Automated prostate multi-regional segmentation in magnetic resonance using fully convolutional neural networks

Fecha de publicación: 01/01/2023 Fecha Ahead of Print: 01/01/2023

Autores de IIS La Fe

Luis Martí Bonmatí

Autor
Ángel Alberich Bayarri

Autor

Participantes ajenos a IIS La Fe

Jimenez-Pastor, Ana
Lopez-Gonzalez, Rafael
Fos-Guarinos, Belen
Garcia-Castro, Fabio
Wittenberg, Mark
Torregrosa-Andres, Asuncion
Garcia-Fontes, Margarita
Duarte, Pablo
Gambini, Juan Pablo
Bittencourt, Leonardo Kayat
Kitamura, Felipe Campos
Venugopal, Vasantha Kumar
Mahajan, Vidur
Ros, Pablo
Soria-Olivas, Emilio

Grupos

Investigacion Biomédica en Imagen (GIBI 2-30)

Investigador Principal

Leonor Cerdá Alberich

Abstract

AbstractObjectiveAutomatic MR imaging segmentation of the prostate provides relevant clinical benefits for prostate cancer evaluation such as calculation of automated PSA density and other critical imaging biomarkers. Further, automated T2-weighted image segmentation of central-transition zone (CZ-TZ), peripheral zone (PZ), and seminal vesicle (SV) can help to evaluate clinically significant cancer following the PI-RADS v2.1 guidelines. Therefore, the main objective of this work was to develop a robust and reproducible CNN-based automatic prostate multi-regional segmentation model using an intercontinental cohort of prostate MRI.MethodsA heterogeneous database of 243 T2-weighted prostate studies from 7 countries and 10 machines of 3 different vendors, with the CZ-TZ, PZ, and SV regions manually delineated by two experienced radiologists (ground truth), was used to train (n = 123) and test (n = 120) a U-Net-based model with deep supervision using a cyclical learning rate. The performance of the model was evaluated by means of dice similarity coefficient (DSC), among others. Segmentation results with a DSC above 0.7 were considered accurate.ResultsThe proposed method obtained a DSC of 0.88 ± 0.01, 0.85 ± 0.02, 0.72 ± 0.02, and 0.72 ± 0.02 for the prostate gland, CZ-TZ, PZ, and SV respectively in the 120 studies of the test set when comparing the predicted segmentations with the ground truth. No statistically significant differences were found in the results obtained between manufacturers or continents.ConclusionProstate multi-regional T2-weighted MR images automatic segmentation can be accurately achieved by U-Net like CNN, generalizable in a highly variable clinical environment with different equipment, acquisition configurations, and population.Key Points• Deep learning techniques allows the accurate segmentation of the prostate in three different regions on MR T2w images.• Multi-centric database proved the generalization of the CNN model on different institutions across different continents.• CNN models can be used to aid on the diagnosis and follow-up of patients with prostate cancer.

Datos de la publicación

ISSN/ISSNe:: 0938-7994, 1432-1084
Tipo:: Article
Páginas:: 5087-5096
DOI:: 10.1007/s00330-023-09410-9
Enlace a otro recurso:: app.dimensions.ai
Factor de Impacto:: 1,730 SCImago ℠
Cuartil:: Q1 SCImago ℠

Documentos

No hay documentos

Métricas

Filiaciones

Jimenez-Pastor, Ana:: 1Quantitative Imaging Biomarkers in Medicine (Quibim S.L.), Aragon Avenue, 30, 13th floor, Office I-J, 46021, Valencia, Spain.
Lopez-Gonzalez, Rafael:: 2Quantitative Imaging Biomarkers in Medicine (Quibim S.L.), Aragon Avenue, 30, 13th floor, Office I-J, 46021, Valencia, Spain.

3Intelligent Data Analysis Laboratory (IDAL), ETSE, Universidad de Valencia, Valencia, Spain.
Fos-Guarinos, Belen:: 2Quantitative Imaging Biomarkers in Medicine (Quibim S.L.), Aragon Avenue, 30, 13th floor, Office I-J, 46021, Valencia, Spain.
Garcia-Castro, Fabio:: 2Quantitative Imaging Biomarkers in Medicine (Quibim S.L.), Aragon Avenue, 30, 13th floor, Office I-J, 46021, Valencia, Spain.
Wittenberg, Mark:: 2Quantitative Imaging Biomarkers in Medicine (Quibim S.L.), Aragon Avenue, 30, 13th floor, Office I-J, 46021, Valencia, Spain.
Torregrosa-Andres, Asuncion:: Instituto de Investigación. 4Department of Radiology and GIBI230 Research Group On Biomedical Imaging, Hospital Universitario Y Politécnico de La Fe and Instituto de Investigación Sanitaria La Fe, Valencia, Spain.
Marti-Bonmati, Luis:: Instituto de Investigación. 4Department of Radiology and GIBI230 Research Group On Biomedical Imaging, Hospital Universitario Y Politécnico de La Fe and Instituto de Investigación Sanitaria La Fe, Valencia, Spain.
Garcia-Fontes, Margarita:: 5Centro Uruguayo de Imagenologia Molecular (CUDIM), Montevideo, Uruguay.
Duarte, Pablo:: 5Centro Uruguayo de Imagenologia Molecular (CUDIM), Montevideo, Uruguay.
Gambini, Juan Pablo:: 5Centro Uruguayo de Imagenologia Molecular (CUDIM), Montevideo, Uruguay.
Bittencourt, Leonardo Kayat:: 6University Hospitals Cleveland Medical Center, Cleveland, USA.

7DasaInova, Dasa, Unifesp, São Paulo, Brazil.
Kitamura, Felipe Campos:: 7DasaInova, Dasa, Unifesp, São Paulo, Brazil.
Venugopal, Vasantha Kumar:: 8Mahajan Imaging, New Delhi, India.
Mahajan, Vidur:: 8Mahajan Imaging, New Delhi, India.
Ros, Pablo:: 9Stony Brook University, Stony Brook, NY, USA.
Soria-Olivas, Emilio:: 3Intelligent Data Analysis Laboratory (IDAL), ETSE, Universidad de Valencia, Valencia, Spain.
Alberich-Bayarri, Angel:: 2Quantitative Imaging Biomarkers in Medicine (Quibim S.L.), Aragon Avenue, 30, 13th floor, Office I-J, 46021, Valencia, Spain.