Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images
Tumor segmentation is one of the key steps in imaging processing. The goals of this study were to assess the inter-observer variability in manual segmentation of neuroblastic tumors and to analyze whether the state-of-the-art deep learning architecture nnU-Net can provide a robust solution to detect...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-07-01
|
| Series: | Cancers |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-6694/14/15/3648 |
| _version_ | 1797414438059376640 |
|---|---|
| author | Diana Veiga-Canuto Leonor Cerdà-Alberich Cinta Sangüesa Nebot Blanca Martínez de las Heras Ulrike Pötschger Michela Gabelloni José Miguel Carot Sierra Sabine Taschner-Mandl Vanessa Düster Adela Cañete Ruth Ladenstein Emanuele Neri Luis Martí-Bonmatí |
| author_facet | Diana Veiga-Canuto Leonor Cerdà-Alberich Cinta Sangüesa Nebot Blanca Martínez de las Heras Ulrike Pötschger Michela Gabelloni José Miguel Carot Sierra Sabine Taschner-Mandl Vanessa Düster Adela Cañete Ruth Ladenstein Emanuele Neri Luis Martí-Bonmatí |
| author_sort | Diana Veiga-Canuto |
| collection | DOAJ |
| description | Tumor segmentation is one of the key steps in imaging processing. The goals of this study were to assess the inter-observer variability in manual segmentation of neuroblastic tumors and to analyze whether the state-of-the-art deep learning architecture nnU-Net can provide a robust solution to detect and segment tumors on MR images. A retrospective multicenter study of 132 patients with neuroblastic tumors was performed. Dice Similarity Coefficient (DSC) and Area Under the Receiver Operating Characteristic Curve (AUC ROC) were used to compare segmentation sets. Two more metrics were elaborated to understand the direction of the errors: the modified version of False Positive (FPRm) and False Negative (FNR) rates. Two radiologists manually segmented 46 tumors and a comparative study was performed. nnU-Net was trained-tuned with 106 cases divided into five balanced folds to perform cross-validation. The five resulting models were used as an ensemble solution to measure training (n = 106) and validation (n = 26) performance, independently. The time needed by the model to automatically segment 20 cases was compared to the time required for manual segmentation. The median DSC for manual segmentation sets was 0.969 (±0.032 IQR). The median DSC for the automatic tool was 0.965 (±0.018 IQR). The automatic segmentation model achieved a better performance regarding the FPRm. MR images segmentation variability is similar between radiologists and nnU-Net. Time leverage when using the automatic model with posterior visual validation and manual adjustment corresponds to 92.8%. |
| first_indexed | 2024-03-09T05:33:01Z |
| format | Article |
| id | doaj.art-34a8e06a1e134411af1ea02f0849ec52 |
| institution | Directory Open Access Journal |
| issn | 2072-6694 |
| language | English |
| last_indexed | 2024-03-09T05:33:01Z |
| publishDate | 2022-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Cancers |
| spelling | doaj.art-34a8e06a1e134411af1ea02f0849ec522023-12-03T12:30:42ZengMDPI AGCancers2072-66942022-07-011415364810.3390/cancers14153648Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance ImagesDiana Veiga-Canuto0Leonor Cerdà-Alberich1Cinta Sangüesa Nebot2Blanca Martínez de las Heras3Ulrike Pötschger4Michela Gabelloni5José Miguel Carot Sierra6Sabine Taschner-Mandl7Vanessa Düster8Adela Cañete9Ruth Ladenstein10Emanuele Neri11Luis Martí-Bonmatí12Grupo de Investigación Biomédica en Imagen, Instituto de Investigación Sanitaria La Fe, Avenida Fernando Abril Martorell, 106 Torre A 7planta, 46026 Valencia, SpainGrupo de Investigación Biomédica en Imagen, Instituto de Investigación Sanitaria La Fe, Avenida Fernando Abril Martorell, 106 Torre A 7planta, 46026 Valencia, SpainÁrea Clínica de Imagen Médica, Hospital Universitario y Politécnico La Fe, Avenida Fernando Abril Martorell, 106 Torre A 7planta, 46026 Valencia, SpainUnidad de Oncohematología Pediátrica, Hospital Universitario y Politécnico La Fe, Avenida Fernando Abril Martorell, 106 Torre A 7planta, 46026 Valencia, SpainSt. Anna Children’s Cancer Research Institute, Zimmermannplatz 10, 1090 Vienna, AustriaAcademic Radiology, Department of Translational Research, University of Pisa, Via Roma, 67, 56126 Pisa, ItalyDepartamento de Estadística e Investigación Operativa Aplicadas y Calidad, Universitat Politècnica de València, Camí de Vera s/n, 46022 Valencia, SpainSt. Anna Children’s Cancer Research Institute, Zimmermannplatz 10, 1090 Vienna, AustriaSt. Anna Children’s Cancer Research Institute, Zimmermannplatz 10, 