High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approach

Abstract Spontaneous intracranial hemorrhages have a high disease burden. Due to increasing medical imaging, new technological solutions for assisting in image interpretation are warranted. We developed a deep learning (DL) solution for spontaneous intracranial hemorrhage detection from head CT scan...

Full description

Saved in:
Bibliographic Details
Main Authors: Juuso Takala, Heikki Peura, Riku Pirinen, Katri Väätäinen, Sergei Terjajev, Ziyuan Lin, Rahul Raj, Miikka Korja
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-15835-7
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849344979580747776
author Juuso Takala
Heikki Peura
Riku Pirinen
Katri Väätäinen
Sergei Terjajev
Ziyuan Lin
Rahul Raj
Miikka Korja
author_facet Juuso Takala
Heikki Peura
Riku Pirinen
Katri Väätäinen
Sergei Terjajev
Ziyuan Lin
Rahul Raj
Miikka Korja
author_sort Juuso Takala
collection DOAJ
description Abstract Spontaneous intracranial hemorrhages have a high disease burden. Due to increasing medical imaging, new technological solutions for assisting in image interpretation are warranted. We developed a deep learning (DL) solution for spontaneous intracranial hemorrhage detection from head CT scans. The DL solution included four base convolutional neural networks (CNNs), which were trained using 300 head CT scans. A metamodel was trained on top of the four base CNNs, and simple post processing steps were applied to improve the solution’s accuracy. The solution performance was evaluated using a retrospective dataset of consecutive emergency head CTs imaged in ten different emergency rooms. 7797 head CT scans were included in the validation dataset and 118 CT scans presented with spontaneous intracranial hemorrhage. The trained metamodel together with a simple rule-based post-processing step showed 89.8% sensitivity and 89.5% specificity for hemorrhage detection at the case-level. The solution detected all 78 spontaneous hemorrhage cases imaged presumably or confirmedly within 12 h from the symptom onset and identified five hemorrhages missed in the initial on-call reports. Although the success of DL algorithms depends on multiple factors, including training data versatility and quality of annotations, using the proposed ensemble-learning approach and rule-based post-processing may help clinicians to develop highly accurate DL solutions for clinical imaging diagnostics.
format Article
id doaj-art-263db48e072e456db131bbec24c580d3
institution Kabale University
issn 2045-2322
language English
publishDate 2025-08-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-263db48e072e456db131bbec24c580d32025-08-20T03:42:31ZengNature PortfolioScientific Reports2045-23222025-08-0115111110.1038/s41598-025-15835-7High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approachJuuso Takala0Heikki Peura1Riku Pirinen2Katri Väätäinen3Sergei Terjajev4Ziyuan Lin5Rahul Raj6Miikka Korja7Department of Neurosurgery, University of Helsinki and Helsinki University HospitalDepartment of Neurosurgery, University of Helsinki and Helsinki University HospitalDepartment of Neurosurgery, University of Helsinki and Helsinki University HospitalDiagnostic Center, Helsinki University HospitalDiagnostic Center, Helsinki University HospitalPlanmecaDepartment of Neurosurgery, University of Helsinki and Helsinki University HospitalDepartment of Neurosurgery, University of Helsinki and Helsinki University HospitalAbstract Spontaneous intracranial hemorrhages have a high disease burden. Due to increasing medical imaging, new technological solutions for assisting in image interpretation are warranted. We developed a deep learning (DL) solution for spontaneous intracranial hemorrhage detection from head CT scans. The DL solution included four base convolutional neural networks (CNNs), which were trained using 300 head CT scans. A metamodel was trained on top of the four base CNNs, and simple post processing steps were applied to improve the solution’s accuracy. The solution performance was evaluated using a retrospective dataset of consecutive emergency head CTs imaged in ten different emergency rooms. 7797 head CT scans were included in the validation dataset and 118 CT scans presented with spontaneous intracranial hemorrhage. The trained metamodel together with a simple rule-based post-processing step showed 89.8% sensitivity and 89.5% specificity for hemorrhage detection at the case-level. The solution detected all 78 spontaneous hemorrhage cases imaged presumably or confirmedly within 12 h from the symptom onset and identified five hemorrhages missed in the initial on-call reports. Although the success of DL algorithms depends on multiple factors, including training data versatility and quality of annotations, using the proposed ensemble-learning approach and rule-based post-processing may help clinicians to develop highly accurate DL solutions for clinical imaging diagnostics.https://doi.org/10.1038/s41598-025-15835-7Intracranial hemorrhageTomography, computedDeep learningEnsemble learningConvolutional neural networks
spellingShingle Juuso Takala
Heikki Peura
Riku Pirinen
Katri Väätäinen
Sergei Terjajev
Ziyuan Lin
Rahul Raj
Miikka Korja
High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approach
Scientific Reports
Intracranial hemorrhage
Tomography, computed
Deep learning
Ensemble learning
Convolutional neural networks
title High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approach
title_full High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approach
title_fullStr High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approach
title_full_unstemmed High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approach
title_short High sensitivity in spontaneous intracranial hemorrhage detection from emergency head CT scans using ensemble-learning approach
title_sort high sensitivity in spontaneous intracranial hemorrhage detection from emergency head ct scans using ensemble learning approach
topic Intracranial hemorrhage
Tomography, computed
Deep learning
Ensemble learning
Convolutional neural networks
url https://doi.org/10.1038/s41598-025-15835-7
work_keys_str_mv AT juusotakala highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach
AT heikkipeura highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach
AT rikupirinen highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach
AT katrivaatainen highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach
AT sergeiterjajev highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach
AT ziyuanlin highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach
AT rahulraj highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach
AT miikkakorja highsensitivityinspontaneousintracranialhemorrhagedetectionfromemergencyheadctscansusingensemblelearningapproach