Ilaria Bonavita1, Xavier Rafael-Palou2, Mario Ceresa3, Gemma Piella4, Vicent Ribas5, Miguel A González Ballester6. 1. Eurecat, Centre Tecnològic de Catalunya, eHealth Unit, Barcelona, Spain. Electronic address: ilaria.bonavita@eurecat.org. 2. Eurecat, Centre Tecnològic de Catalunya, eHealth Unit, Barcelona, Spain; BCN Medtech, Dept. of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain. Electronic address: xavier.rafael@eurecat.org. 3. BCN Medtech, Dept. of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain. Electronic address: mario.ceresa@upf.edu. 4. BCN Medtech, Dept. of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain. Electronic address: gemma.piella@upf.edu. 5. Eurecat, Centre Tecnològic de Catalunya, eHealth Unit, Barcelona, Spain. Electronic address: vicent.ribas@eurecat.org. 6. BCN Medtech, Dept. of Information and Communication Technologies, Universitat Pompeu Fabra, Barcelona, Spain; ICREA, Barcelona, Spain. Electronic address: ma.gonzalez@upf.edu.
Abstract
BACKGROUND AND OBJECTIVE: The early identification of malignant pulmonary nodules is critical for a better lung cancer prognosis and less invasive chemo or radio therapies. Nodule malignancy assessment done by radiologists is extremely useful for planning a preventive intervention but is, unfortunately, a complex, time-consuming and error-prone task. This explains the lack of large datasets containing radiologists malignancy characterization of nodules; METHODS: In this article, we propose to assess nodule malignancy through 3D convolutional neural networks and to integrate it in an automated end-to-end existing pipeline of lung cancer detection. For training and testing purposes we used independent subsets of the LIDC dataset; RESULTS: Adding the probabilities of nodules malignity in a baseline lung cancer pipeline improved its F1-weighted score by 14.7%, whereas integrating the malignancy model itself using transfer learning outperformed the baseline prediction by 11.8% of F1-weighted score; CONCLUSIONS: Despite the limited size of the lung cancer datasets, integrating predictive models of nodule malignancy improves prediction of lung cancer.
BACKGROUND AND OBJECTIVE: The early identification of malignant pulmonary nodules is critical for a better lung cancer prognosis and less invasive chemo or radio therapies. Nodule malignancy assessment done by radiologists is extremely useful for planning a preventive intervention but is, unfortunately, a complex, time-consuming and error-prone task. This explains the lack of large datasets containing radiologists malignancy characterization of nodules; METHODS: In this article, we propose to assess nodule malignancy through 3D convolutional neural networks and to integrate it in an automated end-to-end existing pipeline of lung cancer detection. For training and testing purposes we used independent subsets of the LIDC dataset; RESULTS: Adding the probabilities of nodules malignity in a baseline lung cancer pipeline improved its F1-weighted score by 14.7%, whereas integrating the malignancy model itself using transfer learning outperformed the baseline prediction by 11.8% of F1-weighted score; CONCLUSIONS: Despite the limited size of the lung cancer datasets, integrating predictive models of nodule malignancy improves prediction of lung cancer.