1090 Vienna, AustriaUnidad de Oncohematología Pediátrica, Hospital Universitario y Politécnico La Fe, Avenida Fernando Abril Martorell, 106 Torre A 7planta, 46026 Valencia, SpainSt. Anna Children’s Cancer Research Institute, Zimmermannplatz 10, 1090 Vienna, AustriaAcademic Radiology, Department of Translational Research, University of Pisa, Via Roma, 67, 56126 Pisa, ItalyGrupo de Investigación Biomédica en Imagen, Instituto de Investigación Sanitaria La Fe, Avenida Fernando Abril Martorell, 106 Torre A 7planta, 46026 Valencia, SpainTumor segmentation is one of the key steps in imaging processing. The goals of this study were to assess the inter-observer variability in manual segmentation of neuroblastic tumors and to analyze whether the state-of-the-art deep learning architecture nnU-Net can provide a robust solution to detect and segment tumors on MR images. A retrospective multicenter study of 132 patients with neuroblastic tumors was performed. Dice Similarity Coefficient (DSC) and Area Under the Receiver Operating Characteristic Curve (AUC ROC) were used to compare segmentation sets. Two more metrics were elaborated to understand the direction of the errors: the modified version of False Positive (FPRm) and False Negative (FNR) rates. Two radiologists manually segmented 46 tumors and a comparative study was performed. nnU-Net was trained-tuned with 106 cases divided into five balanced folds to perform cross-validation. The five resulting models were used as an ensemble solution to measure training (n = 106) and validation (n = 26) performance, independently. The time needed by the model to automatically segment 20 cases was compared to the time required for manual segmentation. The median DSC for manual segmentation sets was 0.969 (±0.032 IQR). The median DSC for the automatic tool was 0.965 (±0.018 IQR). The automatic segmentation model achieved a better performance regarding the FPRm. MR images segmentation variability is similar between radiologists and nnU-Net. Time leverage when using the automatic model with posterior visual validation and manual adjustment corresponds to 92.8%.https://www.mdpi.com/2072-6694/14/15/3648tumor segmentationneuroblastic tumorsdeep learningmanual segmentationautomatic segmentationinter-observer variability |
| spellingShingle | Diana Veiga-Canuto Leonor Cerdà-Alberich Cinta Sangüesa Nebot Blanca Martínez de las Heras Ulrike Pötschger Michela Gabelloni José Miguel Carot Sierra Sabine Taschner-Mandl Vanessa Düster Adela Cañete Ruth Ladenstein Emanuele Neri Luis Martí-Bonmatí Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images Cancers tumor segmentation neuroblastic tumors deep learning manual segmentation automatic segmentation inter-observer variability |
| title | Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images |
| title_full | Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images |
| title_fullStr | Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images |
| title_full_unstemmed | Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images |
| title_short | Comparative Multicentric Evaluation of Inter-Observer Variability in Manual and Automatic Segmentation of Neuroblastic Tumors in Magnetic Resonance Images |
| title_sort | comparative multicentric evaluation of inter observer variability in manual and automatic segmentation of neuroblastic tumors in magnetic resonance images |
| topic | tumor segmentation neuroblastic tumors deep learning manual segmentation automatic segmentation inter-observer variability |
| url | https://www.mdpi.com/2072-6694/14/15/3648 |
| work_keys_str_mv | AT dianaveigacanuto comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT leonorcerdaalberich comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT cintasanguesanebot comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT blancamartinezdelasheras comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT ulrikepotschger comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT michelagabelloni comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT josemiguelcarotsierra comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT sabinetaschnermandl comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT vanessaduster comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT adelacanete comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT ruthladenstein comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT emanueleneri comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages AT luismartibonmati comparativemulticentricevaluationofinterobservervariabilityinmanualandautomaticsegmentationofneuroblastictumorsinmagneticresonanceimages |