L Pennig1, R Shahzad1,2, L Caldeira1, S Lennartz1, F Thiele1,2, L Goertz3, D Zopfs1, A-K Meißner4, G Fürtjes3,4, M Perkuhn1,2, C Kabbasch1, S Grau3, J Borggrefe1, K R Laukamp5,6,7. 1. From the Institute for Diagnostic and Interventional Radiology (L.P., R.S., L.C., S.L., F.T., D.Z., M.P., C.K., J.B., K.R.L.). 2. Philips Innovative Technologies (R.S., F.T., M.P.), Aachen, Germany. 3. Center for Neurosurgery (L.G., G.F., S.G.), Faculty of Medicine and University Hospital Cologne, University of Cologne, Cologne, Germany. 4. Department of Stereotaxy and Functional Neurosurgery (A.-K.M., G.F.), Center for Neurosurgery, University Hospital Cologne, Cologne, Germany. 5. From the Institute for Diagnostic and Interventional Radiology (L.P., R.S., L.C., S.L., F.T., D.Z., M.P., C.K., J.B., K.R.L.) kai.laukamp@uk-koeln.de. 6. Department of Radiology (K.R.L.), University Hospitals Cleveland Medical Center, Cleveland, Ohio. 7. Department of Radiology (K.R.L.), Case Western Reserve University, Cleveland, Ohio.
Abstract
BACKGROUND AND PURPOSE: Malignant melanoma is an aggressive skin cancer in which brain metastases are common. Our aim was to establish and evaluate a deep learning model for fully automated detection and segmentation of brain metastases in patients with malignant melanoma using clinical routine MR imaging. MATERIALS AND METHODS: Sixty-nine patients with melanoma with a total of 135 brain metastases at initial diagnosis and available multiparametric MR imaging datasets (T1-/T2-weighted, T1-weighted gadolinium contrast-enhanced, FLAIR) were included. A previously established deep learning model architecture (3D convolutional neural network; DeepMedic) simultaneously operating on the aforementioned MR images was trained on a cohort of 55 patients with 103 metastases using 5-fold cross-validation. The efficacy of the deep learning model was evaluated using an independent test set consisting of 14 patients with 32 metastases. Manual segmentations of metastases in a voxelwise manner (T1-weighted gadolinium contrast-enhanced imaging) performed by 2 radiologists in consensus served as the ground truth. RESULTS: After training, the deep learning model detected 28 of 32 brain metastases (mean volume, 1.0 [SD, 2.4] cm3) in the test cohort correctly (sensitivity of 88%), while false-positive findings of 0.71 per scan were observed. Compared with the ground truth, automated segmentations achieved a median Dice similarity coefficient of 0.75. CONCLUSIONS: Deep learning-based automated detection and segmentation of brain metastases in malignant melanoma yields high detection and segmentation accuracy with false-positive findings of <1 per scan.
BACKGROUND AND PURPOSE: Malignant melanoma is an aggressive skin cancer in which brain metastases are common. Our aim was to establish and evaluate a deep learning model for fully automated detection and segmentation of brain metastases in patients with malignant melanoma using clinical routine MR imaging. MATERIALS AND METHODS: Sixty-nine patients with melanoma with a total of 135 brain metastases at initial diagnosis and available multiparametric MR imaging datasets (T1-/T2-weighted, T1-weighted gadolinium contrast-enhanced, FLAIR) were included. A previously established deep learning model architecture (3D convolutional neural network; DeepMedic) simultaneously operating on the aforementioned MR images was trained on a cohort of 55 patients with 103 metastases using 5-fold cross-validation. The efficacy of the deep learning model was evaluated using an independent test set consisting of 14 patients with 32 metastases. Manual segmentations of metastases in a voxelwise manner (T1-weighted gadolinium contrast-enhanced imaging) performed by 2 radiologists in consensus served as the ground truth. RESULTS: After training, the deep learning model detected 28 of 32 brain metastases (mean volume, 1.0 [SD, 2.4] cm3) in the test cohort correctly (sensitivity of 88%), while false-positive findings of 0.71 per scan were observed. Compared with the ground truth, automated segmentations achieved a median Dice similarity coefficient of 0.75. CONCLUSIONS: Deep learning-based automated detection and segmentation of brain metastases in malignant melanoma yields high detection and segmentation accuracy with false-positive findings of <1 per scan.
Authors: Lenhard Pennig; Ulrike Cornelia Isabel Hoyer; Alexandra Krauskopf; Rahil Shahzad; Stephanie T Jünger; Frank Thiele; Kai Roman Laukamp; Jan-Peter Grunz; Michael Perkuhn; Marc Schlamann; Christoph Kabbasch; Jan Borggrefe; Lukas Goertz Journal: Neuroradiology Date: 2021-04-10 Impact factor: 2.804