DB
Dennis Böhm
info
Please Note
<p>This page displays the records of the person named above and is not linked to a unique person identifier. This record may need to be merged to a profile.</p>
1 records found
1
MKNet-family architectures for auto-segmentation of the residual pancreas after pancreatic resection
A deep learning comparative study
Journal article
(2026)
-
Dennis Böhm, Paul C.M. Andel, Paul A. Akkermans, Bas Boekestijn, Willem van der Geest, Robbert J. de Haas, Jakob W. Kist, Michael Weinmann, Lois A. Daamen, More Authors
Purpose
Accurate interpretation of CT scans after pancreatic resection is crucial for detecting abnormalities, including postoperative complications and cancer recurrence. This study investigates the feasibility and clinical utility of a novel MKNet-family deep learning architecture for auto-segmentation of the residual pancreas on postoperative CT imaging, in comparison to previous approaches.
Method
Novel MKNet, MSKNet and MAKNet architectures were developed. Two datasets were used: the National Institutes of Health (NIH) dataset, comprising 82 annotated normal preoperative CT scans, and the IMPACT Consortium dataset (NCT06055010; https://github.com/IMPACTconsortium/IMPACT), comprising 81 annotated postoperative CT scans obtained < 4 weeks after pancreatectomy. Performance was assessed by Hausdorff Distance (HD), 95th-percentile-HD (HD95) and Normalized Surface Distance (NSD), and secondarily by Dice Similarity Coefficient (DSC), and compared with self-implemented existing models for preoperative pancreas auto-segmentation. Qualitative evaluation was conducted by ten abdominal radiologists.
Results
In the postoperative setting, the MAKNet architecture showed the best performance, with an HD and HD95 of 17.3 ± 11.2 mm and 11.5 ± 10.2 mm, respectively. DSC (64.9 ± 14.8%) and NSD (27.2 ± 8.2%) were comparable to the Attention-U-Net (DSC 66.0 ± 13.8%; NSD 27.8 ± 8.4%). Clinical evaluation indicated that the MKNet-family accurately defined the postoperative pancreas (i.e., requiring minimal or no modifications) in 64 of 81 segmentations (79%).
Conclusion
This study demonstrates the effectiveness of novel MKNet-family architectures to accurately segment the residual pancreas on postoperative CT imaging over previous approaches. This advances the state-of-the-art in pancreas auto-segmentation and may be beneficial for medical application and education, acceleration of data annotation, and future research. ...
Accurate interpretation of CT scans after pancreatic resection is crucial for detecting abnormalities, including postoperative complications and cancer recurrence. This study investigates the feasibility and clinical utility of a novel MKNet-family deep learning architecture for auto-segmentation of the residual pancreas on postoperative CT imaging, in comparison to previous approaches.
Method
Novel MKNet, MSKNet and MAKNet architectures were developed. Two datasets were used: the National Institutes of Health (NIH) dataset, comprising 82 annotated normal preoperative CT scans, and the IMPACT Consortium dataset (NCT06055010; https://github.com/IMPACTconsortium/IMPACT), comprising 81 annotated postoperative CT scans obtained < 4 weeks after pancreatectomy. Performance was assessed by Hausdorff Distance (HD), 95th-percentile-HD (HD95) and Normalized Surface Distance (NSD), and secondarily by Dice Similarity Coefficient (DSC), and compared with self-implemented existing models for preoperative pancreas auto-segmentation. Qualitative evaluation was conducted by ten abdominal radiologists.
Results
In the postoperative setting, the MAKNet architecture showed the best performance, with an HD and HD95 of 17.3 ± 11.2 mm and 11.5 ± 10.2 mm, respectively. DSC (64.9 ± 14.8%) and NSD (27.2 ± 8.2%) were comparable to the Attention-U-Net (DSC 66.0 ± 13.8%; NSD 27.8 ± 8.4%). Clinical evaluation indicated that the MKNet-family accurately defined the postoperative pancreas (i.e., requiring minimal or no modifications) in 64 of 81 segmentations (79%).
