Dejan Kostyszyn1,2, Tobias Fechter1,3,4, Nico Bartl3,4, Anca L Grosu3,4, Christian Gratzke5, August Sigle5, Michael Mix6, Juri Ruf6, Thomas F Fassbender6, Selina Kiefer7, Alisa S Bettermann4, Nils H Nicolay3,4, Simon Spohn3,4, Maria U Kramer4, Peter Bronsert3,7, Hongqian Guo8, Xuefeng Qiu8, Feng Wang9, Christoph Henkenberens10, Rudolf A Werner11, Dimos Baltas1,3, Philipp T Meyer6, Thorsten Derlin11, Mengxia Chen8, Constantinos Zamboglou12,4. 1. Division of Medical Physics, Department of Radiation Oncology, Medical Center-University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany. 2. Faculty of Engineering, University of Freiburg, Freiburg, Germany. 3. German Cancer Consortium (DKTK), Partner Site Freiburg, Germany. 4. Department of Radiation Oncology, Medical Center-University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany. 5. Department of Urology, Medical Center-University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany. 6. Department of Nuclear Medicine, Medical Center-University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany. 7. Institute for Surgical Pathology, Medical Center-University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany. 8. Department of Urology, Medical School of Nanjing University, Affiliated Drum Tower Hospital, Nanjing, China. 9. Department of Nuclear Medicine, Medical School of Nanjing University, Affiliated Drum Tower Hospital, Nanjing, China. 10. Department of Radiation Oncology, Hannover Medical School, Hannover, Germany. 11. Department of Nuclear Medicine, Hannover Medical School, Hannover, Germany. 12. German Cancer Consortium (DKTK), Partner Site Freiburg, Germany constantinos.zamboglou@uniklinik-freiburg.de.
Abstract
Accurate delineation of the intraprostatic gross tumor volume (GTV) is a prerequisite for treatment approaches in patients with primary prostate cancer (PCa). Prostate-specific membrane antigen PET (PSMA PET) may outperform MRI in GTV detection. However, visual GTV delineation underlies interobserver heterogeneity and is time consuming. The aim of this study was to develop a convolutional neural network (CNN) for automated segmentation of intraprostatic tumor (GTV-CNN) in PSMA PET. Methods: The CNN (3D U-Net) was trained on the 68Ga-PSMA PET images of 152 patients from 2 different institutions, and the training labels were generated manually using a validated technique. The CNN was tested on 2 independent internal (cohort 1: 68Ga-PSMA PET, n = 18 and cohort 2: 18F-PSMA PET, n = 19) and 1 external (cohort 3: 68Ga-PSMA PET, n = 20) test datasets. Accordance between manual contours and GTV-CNN was assessed with the Dice-Sørensen coefficient (DSC). Sensitivity and specificity were calculated for the 2 internal test datasets (cohort 1: n = 18, cohort 2: n = 11) using whole-mount histology. Results: The median DSCs for cohorts 1-3 were 0.84 (range: 0.32-0.95), 0.81 (range: 0.28-0.93), and 0.83 (range: 0.32-0.93), respectively. Sensitivities and specificities for the GTV-CNN were comparable with manual expert contours: 0.98 and 0.76 (cohort 1) and 1 and 0.57 (cohort 2), respectively. Computation time was around 6 s for a standard dataset. Conclusion: The application of a CNN for automated contouring of intraprostatic GTV in 68Ga-PSMA and 18F-PSMA PET images resulted in a high concordance with expert contours and in high sensitivities and specificities in comparison with histology as a reference. This robust, accurate and fast technique may be implemented for treatment concepts in primary prostate cancer. The trained model and the study's source code are available in an open source repository. COPYRIGHT
Accurate delineation of the intraprostatic gross tumor volume (GTV) is a prerequisite for treatment approaches in patients with primary prostate cancer (PCa). Prostate-specific membrane antigen PET (PSMA PET) may outperform MRI in GTV detection. However, visual GTV delineation underlies interobserver heterogeneity and is time consuming. The aim of this study was to develop a convolutional neural network (CNN) for automated segmentation of intraprostatic tumor (GTV-CNN) in PSMA PET. Methods: The CNN (3D U-Net) was trained on the 68Ga-PSMA PET images of 152 patients from 2 different institutions, and the training labels were generated manually using a validated technique. The CNN was tested on 2 independent internal (cohort 1: 68Ga-PSMA PET, n = 18 and cohort 2: 18F-PSMA PET, n = 19) and 1 external (cohort 3: 68Ga-PSMA PET, n = 20) test datasets. Accordance between manual contours and GTV-CNN was assessed with the Dice-Sørensen coefficient (DSC). Sensitivity and specificity were calculated for the 2 internal test datasets (cohort 1: n = 18, cohort 2: n = 11) using whole-mount histology. Results: The median DSCs for cohorts 1-3 were 0.84 (range: 0.32-0.95), 0.81 (range: 0.28-0.93), and 0.83 (range: 0.32-0.93), respectively. Sensitivities and specificities for the GTV-CNN were comparable with manual expert contours: 0.98 and 0.76 (cohort 1) and 1 and 0.57 (cohort 2), respectively. Computation time was around 6 s for a standard dataset. Conclusion: The application of a CNN for automated contouring of intraprostatic GTV in 68Ga-PSMA and 18F-PSMA PET images resulted in a high concordance with expert contours and in high sensitivities and specificities in comparison with histology as a reference. This robust, accurate and fast technique may be implemented for treatment concepts in primary prostate cancer. The trained model and the study's source code are available in an open source repository. COPYRIGHT
Authors: Jeremie Calais; Amar U Kishan; Minsong Cao; Wolfgang P Fendler; Matthias Eiber; Ken Herrmann; Francesco Ceci; Robert E Reiter; Matthew B Rettig; John V Hegde; Narek Shaverdian; Chris R King; Michael L Steinberg; Johannes Czernin; Nicholas G Nickols Journal: J Nucl Med Date: 2018-04-13 Impact factor: 10.057
Authors: Constantinos Zamboglou; Thomas F Fassbender; Lina Steffan; Florian Schiller; Tobias Fechter; Montserrat Carles; Selina Kiefer; Hans C Rischke; Kathrin Reichel; Nina-Sophie Schmidt-Hegemann; Harun Ilhan; Alin F Chirindel; Guillaume Nicolas; Christoph Henkenberens; Thorsten Derlin; Peter Bronsert; Panayiotis Mavroidis; Ronald C Chen; Philipp T Meyer; Juri Ruf; Anca L Grosu Journal: Radiother Oncol Date: 2019-08-17 Impact factor: 6.280
Authors: Constantinos Zamboglou; Vanessa Drendel; Cordula A Jilg; Hans C Rischke; Teresa I Beck; Wolfgang Schultze-Seemann; Tobias Krauss; Michael Mix; Florian Schiller; Ulrich Wetterauer; Martin Werner; Mathias Langer; Michael Bock; Philipp T Meyer; Anca L Grosu Journal: Theranostics Date: 2017-01-01 Impact factor: 11.556
Authors: Constantinos Zamboglou; Montserrat Carles; Tobias Fechter; Selina Kiefer; Kathrin Reichel; Thomas F Fassbender; Peter Bronsert; Goeran Koeber; Oliver Schilling; Juri Ruf; Martin Werner; Cordula A Jilg; Dimos Baltas; Michael Mix; Anca L Grosu Journal: Theranostics Date: 2019-04-13 Impact factor: 11.556
Authors: Constantinos Zamboglou; Benedikt Thomann; Khodor Koubar; Peter Bronsert; Tobias Krauss; Hans C Rischke; Ilias Sachpazidis; Vanessa Drendel; Nasr Salman; Kathrin Reichel; Cordula A Jilg; Martin Werner; Philipp T Meyer; Michael Bock; Dimos Baltas; Anca L Grosu Journal: Radiat Oncol Date: 2018-05-02 Impact factor: 3.481
Authors: Kerstin Johnsson; Johan Brynolfsson; Hannicka Sahlstedt; Nicholas G Nickols; Matthew Rettig; Stephan Probst; Michael J Morris; Anders Bjartell; Mathias Eiber; Aseem Anand Journal: Eur J Nucl Med Mol Imaging Date: 2021-08-31 Impact factor: 10.057
Authors: Mathieu Gaudreault; David Chang; Nicholas Hardcastle; Price Jackson; Tomas Kron; Michael S Hofman; Shankar Siva Journal: Clin Transl Radiat Oncol Date: 2022-05-17
Authors: Thomas S C Ng; Xin Gao; Keyan Salari; Dimitar V Zlatev; Pedram Heidari; Sophia C Kamran Journal: Front Oncol Date: 2021-07-28 Impact factor: 6.244
Authors: Constantin Lapa; Ursula Nestle; Nathalie L Albert; Christian Baues; Ambros Beer; Andreas Buck; Volker Budach; Rebecca Bütof; Stephanie E Combs; Thorsten Derlin; Matthias Eiber; Wolfgang P Fendler; Christian Furth; Cihan Gani; Eleni Gkika; Anca-L Grosu; Christoph Henkenberens; Harun Ilhan; Steffen Löck; Simone Marnitz-Schulze; Matthias Miederer; Michael Mix; Nils H Nicolay; Maximilian Niyazi; Christoph Pöttgen; Claus M Rödel; Imke Schatka; Sarah M Schwarzenboeck; Andrei S Todica; Wolfgang Weber; Simone Wegen; Thomas Wiegel; Constantinos Zamboglou; Daniel Zips; Klaus Zöphel; Sebastian Zschaeck; Daniela Thorwarth; Esther G C Troost Journal: Strahlenther Onkol Date: 2021-07-14 Impact factor: 3.621
Authors: Simon K B Spohn; Alisa S Bettermann; Fabian Bamberg; Matthias Benndorf; Michael Mix; Nils H Nicolay; Tobias Fechter; Tobias Hölscher; Radu Grosu; Arturo Chiti; Anca L Grosu; Constantinos Zamboglou Journal: Theranostics Date: 2021-07-06 Impact factor: 11.556