Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly Segmentation
Stroke is the second most common cause of death worldwide, responsible for 6.24 million deaths in 2015 (about 11% of all deaths). Three out of four stroke survivors suffer long term disability, as many cannot return to their prior employment or live independently. Eighty-seven percent of strokes are...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2018-04-01
|
| Series: | Frontiers in Neuroinformatics |
| Subjects: | |
| Online Access: | http://journal.frontiersin.org/article/10.3389/fninf.2018.00021/full |
| _version_ | 1818897797051580416 |
|---|---|
| author | Jens K. Boldsen Thorbjørn S. Engedal Salvador Pedraza Tae-Hee Cho Tae-Hee Cho Götz Thomalla Norbert Nighoghossian Norbert Nighoghossian Jean-Claude Baron Jean-Claude Baron Jens Fiehler Leif Østergaard Leif Østergaard Kim Mouridsen |
| author_facet | Jens K. Boldsen Thorbjørn S. Engedal Salvador Pedraza Tae-Hee Cho Tae-Hee Cho Götz Thomalla Norbert Nighoghossian Norbert Nighoghossian Jean-Claude Baron Jean-Claude Baron Jens Fiehler Leif Østergaard Leif Østergaard Kim Mouridsen |
| author_sort | Jens K. Boldsen |
| collection | DOAJ |
| description | Stroke is the second most common cause of death worldwide, responsible for 6.24 million deaths in 2015 (about 11% of all deaths). Three out of four stroke survivors suffer long term disability, as many cannot return to their prior employment or live independently. Eighty-seven percent of strokes are ischemic. As an increasing volume of ischemic brain tissue proceeds to permanent infarction in the hours following the onset, immediate treatment is pivotal to increase the likelihood of good clinical outcome for the patient. Triaging stroke patients for active therapy requires assessment of the volume of salvageable and irreversible damaged tissue, respectively. With Magnetic Resonance Imaging (MRI), diffusion-weighted imaging is commonly used to assess the extent of permanently damaged tissue, the core lesion. To speed up and standardize decision-making in acute stroke management we present a fully automated algorithm, ATLAS, for delineating the core lesion. We compare performance to widely used threshold based methodology, as well as a recently proposed state-of-the-art algorithm: COMBAT Stroke. ATLAS is a machine learning algorithm trained to match the lesion delineation by human experts. The algorithm utilizes decision trees along with spatial pre- and post-regularization to outline the lesion. As input data the algorithm takes images from 108 patients with acute anterior circulation stroke from the I-Know multicenter study. We divided the data into training and test data using leave-one-out cross validation to assess performance in independent patients. Performance was quantified by the Dice index. The median Dice coefficient of ATLAS algorithm was 0.6122, which was significantly higher than COMBAT Stroke, with a median Dice coefficient of 0.5636 (p < 0.0001) and the best possible performing methods based on thresholding of the diffusion weighted images (median Dice coefficient: 0.3951) or the apparent diffusion coefficient (median Dice coefficeint: 0.2839). Furthermore, the volume of the ATLAS segmentation was compared to the volume of the expert segmentation, yielding a standard deviation of the residuals of 10.25 ml compared to 17.53 ml for COMBAT Stroke. Since accurate quantification of the volume of permanently damaged tissue is essential in acute stroke patients, ATLAS may contribute to more optimal patient triaging for active or supportive therapy. |
| first_indexed | 2024-12-19T19:21:53Z |
| format | Article |
| id | doaj.art-7b2cd9f8aaaa40e6958290d9d5e95c09 |
| institution | Directory Open Access Journal |
| issn | 1662-5196 |
| language | English |
| last_indexed | 2024-12-19T19:21:53Z |
| publishDate | 2018-04-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neuroinformatics |
| spelling | doaj.art-7b2cd9f8aaaa40e6958290d9d5e95c092022-12-21T20:08:59ZengFrontiers Media S.A.Frontiers in Neuroinformatics1662-51962018-04-011210.3389/fninf.2018.00021322336Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly SegmentationJens K. Boldsen0Thorbjørn S. Engedal1Salvador Pedraza2Tae-Hee Cho3Tae-Hee Cho4Götz Thomalla5Norbert Nighoghossian6Norbert Nighoghossian7Jean-Claude Baron8Jean-Claude Baron9Jens Fiehler10Leif Østergaard11Leif Østergaard12Kim Mouridsen13Department of Clinical Medicine, Center of Functional Integrative Neuroscience, Aarhus University, Aarhus, DenmarkDepartment of Clinical Medicine, Center of Functional Integrative Neuroscience, Aarhus University, Aarhus, DenmarkRadiology Department, IDI, Hospital Dr Josep Trueta, Institut d'Investigació Biomèdica de Girona (IDIBGI), University of Girona, Girona, SpainStroke Medicine Department, Hôpital Neurologique, Hospices Civils de Lyon, Lyon, FranceCreatis, Centre National de la Recherche Scientifique UMR 5220, Institut National de la Santé et de la Recherche Médicale U1206, INSA de Lyon, Université Lyon 1, Lyon, FranceKlinik und Poliklinik für Neurologie, Kopf- und Neurozentrum, Universitätsklinikum Hamburg-Eppendorf, Hamburg, GermanyStroke Medicine Department, Hôpital Neurologique, Hospices Civils de Lyon, Lyon, FranceCreatis, Centre National de la Recherche Scientifique UMR 5220, Institut National de la Santé et de la Recherche Médicale U1206, INSA de Lyon, Université Lyon 1, Lyon, FranceDepartment of Neurology, Sainte-Anne