Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks
Facing an ongoing organ shortage in transplant medicine, strategies to increase the use of organs from marginal donors by objective organ assessment are being fostered. In this context, normothermic machine perfusion provides a platform for ex vivo organ evaluation during preservation. Consequently,...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Biomedicines |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-9059/10/2/397 |
| _version_ | 1827656708357357568 |
|---|---|
| author | Florian Sommer Bingrui Sun Julian Fischer Miriam Goldammer Christine Thiele Hagen Malberg Wenke Markgraf |
| author_facet | Florian Sommer Bingrui Sun Julian Fischer Miriam Goldammer Christine Thiele Hagen Malberg Wenke Markgraf |
| author_sort | Florian Sommer |
| collection | DOAJ |
| description | Facing an ongoing organ shortage in transplant medicine, strategies to increase the use of organs from marginal donors by objective organ assessment are being fostered. In this context, normothermic machine perfusion provides a platform for ex vivo organ evaluation during preservation. Consequently, analytical tools are emerging to determine organ quality. In this study, hyperspectral imaging (HSI) in the wavelength range of 550–995 nm was applied. Classification of 26 kidneys based on HSI was established using KidneyResNet, a convolutional neural network (CNN) based on the ResNet-18 architecture, to predict inulin clearance behavior. HSI preprocessing steps were implemented, including automated region of interest (ROI) selection, before executing the KidneyResNet algorithm. Training parameters and augmentation methods were investigated concerning their influence on the prediction. When classifying individual ROIs, the optimized KidneyResNet model achieved 84% and 62% accuracy in the validation and test set, respectively. With a majority decision on all ROIs of a kidney, the accuracy increased to 96% (validation set) and 100% (test set). These results demonstrate the feasibility of HSI in combination with KidneyResNet for non-invasive prediction of ex vivo kidney function. This knowledge of preoperative renal quality may support the organ acceptance decision. |
| first_indexed | 2024-03-09T22:32:10Z |
| format | Article |
| id | doaj.art-b291f36f0b5e4b17afa236f2de052adc |
| institution | Directory Open Access Journal |
| issn | 2227-9059 |
| language | English |
| last_indexed | 2024-03-09T22:32:10Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Biomedicines |
| spelling | doaj.art-b291f36f0b5e4b17afa236f2de052adc2023-11-23T18:54:55ZengMDPI AGBiomedicines2227-90592022-02-0110239710.3390/biomedicines10020397Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural NetworksFlorian Sommer0Bingrui Sun1Julian Fischer2Miriam Goldammer3Christine Thiele4Hagen Malberg5Wenke Markgraf6Institute of Biomedical Engineering, Technische Universität Dresden, 01307 Dresden, GermanyInstitute of Biomedical Engineering, Technische Universität Dresden, 01307 Dresden, GermanyInstitute of Biomedical Engineering, Technische Universität Dresden, 01307 Dresden, GermanyInstitute of Biomedical Engineering, Technische Universität Dresden, 01307 Dresden, GermanyInstitute of Biomedical Engineering, Technische Universität Dresden, 01307 Dresden, GermanyInstitute of Biomedical Engineering, Technische Universität Dresden, 01307 Dresden, GermanyInstitute of Biomedical Engineering, Technische Universität Dresden, 01307 Dresden, GermanyFacing an ongoing organ shortage in transplant medicine, strategies to increase the use of organs from marginal donors by objective organ assessment are being fostered. In this context, normothermic machine perfusion provides a platform for ex vivo organ evaluation during preservation. Consequently, analytical tools are emerging to determine organ quality. In this study, hyperspectral imaging (HSI) in the wavelength range of 550–995 nm was applied. Classification of 26 kidneys based on HSI was established using KidneyResNet, a convolutional neural network (CNN) based on the ResNet-18 architecture, to predict inulin clearance behavior. HSI preprocessing steps were implemented, including automated region of interest (ROI) selection, before executing the KidneyResNet algorithm. Training parameters and augmentation methods were investigated concerning their influence on the prediction. When classifying individual ROIs, the optimized KidneyResNet model achieved 84% and 62% accuracy in the validation and test set, respectively. With a majority decision on all ROIs of a kidney, the accuracy increased to 96% (validation set) and 100% (test set). These results demonstrate the feasibility of HSI in combination with KidneyResNet for non-invasive prediction of ex vivo kidney function. This knowledge of preoperative renal quality may support the organ acceptance decision.https://www.mdpi.com/2227-9059/10/2/397normothermic machine perfusionorgan preservationkidneybiomedical optical imaginghyperspectral imagingmachine learning |
| spellingShingle | Florian Sommer Bingrui Sun Julian Fischer Miriam Goldammer Christine Thiele Hagen Malberg Wenke Markgraf Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks Biomedicines normothermic machine perfusion organ preservation kidney biomedical optical imaging hyperspectral imaging machine learning |
| title | Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks |
| title_full | Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks |
| title_fullStr | Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks |
| title_full_unstemmed | Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks |
| title_short | Hyperspectral Imaging during Normothermic Machine Perfusion—A Functional Classification of Ex Vivo Kidneys Based on Convolutional Neural Networks |
| title_sort | hyperspectral imaging during normothermic machine perfusion a functional classification of ex vivo kidneys based on convolutional neural networks |
| topic | normothermic machine perfusion organ preservation kidney biomedical optical imaging hyperspectral imaging machine learning |
| url | https://www.mdpi.com/2227-9059/10/2/397 |
| work_keys_str_mv | AT floriansommer hyperspectralimagingduringnormothermicmachineperfusionafunctionalclassificationofexvivokidneysbasedonconvolutionalneuralnetworks AT bingruisun hyperspectralimagingduringnormothermicmachineperfusionafunctionalclassificationofexvivokidneysbasedonconvolutionalneuralnetworks AT julianfischer hyperspectralimagingduringnormothermicmachineperfusionafunctionalclassificationofexvivokidneysbasedonconvolutionalneuralnetworks AT miriamgoldammer hyperspectralimagingduringnormothermicmachineperfusionafunctionalclassificationofexvivokidneysbasedonconvolutionalneuralnetworks AT christinethiele hyperspectralimagingduringnormothermicmachineperfusionafunctionalclassificationofexvivokidneysbasedonconvolutionalneuralnetworks AT hagenmalberg hyperspectralimagingduringnormothermicmachineperfusionafunctionalclassificationofexvivokidneysbasedonconvolutionalneuralnetworks AT wenkemarkgraf hyperspectralimagingduringnormothermicmachineperfusionafunctionalclassificationofexvivokidneysbasedonconvolutionalneuralnetworks |