Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural Network
<b>Background:</b> Accurate delineation of lesions in acute ischemic stroke is important for determining the extent of tissue damage and the identification of potentially salvageable brain tissues. Automatic segmentation on CT images is challenging due to the poor contrast-to-noise ratio...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | AI |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2673-2688/6/1/15 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589416735965184 |
|---|---|
| author | Ankit Kandpal Rakesh Kumar Gupta Anup Singh |
| author_facet | Ankit Kandpal Rakesh Kumar Gupta Anup Singh |
| author_sort | Ankit Kandpal |
| collection | DOAJ |
| description | <b>Background:</b> Accurate delineation of lesions in acute ischemic stroke is important for determining the extent of tissue damage and the identification of potentially salvageable brain tissues. Automatic segmentation on CT images is challenging due to the poor contrast-to-noise ratio. Quantitative CT perfusion images improve the estimation of the perfusion deficit regions; however, they are limited by a poor signal-to-noise ratio. The study aims to investigate the potential of deep learning (DL) algorithms for the improved segmentation of ischemic lesions. <b>Methods:</b> This study proposes a novel DL architecture, DenseResU-NetCTPSS, for stroke segmentation using multiparametric CT perfusion images. The proposed network is benchmarked against state-of-the-art DL models. Its performance is assessed using the ISLES-2018 challenge dataset, a widely recognized dataset for stroke segmentation in CT images. The proposed network was evaluated on both training and test datasets. <b>Results:</b> The final optimized network takes three image sequences, namely CT, cerebral blood volume (CBV), and time to max (Tmax), as input to perform segmentation. The network achieved a dice score of 0.65 ± 0.19 and 0.45 ± 0.32 on the training and testing datasets. The model demonstrated a notable improvement over existing state-of-the-art DL models. <b>Conclusions:</b> The optimized model combines CT, CBV, and T<sub>max</sub> images, enabling automatic lesion identification with reasonable accuracy and aiding radiologists in faster, more objective assessments. |
| format | Article |
| id | doaj-art-9a0d571500eb42a488f866e1565442f3 |
| institution | Kabale University |
| issn | 2673-2688 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | AI |
| spelling | doaj-art-9a0d571500eb42a488f866e1565442f32025-01-24T13:17:24ZengMDPI AGAI2673-26882025-01-01611510.3390/ai6010015Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural NetworkAnkit Kandpal0Rakesh Kumar Gupta1Anup Singh2Centre for Biomedical Engineering, Indian Institute of Technology Delhi, New Delhi 110010, IndiaDepartment of Radiology, Fortis Memorial Research Institute, Gurugram 122002, IndiaCentre for Biomedical Engineering, Indian Institute of Technology Delhi, New Delhi 110010, India<b>Background:</b> Accurate delineation of lesions in acute ischemic stroke is important for determining the extent of tissue damage and the identification of potentially salvageable brain tissues. Automatic segmentation on CT images is challenging due to the poor contrast-to-noise ratio. Quantitative CT perfusion images improve the estimation of the perfusion deficit regions; however, they are limited by a poor signal-to-noise ratio. The study aims to investigate the potential of deep learning (DL) algorithms for the improved segmentation of ischemic lesions. <b>Methods:</b> This study proposes a novel DL architecture, DenseResU-NetCTPSS, for stroke segmentation using multiparametric CT perfusion images. The proposed network is benchmarked against state-of-the-art DL models. Its performance is assessed using the ISLES-2018 challenge dataset, a widely recognized dataset for stroke segmentation in CT images. The proposed network was evaluated on both training and test datasets. <b>Results:</b> The final optimized network takes three image sequences, namely CT, cerebral blood volume (CBV), and time to max (Tmax), as input to perform segmentation. The network achieved a dice score of 0.65 ± 0.19 and 0.45 ± 0.32 on the training and testing datasets. The model demonstrated a notable improvement over existing state-of-the-art DL models. <b>Conclusions:</b> The optimized model combines CT, CBV, and T<sub>max</sub> images, enabling automatic lesion identification with reasonable accuracy and aiding radiologists in faster, more objective assessments.https://www.mdpi.com/2673-2688/6/1/15computer aided diagnosiscomputed tomographydeep learningischemic strokeCT perfusionmedical image segmentation |
| spellingShingle | Ankit Kandpal Rakesh Kumar Gupta Anup Singh Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural Network AI computer aided diagnosis computed tomography deep learning ischemic stroke CT perfusion medical image segmentation |
| title | Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural Network |
| title_full | Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural Network |
| title_fullStr | Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural Network |
| title_full_unstemmed | Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural Network |
| title_short | Ischemic Stroke Lesion Segmentation on Multiparametric CT Perfusion Maps Using Deep Neural Network |
| title_sort | ischemic stroke lesion segmentation on multiparametric ct perfusion maps using deep neural network |
| topic | computer aided diagnosis computed tomography deep learning ischemic stroke CT perfusion medical image segmentation |
| url | https://www.mdpi.com/2673-2688/6/1/15 |
| work_keys_str_mv | AT ankitkandpal ischemicstrokelesionsegmentationonmultiparametricctperfusionmapsusingdeepneuralnetwork AT rakeshkumargupta ischemicstrokelesionsegmentationonmultiparametricctperfusionmapsusingdeepneuralnetwork AT anupsingh ischemicstrokelesionsegmentationonmultiparametricctperfusionmapsusingdeepneuralnetwork |