Conclusion
This study demonstrates the effectiveness of novel MKNet-family architectures to accurately segment the residual pancreas on postoperative CT imaging over previous approaches. This advances the state-of-the-art in pancreas auto-segmentation and may be beneficial for medical application and education, acceleration of data annotation, and future research. ...
Purpose
Accurate interpretation of CT scans after pancreatic resection is crucial for detecting abnormalities, including postoperative complications and cancer recurrence. This study investigates the feasibility and clinical utility of a novel MKNet-family deep learning architecture for auto-segmentation of the residual pancreas on postoperative CT imaging, in comparison to previous approaches.
Method
Novel MKNet, MSKNet and MAKNet architectures were developed. Two datasets were used: the National Institutes of Health (NIH) dataset, comprising 82 annotated normal preoperative CT scans, and the IMPACT Consortium dataset (NCT06055010; https://github.com/IMPACTconsortium/IMPACT), comprising 81 annotated postoperative CT scans obtained < 4 weeks after pancreatectomy. Performance was assessed by Hausdorff Distance (HD), 95th-percentile-HD (HD95) and Normalized Surface Distance (NSD), and secondarily by Dice Similarity Coefficient (DSC), and compared with self-implemented existing models for preoperative pancreas auto-segmentation. Qualitative evaluation was conducted by ten abdominal radiologists.
Results
In the postoperative setting, the MAKNet architecture showed the best performance, with an HD and HD95 of 17.3 ± 11.2 mm and 11.5 ± 10.2 mm, respectively. DSC (64.9 ± 14.8%) and NSD (27.2 ± 8.2%) were comparable to the Attention-U-Net (DSC 66.0 ± 13.8%; NSD 27.8 ± 8.4%). Clinical evaluation indicated that the MKNet-family accurately defined the postoperative pancreas (i.e., requiring minimal or no modifications) in 64 of 81 segmentations (79%).
Conclusion
This study demonstrates the effectiveness of novel MKNet-family architectures to accurately segment the residual pancreas on postoperative CT imaging over previous approaches. This advances the state-of-the-art in pancreas auto-segmentation and may be beneficial for medical application and education, acceleration of data annotation, and future research.
Accurate interpretation of CT scans after pancreatic resection is crucial for detecting abnormalities, including postoperative complications and cancer recurrence. This study investigates the feasibility and clinical utility of a novel MKNet-family deep learning architecture for auto-segmentation of the residual pancreas on postoperative CT imaging, in comparison to previous approaches.
Method
Novel MKNet, MSKNet and MAKNet architectures were developed. Two datasets were used: the National Institutes of Health (NIH) dataset, comprising 82 annotated normal preoperative CT scans, and the IMPACT Consortium dataset (NCT06055010; https://github.com/IMPACTconsortium/IMPACT), comprising 81 annotated postoperative CT scans obtained < 4 weeks after pancreatectomy. Performance was assessed by Hausdorff Distance (HD), 95th-percentile-HD (HD95) and Normalized Surface Distance (NSD), and secondarily by Dice Similarity Coefficient (DSC), and compared with self-implemented existing models for preoperative pancreas auto-segmentation. Qualitative evaluation was conducted by ten abdominal radiologists.
Results
In the postoperative setting, the MAKNet architecture showed the best performance, with an HD and HD95 of 17.3 ± 11.2 mm and 11.5 ± 10.2 mm, respectively. DSC (64.9 ± 14.8%) and NSD (27.2 ± 8.2%) were comparable to the Attention-U-Net (DSC 66.0 ± 13.8%; NSD 27.8 ± 8.4%). Clinical evaluation indicated that the MKNet-family accurately defined the postoperative pancreas (i.e., requiring minimal or no modifications) in 64 of 81 segmentations (79%).
Conclusion
This study demonstrates the effectiveness of novel MKNet-family architectures to accurately segment the residual pancreas on postoperative CT imaging over previous approaches. This advances the state-of-the-art in pancreas auto-segmentation and may be beneficial for medical application and education, acceleration of data annotation, and future research.