Repositorio Dspace

Assessing robustness and generalization of a deep neural network for brain MS lesion segmentation on real-world data

Mostrar el registro sencillo del ítem

dc.contributor.author Chaves, Hernán
dc.contributor.author Serra, María Mercedes
dc.contributor.author Shalom, Diego E.
dc.contributor.author Ananía, Pilar
dc.contributor.author Rueda, Fernanda
dc.contributor.author Osa Sanz, Emilia
dc.contributor.author Stefanoff, Nadia Ivanna
dc.contributor.author Rodríguez Murúa, Sofía
dc.contributor.author Costa, Martín Elías
dc.contributor.author Kitamura, Felipe C.
dc.contributor.author Yañez, Paulina
dc.contributor.author Cejas, Claudia Patricia
dc.contributor.author Correale, Jorge
dc.contributor.author Ferrante, Enzo
dc.contributor.author Fernández Slezak, Diego
dc.contributor.author Farez, Mauricio Franco
dc.date.accessioned 2024-02-23T15:28:37Z
dc.date.available 2024-02-23T15:28:37Z
dc.date.issued 2023-08-31
dc.identifier.citation Chaves H, Serra MM, Shalom DE, Ananía P, Rueda F, Osa Sanz E, Stefanoff NI, Rodríguez Murúa S, Costa ME, Kitamura FC, Yañez P, Cejas C, Correale J, Ferrante E, Fernández Slezak D, Farez MF. Assessing robustness and generalization of a deep neural network for brain MS lesion segmentation on real-world data. Eur Radiol. 2024 Mar;34(3):2024-2035. doi: 10.1007/s00330-023-10093-5 es_ES
dc.identifier.uri https://doi.org/10.1007/s00330-023-10093-5
dc.identifier.uri https://repositorio.fleni.org.ar/xmlui/handle/123456789/1012
dc.description.abstract Objectives: Evaluate the performance of a deep learning (DL)-based model for multiple sclerosis (MS) lesion segmentation and compare it to other DL and non-DL algorithms. Methods: This ambispective, multicenter study assessed the performance of a DL-based model for MS lesion segmentation and compared it to alternative DL- and non-DL-based methods. Models were tested on internal (n = 20) and external (n = 18) datasets from Latin America, and on an external dataset from Europe (n = 49). We also examined robustness by rescanning six patients (n = 6) from our MS clinical cohort. Moreover, we studied inter-human annotator agreement and discussed our findings in light of these results. Performance and robustness were assessed using intraclass correlation coefficient (ICC), Dice coefficient (DC), and coefficient of variation (CV). Results: Inter-human ICC ranged from 0.89 to 0.95, while spatial agreement among annotators showed a median DC of 0.63. Using expert manual segmentations as ground truth, our DL model achieved a median DC of 0.73 on the internal, 0.66 on the external, and 0.70 on the challenge datasets. The performance of our DL model exceeded that of the alternative algorithms on all datasets. In the robustness experiment, our DL model also achieved higher DC (ranging from 0.82 to 0.90) and lower CV (ranging from 0.7 to 7.9%) when compared to the alternative methods. Conclusion: Our DL-based model outperformed alternative methods for brain MS lesion segmentation. The model also proved to generalize well on unseen data and has a robust performance and low processing times both on real-world and challenge-based data. Clinical relevance statement: Our DL-based model demonstrated superior performance in accurately segmenting brain MS lesions compared to alternative methods, indicating its potential for clinical application with improved accuracy, robustness, and efficiency. Key points: • Automated lesion load quantification in MS patients is valuable; however, more accurate methods are still necessary. • A novel deep learning model outperformed alternative MS lesion segmentation methods on multisite datasets. • Deep learning models are particularly suitable for MS lesion segmentation in clinical scenarios. es_ES
dc.language.iso eng es_ES
dc.publisher Springer es_ES
dc.subject Algorithms es_ES
dc.subject Algoritmos es_ES
dc.subject Brain es_ES
dc.subject Encéfalo es_ES
dc.subject Magnetic Resonance Imaging es_ES
dc.subject Imagen por Resonancia Magnética es_ES
dc.subject Multiple Sclerosis es_ES
dc.subject Esclerosis Múltiple es_ES
dc.subject Neural Networks, Computer es_ES
dc.subject Redes Neurales de la Computación es_ES
dc.title Assessing robustness and generalization of a deep neural network for brain MS lesion segmentation on real-world data es_ES
dc.type info:eu-repo/semantics/article es_ES
dc.description.fil Fil: Chaves, Hernán. Fleni. Departamento de Diagnóstico por Imágenes; Argentina.
dc.description.fil Fil: Serra, María Mercedes. Fleni. Departamento de Diagnóstico por Imágenes; Argentina.
dc.description.fil Fil: Osa Sanz, Emilia. Fleni. Departamento de Diagnóstico por Imágenes; Argentina.
dc.description.fil Fil: Stefanoff, Nadia Ivanna. Fleni. Departamento de Diagnóstico por Imágenes; Argentina.
dc.description.fil Fil: Rodríguez Murúa, Sofía. Fleni. Centro para la Investigación de Enfermedades Neuroinmunológicas; Argentina.
dc.description.fil Fil: Yañez, Paulina. Fleni. Departamento de Diagnóstico por Imágenes; Argentina.
dc.description.fil Fil: Cejas, Claudia Patricia. Fleni. Departamento de Diagnóstico por Imágenes; Argentina.
dc.description.fil Fil: Correale, Jorge. Fleni. Departamento de Neurología. Servicio de Neuroinmunología y Enfermedades Desmielinizantes; Argentina.
dc.description.fil Fil: Fernández Slezak, Diego. Fleni. Centro para la Investigación de Enfermedades Neuroinmunológicas; Argentina.
dc.description.fil Fil: Farez, Mauricio Franco. Fleni. Centro para la Investigación de Enfermedades Neuroinmunológicas; Argentina.
dc.relation.ispartofVOLUME 34
dc.relation.ispartofNUMBER 3
dc.relation.ispartofPAGINATION 2024-2035
dc.relation.ispartofCOUNTRY Alemania
dc.relation.ispartofCITY Berlín
dc.relation.ispartofTITLE European radiology
dc.relation.ispartofISSN 1432-1084
dc.type.snrd info:ar-repo/semantics/artículo es_ES


Ficheros en el ítem

Este ítem aparece en la(s) siguiente(s) colección(ones)

Mostrar el registro sencillo del ítem

Buscar en DSpace


Listar

Mi cuenta

Estadísticas