Hôpital, Paris Descartes University, Institut National de la Santé et de la Recherche Médicale U894, Paris, FranceDepartment of Clinical Neurosciences, University of Cambridge, Cambridge, United KingdomDepartment of Diagnostic and Interventional Neuroradiology, University Medical Center Hamburg-Eppendorf, Hamburg, GermanyDepartment of Clinical Medicine, Center of Functional Integrative Neuroscience, Aarhus University, Aarhus, DenmarkDepartment of Neuroradiology, Aarhus University Hospital, Aarhus, DenmarkDepartment of Clinical Medicine, Center of Functional Integrative Neuroscience, Aarhus University, Aarhus, DenmarkStroke is the second most common cause of death worldwide, responsible for 6.24 million deaths in 2015 (about 11% of all deaths). Three out of four stroke survivors suffer long term disability, as many cannot return to their prior employment or live independently. Eighty-seven percent of strokes are ischemic. As an increasing volume of ischemic brain tissue proceeds to permanent infarction in the hours following the onset, immediate treatment is pivotal to increase the likelihood of good clinical outcome for the patient. Triaging stroke patients for active therapy requires assessment of the volume of salvageable and irreversible damaged tissue, respectively. With Magnetic Resonance Imaging (MRI), diffusion-weighted imaging is commonly used to assess the extent of permanently damaged tissue, the core lesion. To speed up and standardize decision-making in acute stroke management we present a fully automated algorithm, ATLAS, for delineating the core lesion. We compare performance to widely used threshold based methodology, as well as a recently proposed state-of-the-art algorithm: COMBAT Stroke. ATLAS is a machine learning algorithm trained to match the lesion delineation by human experts. The algorithm utilizes decision trees along with spatial pre- and post-regularization to outline the lesion. As input data the algorithm takes images from 108 patients with acute anterior circulation stroke from the I-Know multicenter study. We divided the data into training and test data using leave-one-out cross validation to assess performance in independent patients. Performance was quantified by the Dice index. The median Dice coefficient of ATLAS algorithm was 0.6122, which was significantly higher than COMBAT Stroke, with a median Dice coefficient of 0.5636 (p < 0.0001) and the best possible performing methods based on thresholding of the diffusion weighted images (median Dice coefficient: 0.3951) or the apparent diffusion coefficient (median Dice coefficeint: 0.2839). Furthermore, the volume of the ATLAS segmentation was compared to the volume of the expert segmentation, yielding a standard deviation of the residuals of 10.25 ml compared to 17.53 ml for COMBAT Stroke. Since accurate quantification of the volume of permanently damaged tissue is essential in acute stroke patients, ATLAS may contribute to more optimal patient triaging for active or supportive therapy.http://journal.frontiersin.org/article/10.3389/fninf.2018.00021/fullstrokediffusion MRIsegmentationdiffusion lesioncomputer learningdecision trees |
| spellingShingle | Jens K. Boldsen Thorbjørn S. Engedal Salvador Pedraza Tae-Hee Cho Tae-Hee Cho Götz Thomalla Norbert Nighoghossian Norbert Nighoghossian Jean-Claude Baron Jean-Claude Baron Jens Fiehler Leif Østergaard Leif Østergaard Kim Mouridsen Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly Segmentation Frontiers in Neuroinformatics stroke diffusion MRI segmentation diffusion lesion computer learning decision trees |
| title | Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly Segmentation |
| title_full | Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly Segmentation |
| title_fullStr | Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly Segmentation |
| title_full_unstemmed | Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly Segmentation |
| title_short | Better Diffusion Segmentation in Acute Ischemic Stroke Through Automatic Tree Learning Anomaly Segmentation |
| title_sort | better diffusion segmentation in acute ischemic stroke through automatic tree learning anomaly segmentation |
| topic | stroke diffusion MRI segmentation diffusion lesion computer learning decision trees |
| url | http://journal.frontiersin.org/article/10.3389/fninf.2018.00021/full |
| work_keys_str_mv | AT jenskboldsen betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT thorbjørnsengedal betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT salvadorpedraza betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT taeheecho betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT taeheecho betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT gotzthomalla betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT norbertnighoghossian betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT norbertnighoghossian betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT jeanclaudebaron betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT jeanclaudebaron betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT jensfiehler betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT leiføstergaard betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT leiføstergaard betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation AT kimmouridsen betterdiffusionsegmentationinacuteischemicstrokethroughautomatictreelearninganomalysegmentation |