Contrast phase recognition in liver computer tomography using deep learning.

Autor: Rocha BA; InRad, Institute of Radiology, University of São Paulo, School of Medicine, Rua Dr. Ovídio Pires de Campos, 75 Cerqueira César, São Paulo, SP, 05403-010, Brazil. bruno@machiron.com.br.; Machiron Ltd., Rua Capote Valente, 671, São Paulo, 05409-002, Brazil. bruno@machiron.com.br., Ferreira LC; InRad, Institute of Radiology, University of São Paulo, School of Medicine, Rua Dr. Ovídio Pires de Campos, 75 Cerqueira César, São Paulo, SP, 05403-010, Brazil., Vianna LGR; Machiron Ltd., Rua Capote Valente, 671, São Paulo, 05409-002, Brazil., Ferreira LGG; Machiron Ltd., Rua Capote Valente, 671, São Paulo, 05409-002, Brazil., Ciconelle ACM; Machiron Ltd., Rua Capote Valente, 671, São Paulo, 05409-002, Brazil., Da Silva Noronha A; Machiron Ltd., Rua Capote Valente, 671, São Paulo, 05409-002, Brazil., Cortez Filho JM; Department of Gastroenterology, University of São Paulo, School of Medicine (FMUSP), Hospital das Clínicas (HCFMUSP), Rua Dr. Enéas Carvalho de Aguiar, 225, São Paulo, SP, 05403-000, Brazil., Nogueira LSL; Department of Gastroenterology, University of São Paulo, School of Medicine (FMUSP), Hospital das Clínicas (HCFMUSP), Rua Dr. Enéas Carvalho de Aguiar, 225, São Paulo, SP, 05403-000, Brazil., Leite JMRS; Machiron Ltd., Rua Capote Valente, 671, São Paulo, 05409-002, Brazil., da Silva Filho MRM; InRad, Institute of Radiology, University of São Paulo, School of Medicine, Rua Dr. Ovídio Pires de Campos, 75 Cerqueira César, São Paulo, SP, 05403-010, Brazil., da Costa Leite C; InRad, Institute of Radiology, University of São Paulo, School of Medicine, Rua Dr. Ovídio Pires de Campos, 75 Cerqueira César, São Paulo, SP, 05403-010, Brazil., de Maria Felix M; InRad, Institute of Radiology, University of São Paulo, School of Medicine, Rua Dr. Ovídio Pires de Campos, 75 Cerqueira César, São Paulo, SP, 05403-010, Brazil., Gutierrez MA; Informatics Department, The Heart Institute, Hospital das Clínicas (HCFMUSP), University of São Paulo, School of Medicine, Rua Dr. Enéas de Carvalho Aguiar 44, São Paulo, SP, 05403-000, Brazil., Nomura CH; InRad, Institute of Radiology, University of São Paulo, School of Medicine, Rua Dr. Ovídio Pires de Campos, 75 Cerqueira César, São Paulo, SP, 05403-010, Brazil., Cerri GG; InRad, Institute of Radiology, University of São Paulo, School of Medicine, Rua Dr. Ovídio Pires de Campos, 75 Cerqueira César, São Paulo, SP, 05403-010, Brazil., Carrilho FJ; Department of Gastroenterology, University of São Paulo, School of Medicine (FMUSP), Hospital das Clínicas (HCFMUSP), Rua Dr. Enéas Carvalho de Aguiar, 225, São Paulo, SP, 05403-000, Brazil., Ono SK; Department of Gastroenterology, University of São Paulo, School of Medicine (FMUSP), Hospital das Clínicas (HCFMUSP), Rua Dr. Enéas Carvalho de Aguiar, 225, São Paulo, SP, 05403-000, Brazil.
Jazyk: angličtina
Zdroj: Scientific reports [Sci Rep] 2022 Nov 24; Vol. 12 (1), pp. 20315. Date of Electronic Publication: 2022 Nov 24.
DOI: 10.1038/s41598-022-24485-y
Abstrakt: Hepatocellular carcinoma (HCC) has become the 4th leading cause of cancer-related deaths, with high social, economical and health implications. Imaging techniques such as multiphase computed tomography (CT) have been successfully used for diagnosis of liver tumors such as HCC in a feasible and accurate way and its interpretation relies mainly on comparing the appearance of the lesions in the different contrast phases of the exam. Recently, some researchers have been dedicated to the development of tools based on machine learning (ML) algorithms, especially by deep learning techniques, to improve the diagnosis of liver lesions in imaging exams. However, the lack of standardization in the naming of the CT contrast phases in the DICOM metadata is a problem for real-life deployment of machine learning tools. Therefore, it is important to correctly identify the exam phase based only on the image and not on the exam metadata, which is unreliable. Motivated by this problem, we successfully created an annotation platform and implemented a convolutional neural network (CNN) to automatically identify the CT scan phases in the HCFMUSP database in the city of São Paulo, Brazil. We improved this algorithm with hyperparameter tuning and evaluated it with cross validation methods. Comparing its predictions with the radiologists annotation, it achieved an accuracy of 94.6%, 98% and 100% in the testing dataset for the slice, volume and exam evaluation, respectively.
(© 2022. The Author(s).)
Databáze: MEDLINE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje