Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection images
Abstract Background Early diagnosis of the potentially fatal but curable chronic pulmonary embolism (CPE) is challenging. We have developed and investigated a novel convolutional neural network (CNN) model to recognise CPE from CT pulmonary angiograms (CTPA) based on the general vascular morphology...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
SpringerOpen
2023-06-01
|
| Series: | European Radiology Experimental |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s41747-023-00346-9 |
| _version_ | 1797795854593032192 |
|---|---|
| author | Tuomas Vainio Teemu Mäkelä Anssi Arkko Sauli Savolainen Marko Kangasniemi |
| author_facet | Tuomas Vainio Teemu Mäkelä Anssi Arkko Sauli Savolainen Marko Kangasniemi |
| author_sort | Tuomas Vainio |
| collection | DOAJ |
| description | Abstract Background Early diagnosis of the potentially fatal but curable chronic pulmonary embolism (CPE) is challenging. We have developed and investigated a novel convolutional neural network (CNN) model to recognise CPE from CT pulmonary angiograms (CTPA) based on the general vascular morphology in two-dimensional (2D) maximum intensity projection images. Methods A CNN model was trained on a curated subset of a public pulmonary embolism CT dataset (RSPECT) with 755 CTPA studies, including patient-level labels of CPE, acute pulmonary embolism (APE), or no pulmonary embolism. CPE patients with right-to-left-ventricular ratio (RV/LV) < 1 and APE patients with RV/LV ≥ 1 were excluded from the training. Additional CNN model selection and testing were done on local data with 78 patients without the RV/LV-based exclusion. We calculated area under the receiver operating characteristic curves (AUC) and balanced accuracies to evaluate the CNN performance. Results We achieved a very high CPE versus no-CPE classification AUC 0.94 and balanced accuracy 0.89 on the local dataset using an ensemble model and considering CPE to be present in either one or both lungs. Conclusions We propose a novel CNN model with excellent predictive accuracy to differentiate chronic pulmonary embolism with RV/LV ≥ 1 from acute pulmonary embolism and non-embolic cases from 2D maximum intensity projection reconstructions of CTPA. Relevance statement A DL CNN model identifies chronic pulmonary embolism from CTA with an excellent predictive accuracy. Key points • Automatic recognition of CPE from computed tomography pulmonary angiography was developed. • Deep learning was applied on two-dimensional maximum intensity projection images. • A large public dataset was used for training the deep learning model. • The proposed model showed an excellent predictive accuracy. Graphical Abstract |
| first_indexed | 2024-03-13T03:24:17Z |
| format | Article |
| id | doaj.art-7d350180478842e0bde767e32ec86018 |
| institution | Directory Open Access Journal |
| issn | 2509-9280 |
| language | English |
| last_indexed | 2024-03-13T03:24:17Z |
| publishDate | 2023-06-01 |
| publisher | SpringerOpen |
| record_format | Article |
| series | European Radiology Experimental |
| spelling | doaj.art-7d350180478842e0bde767e32ec860182023-06-25T11:08:35ZengSpringerOpenEuropean Radiology Experimental2509-92802023-06-017111310.1186/s41747-023-00346-9Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection imagesTuomas Vainio0Teemu Mäkelä1Anssi Arkko2Sauli Savolainen3Marko Kangasniemi4Radiology, HUS Medical Imaging Center, University of Helsinki and Helsinki University HospitalRadiology, HUS Medical Imaging Center, University of Helsinki and Helsinki University HospitalRadiology, HUS Medical Imaging Center, University of Helsinki and Helsinki University HospitalRadiology, HUS Medical Imaging Center, University of Helsinki and Helsinki University HospitalRadiology, HUS Medical Imaging Center, University of Helsinki and Helsinki University HospitalAbstract Background Early diagnosis of the potentially fatal but curable chronic pulmonary embolism (CPE) is challenging. We have developed and investigated a novel convolutional neural network (CNN) model to recognise CPE from CT pulmonary angiograms (CTPA) based on the general vascular morphology in two-dimensional (2D) maximum intensity projection images. Methods A CNN model was trained on a curated subset of a public pulmonary embolism CT dataset (RSPECT) with 755 CTPA studies, including patient-level labels of CPE, acute pulmonary embolism (APE), or no pulmonary embolism. CPE patients with right-to-left-ventricular ratio (RV/LV) < 1 and APE patients with RV/LV ≥ 1 were excluded from the training. Additional CNN model selection and testing were done on local data with 78 patients without the RV/LV-based exclusion. We calculated area under the receiver operating characteristic curves (AUC) and balanced accuracies to evaluate the CNN performance. Results We achieved a very high CPE versus no-CPE classification AUC 0.94 and balanced accuracy 0.89 on the local dataset using an ensemble model and considering CPE to be present in either one or both lungs. Conclusions We propose a novel CNN model with excellent predictive accuracy to differentiate chronic pulmonary embolism with RV/LV ≥ 1 from acute pulmonary embolism and non-embolic cases from 2D maximum intensity projection reconstructions of CTPA. Relevance statement A DL CNN model identifies chronic pulmonary embolism from CTA with an excellent predictive accuracy. Key points • Automatic recognition of CPE from computed tomography pulmonary angiography was developed. • Deep learning was applied on two-dimensional maximum intensity projection images. • A large public dataset was used for training the deep learning model. • The proposed model showed an excellent predictive accuracy. Graphical Abstracthttps://doi.org/10.1186/s41747-023-00346-9Artificial intelligenceComputed tomography angiographyDeep learningNeural networks (computer)Pulmonary embolism |
| spellingShingle | Tuomas Vainio Teemu Mäkelä Anssi Arkko Sauli Savolainen Marko Kangasniemi Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection images European Radiology Experimental Artificial intelligence Computed tomography angiography Deep learning Neural networks (computer) Pulmonary embolism |
| title | Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection images |
| title_full | Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection images |
| title_fullStr | Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection images |
| title_full_unstemmed | Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection images |
| title_short | Leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from CT angiogram maximum intensity projection images |
| title_sort | leveraging open dataset and transfer learning for accurate recognition of chronic pulmonary embolism from ct angiogram maximum intensity projection images |
| topic | Artificial intelligence Computed tomography angiography Deep learning Neural networks (computer) Pulmonary embolism |
| url | https://doi.org/10.1186/s41747-023-00346-9 |
| work_keys_str_mv | AT tuomasvainio leveragingopendatasetandtransferlearningforaccuraterecognitionofchronicpulmonaryembolismfromctangiogrammaximumintensityprojectionimages AT teemumakela leveragingopendatasetandtransferlearningforaccuraterecognitionofchronicpulmonaryembolismfromctangiogrammaximumintensityprojectionimages AT anssiarkko leveragingopendatasetandtransferlearningforaccuraterecognitionofchronicpulmonaryembolismfromctangiogrammaximumintensityprojectionimages AT saulisavolainen leveragingopendatasetandtransferlearningforaccuraterecognitionofchronicpulmonaryembolismfromctangiogrammaximumintensityprojectionimages AT markokangasniemi leveragingopendatasetandtransferlearningforaccuraterecognitionofchronicpulmonaryembolismfromctangiogrammaximumintensityprojectionimages |