A machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomography
Abstract Background In most medical centers, particularly in primary hospitals, non-contrast computed tomography (NCCT) serves as the primary imaging modality for diagnosing acute ischemic stroke. However, due to the small density difference between the infarct and the surrounding normal brain tissu...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-07-01
|
| Series: | BMC Medical Imaging |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12880-025-01822-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849332103437615104 |
|---|---|
| author | Jiahe Tan Mengjun Xiao Zhipeng Wang Shuzhen Wu Kun Han Haiyan Wang Yong Huang |
| author_facet | Jiahe Tan Mengjun Xiao Zhipeng Wang Shuzhen Wu Kun Han Haiyan Wang Yong Huang |
| author_sort | Jiahe Tan |
| collection | DOAJ |
| description | Abstract Background In most medical centers, particularly in primary hospitals, non-contrast computed tomography (NCCT) serves as the primary imaging modality for diagnosing acute ischemic stroke. However, due to the small density difference between the infarct and the surrounding normal brain tissue on NCCT images within the initial 6 h post-onset, it poses significant challenges in promptly and accurately positioning and quantifying the infarct at the early stage. Aims To investigate whether a radiomics-based model using NCCT could effectively assess the risk of acute ischemic stroke (AIS). Methods This study proposed a machine learning (ML) for infarct detection, enabling automated quantitative assessment of AIS lesions on NCCT images. In this retrospective study, NCCT images from 228 patients with AIS (< 6 h from onset) were included, and paired with MRI-diffusion-weighted imaging (DWI) images (attained within 1 to 7 days of onset). NCCT and DWI images were co-registered using the Elastix toolbox. The internal dataset (153 AIS patients) included 179 AIS VOIs and 153 non-AIS VOIs as the training and validation groups. Subsequent cases (75 patients) after 2021 served as the independent test set, comprising 94 AIS VOIs and 75 non-AIS VOIs. Results The random forest (RF) model demonstrated robust diagnostic performance across the training, validation, and independent test sets. The areas under the receiver operating characteristic (ROC) curves were 0.858 (95% CI: 0.808–0.908), 0.829 (95% CI: 0.748–0.910), and 0.789 (95% CI: 0.717–0.860), respectively. Accuracies were 79.399%, 77.778%, and 73.965%, while sensitivities were 81.679%, 77.083%, and 68.085%. Specificities were 76.471%, 78.431%, and 81.333%, respectively. Conclusion NCCT-based radiomics combined with a machine learning model could discriminate between AIS and non-AIS patients within less than 6 h of onset. This approach holds promise for improving early stroke diagnosis and patient outcomes. Clinical trial number Not applicable. |
| format | Article |
| id | doaj-art-285d6ab730a04558a9697d246e7e573c |
| institution | Kabale University |
| issn | 1471-2342 |
| language | English |
| publishDate | 2025-07-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Medical Imaging |
| spelling | doaj-art-285d6ab730a04558a9697d246e7e573c2025-08-20T03:46:20ZengBMCBMC Medical Imaging1471-23422025-07-012511910.1186/s12880-025-01822-xA machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomographyJiahe Tan0Mengjun Xiao1Zhipeng Wang2Shuzhen Wu3Kun Han4Haiyan Wang5Yong Huang6Computer Science, Graduate Studies, University of CaliforniaDepartment of Radiology, Shandong Provincial Hospital, Shandong First Medical UniversityDepartment of Radiology, Shandong Provincial Hospital, Shandong First Medical UniversityDepartment of Radiology, Shandong Provincial Hospital, Shandong First Medical UniversityDepartment of Radiology, Shandong Provincial Hospital, Shandong First Medical UniversityDepartment of Radiology, Shandong Provincial Hospital, Shandong First Medical UniversityDepartment of Radiology, Shandong Cancer Hospital and Institute, Shandong First Medical University, Shandong Academy of Medical SciencesAbstract Background In most medical centers, particularly in primary hospitals, non-contrast computed tomography (NCCT) serves as the primary imaging modality for diagnosing acute ischemic stroke. However, due to the small density difference between the infarct and the surrounding normal brain tissue on NCCT images within the initial 6 h post-onset, it poses significant challenges in promptly and accurately positioning and quantifying the infarct at the early stage. Aims To investigate whether a radiomics-based model using NCCT could effectively assess the risk of acute ischemic stroke (AIS). Methods This study proposed a machine learning (ML) for infarct detection, enabling automated quantitative assessment of AIS lesions on NCCT images. In this retrospective study, NCCT images from 228 patients with AIS (< 6 h from onset) were included, and paired with MRI-diffusion-weighted imaging (DWI) images (attained within 1 to 7 days of onset). NCCT and DWI images were co-registered using the Elastix toolbox. The internal dataset (153 AIS patients) included 179 AIS VOIs and 153 non-AIS VOIs as the training and validation groups. Subsequent cases (75 patients) after 2021 served as the independent test set, comprising 94 AIS VOIs and 75 non-AIS VOIs. Results The random forest (RF) model demonstrated robust diagnostic performance across the training, validation, and independent test sets. The areas under the receiver operating characteristic (ROC) curves were 0.858 (95% CI: 0.808–0.908), 0.829 (95% CI: 0.748–0.910), and 0.789 (95% CI: 0.717–0.860), respectively. Accuracies were 79.399%, 77.778%, and 73.965%, while sensitivities were 81.679%, 77.083%, and 68.085%. Specificities were 76.471%, 78.431%, and 81.333%, respectively. Conclusion NCCT-based radiomics combined with a machine learning model could discriminate between AIS and non-AIS patients within less than 6 h of onset. This approach holds promise for improving early stroke diagnosis and patient outcomes. Clinical trial number Not applicable.https://doi.org/10.1186/s12880-025-01822-xArtificial intelligenceRadiomicsAcute ischaemic strokeComputed tomographyMachine learning |
| spellingShingle | Jiahe Tan Mengjun Xiao Zhipeng Wang Shuzhen Wu Kun Han Haiyan Wang Yong Huang A machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomography BMC Medical Imaging Artificial intelligence Radiomics Acute ischaemic stroke Computed tomography Machine learning |
| title | A machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomography |
| title_full | A machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomography |
| title_fullStr | A machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomography |
| title_full_unstemmed | A machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomography |
| title_short | A machine learning model reveals invisible microscopic variation in acute ischaemic stroke (≤ 6 h) with non-contrast computed tomography |
| title_sort | machine learning model reveals invisible microscopic variation in acute ischaemic stroke ≤ 6 h with non contrast computed tomography |
| topic | Artificial intelligence Radiomics Acute ischaemic stroke Computed tomography Machine learning |
| url | https://doi.org/10.1186/s12880-025-01822-x |
| work_keys_str_mv | AT jiahetan amachinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT mengjunxiao amachinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT zhipengwang amachinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT shuzhenwu amachinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT kunhan amachinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT haiyanwang amachinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT yonghuang amachinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT jiahetan machinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT mengjunxiao machinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT zhipengwang machinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT shuzhenwu machinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT kunhan machinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT haiyanwang machinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography AT yonghuang machinelearningmodelrevealsinvisiblemicroscopicvariationinacuteischaemicstroke6hwithnoncontrastcomputedtomography |