach. We demonstrated that the best performance was obtained by T2-W extracted textural metrics.
Authors: Roberto Fornell-Perez; Valentina Vivas-Escalona; Joel Aranda-Sanchez; M Carmen Gonzalez-Dominguez; Jano Rubio-Garcia; Patricia Aleman-Flores; Alvaro Lozano-Rodriguez; Gabriela Porcel-de-Peralta; Juan Francisco Loro-Ferrer Journal: Radiol Med Date: 2020-02-04 Impact factor: 3.469
Authors: Nicolò Schicchi; Marco Fogante; Pierpaolo Palumbo; Giacomo Agliata; Paolo Esposto Pirani; Ernesto Di Cesare; Andrea Giovagnoni Journal: Radiol Med Date: 2020-09-15 Impact factor: 3.469
Authors: V Granata; R Grassi; R Fusco; F Izzo; L Brunese; P Delrio; A Avallone; B Pecori; A Petrillo Journal: Eur Rev Med Pharmacol Sci Date: 2020-12 Impact factor: 3.507
Authors: M J Gunter; S Alhomoud; M Arnold; H Brenner; J Burn; G Casey; A T Chan; A J Cross; E Giovannucci; R Hoover; R Houlston; M Jenkins; P Laurent-Puig; U Peters; D Ransohoff; E Riboli; R Sinha; Z K Stadler; P Brennan; S J Chanock Journal: Ann Oncol Date: 2019-04-01 Impact factor: 32.976