Leonardus B van den Oever1, Ludo Cornelissen1, Marleen Vonder2, Congying Xia3, Jurjen N van Bolhuis4, Rozemarijn Vliegenthart3, Raymond N J Veldhuis5, Geertruida H de Bock2, Matthijs Oudkerk6, Peter M A van Ooijen7. 1. University of Groningen, University Medical Center Groningen, Department of Radiation Oncology, Groningen, the Netherlands. 2. University of Groningen, University Medical Center Groningen, Department of Epidemiology, Groningen, the Netherlands. 3. University of Groningen, University Medical Center Groningen, Department of Radiology, Groningen, the Netherlands. 4. Lifelines Cohort Study, Groningen, the Netherlands. 5. University of Twente, Department of Electrical Engineering, Computer Science and Mathematics, Enschede, the Netherlands. 6. University of Groningen, Faculty of Medical Sciences, Groningen, the Netherlands. 7. University of Groningen, University Medical Center Groningen, Department of Radiation Oncology, Groningen, the Netherlands. Electronic address: p.m.a.van.ooijen@umcg.nl.
Abstract
PURPOSE: Coronary artery calcium (CAC) score has shown to be an accurate predictor of future cardiovascular events. Early detection by CAC scoring might reduce the number of deaths by cardiovascular disease (CVD). Automatically excluding scans which test negative for CAC could significantly reduce the workload of radiologists. We propose an algorithm that both excludes negative scans and segments the CAC. METHOD: The training and internal validation data were collected from the ROBINSCA study. The external validation data were collected from the ImaLife study. Both contain annotated low-dose non-contrast cardiac CT scans. 60 scans of participants were used for training and 2 sets of 50 CT scans of participants without CAC and 50 CT scans of participants with an Agatston score between 10 and 20 were collected for both internal and external validation. The effect of dilated convolutional layers was tested by using 2 CNN architectures. We used the patient-level accuracy as metric for assessing the accuracy of our pipeline for detection of CAC and the Dice coefficient score as metric for the segmentation of CAC. RESULTS: Of the 50 negative cases in the internal and external validation set, 62 % and 86 % were classified correctly, respectively. There were no false negative predictions. For the segmentation task, Dice Coefficient scores of 0.63 and 0.84 were achieved for the internal and external validation datasets, respectively. CONCLUSIONS: Our algorithm excluded 86 % of all scans without CAC. Radiologists might need to spend less time on participants without CAC and could spend more time on participants that need their attention.
PURPOSE: Coronary artery calcium (CAC) score has shown to be an accurate predictor of future cardiovascular events. Early detection by CAC scoring might reduce the number of deaths by cardiovascular disease (CVD). Automatically excluding scans which test negative for CAC could significantly reduce the workload of radiologists. We propose an algorithm that both excludes negative scans and segments the CAC. METHOD: The training and internal validation data were collected from the ROBINSCA study. The external validation data were collected from the ImaLife study. Both contain annotated low-dose non-contrast cardiac CT scans. 60 scans of participants were used for training and 2 sets of 50 CT scans of participants without CAC and 50 CT scans of participants with an Agatston score between 10 and 20 were collected for both internal and external validation. The effect of dilated convolutional layers was tested by using 2 CNN architectures. We used the patient-level accuracy as metric for assessing the accuracy of our pipeline for detection of CAC and the Dice coefficient score as metric for the segmentation of CAC. RESULTS: Of the 50 negative cases in the internal and external validation set, 62 % and 86 % were classified correctly, respectively. There were no false negative predictions. For the segmentation task, Dice Coefficient scores of 0.63 and 0.84 were achieved for the internal and external validation datasets, respectively. CONCLUSIONS: Our algorithm excluded 86 % of all scans without CAC. Radiologists might need to spend less time on participants without CAC and could spend more time on participants that need their attention.
Authors: Riemer H J A Slart; Michelle C Williams; Luis Eduardo Juarez-Orozco; Christoph Rischpler; Marc R Dweck; Andor W J M Glaudemans; Alessia Gimelli; Panagiotis Georgoulias; Olivier Gheysens; Oliver Gaemperli; Gilbert Habib; Roland Hustinx; Bernard Cosyns; Hein J Verberne; Fabien Hyafil; Paola A Erba; Mark Lubberink; Piotr Slomka; Ivana Išgum; Dimitris Visvikis; Márton Kolossváry; Antti Saraste Journal: Eur J Nucl Med Mol Imaging Date: 2021-04-17 Impact factor: 9.236
Authors: L B van den Oever; W A van Veldhuizen; L J Cornelissen; D S Spoor; T P Willems; G Kramer; T Stigter; M Rook; A P G Crijns; M Oudkerk; R N J Veldhuis; G H de Bock; P M A van Ooijen Journal: J Digit Imaging Date: 2022-01-26 Impact factor: 4.056
Authors: L B van den Oever; D S Spoor; A P G Crijns; R Vliegenthart; M Oudkerk; R N J Veldhuis; G H de Bock; P M A van Ooijen Journal: J Med Syst Date: 2022-03-25 Impact factor: 4.920