Zhilong Yi1,2, Siqi Hu2, Xiaofeng Lin1, Qiong Zou2, MinHong Zou3, Zhanlei Zhang4, Lei Xu2, Ningyi Jiang5,6, Yong Zhang7. 1. Department of Nuclear Medicine, The Seventh Affiliated Hospital, Sun Yat-Sen University, Shenzhen, China. 2. Department of Nuclear Medicine, The Third Affiliated Hospital of Sun Yat-Sen University, Guangzhou, Guangdong, China. 3. Department of Ultrasound, The Third Affiliated Hospital of Sun Yat-Sen University, Guangzhou, Guangdong, China. 4. Department of Nuclear Medicine, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, Guangzhou, Guangdong, China. 5. Department of Nuclear Medicine, The Seventh Affiliated Hospital, Sun Yat-Sen University, Shenzhen, China. jiangny@mail.sysu.edu.cn. 6. Department of Nuclear Medicine, Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, Guangzhou, Guangdong, China. jiangny@mail.sysu.edu.cn. 7. Department of Nuclear Medicine, The Third Affiliated Hospital of Sun Yat-Sen University, Guangzhou, Guangdong, China. zhangyn9@mail.sysu.edu.cn.
Abstract
PURPOSE: 68 Ga-PSMA PET/CT has high specificity and sensitivity for the detection of both intraprostatic tumor focal lesions and metastasis. However, approximately 10% of primary prostate cancer are invisible on PSMA-PET (exhibit no or minimal uptake). In this work, we investigated whether machine learning-based radiomics models derived from PSMA-PET images could predict invisible intraprostatic lesions on 68 Ga-PSMA-11 PET in patients with primary prostate cancer. METHODS: In this retrospective study, patients with or without prostate cancer who underwent 68 Ga-PSMA PET/CT and presented negative on PSMA-PET image at either of two different institutions were included: institution 1 (between 2017 and 2020) for the training set and institution 2 (between 2019 and 2020) for the external test set. Three random forest (RF) models were built using selected features extracted from standard PET images, delayed PET images, and both standard and delayed PET images. Then, subsequent tenfold cross-validation was performed. In the test phase, the three RF models and PSA density (PSAD, cut-off value: 0.15 ng/ml/ml) were tested with the external test set. The area under the receiver operating characteristic curve (AUC) was calculated for the models and PSAD. The AUCs of the radiomics model and PSAD were compared. RESULTS: A total of 64 patients (39 with prostate cancer and 25 with benign prostate disease) were in the training set, and 36 (21 with prostate cancer and 15 with benign prostate disease) were in the test set. The average AUCs of the three RF models from tenfold cross-validation were 0.87 (95% CI: 0.72, 1.00), 0.86 (95% CI: 0.63, 1.00), and 0.91 (95% CI: 0.69, 1.00), respectively. In the test set, the AUCs of the three trained RF models and PSAD were 0.903 (95% CI: 0.830, 0.975), 0.856 (95% CI: 0.748, 0.964), 0.925 (95% CI:0.838, 1.00), and 0.662 (95% CI: 0.510, 0.813). The AUCs of the three radiomics models were higher than that of PSAD (0.903, 0.856, and 0.925 vs. 0.662, respectively; P = .007, P = .045, and P = .005, respectively). CONCLUSION: Random forest models developed by 68 Ga-PSMA-11 PET-based radiomics features were proven useful for accurate prediction of invisible intraprostatic lesion on 68 Ga-PSMA-11 PET in patients with primary prostate cancer and showed better diagnostic performance compared with PSAD.
PURPOSE: 68 Ga-PSMA PET/CT has high specificity and sensitivity for the detection of both intraprostatic tumor focal lesions and metastasis. However, approximately 10% of primary prostate cancer are invisible on PSMA-PET (exhibit no or minimal uptake). In this work, we investigated whether machine learning-based radiomics models derived from PSMA-PET images could predict invisible intraprostatic lesions on 68 Ga-PSMA-11 PET in patients with primary prostate cancer. METHODS: In this retrospective study, patients with or without prostate cancer who underwent 68 Ga-PSMA PET/CT and presented negative on PSMA-PET image at either of two different institutions were included: institution 1 (between 2017 and 2020) for the training set and institution 2 (between 2019 and 2020) for the external test set. Three random forest (RF) models were built using selected features extracted from standard PET images, delayed PET images, and both standard and delayed PET images. Then, subsequent tenfold cross-validation was performed. In the test phase, the three RF models and PSA density (PSAD, cut-off value: 0.15 ng/ml/ml) were tested with the external test set. The area under the receiver operating characteristic curve (AUC) was calculated for the models and PSAD. The AUCs of the radiomics model and PSAD were compared. RESULTS: A total of 64 patients (39 with prostate cancer and 25 with benign prostate disease) were in the training set, and 36 (21 with prostate cancer and 15 with benign prostate disease) were in the test set. The average AUCs of the three RF models from tenfold cross-validation were 0.87 (95% CI: 0.72, 1.00), 0.86 (95% CI: 0.63, 1.00), and 0.91 (95% CI: 0.69, 1.00), respectively. In the test set, the AUCs of the three trained RF models and PSAD were 0.903 (95% CI: 0.830, 0.975), 0.856 (95% CI: 0.748, 0.964), 0.925 (95% CI:0.838, 1.00), and 0.662 (95% CI: 0.510, 0.813). The AUCs of the three radiomics models were higher than that of PSAD (0.903, 0.856, and 0.925 vs. 0.662, respectively; P = .007, P = .045, and P = .005, respectively). CONCLUSION: Random forest models developed by 68 Ga-PSMA-11 PET-based radiomics features were proven useful for accurate prediction of invisible intraprostatic lesion on 68 Ga-PSMA-11 PET in patients with primary prostate cancer and showed better diagnostic performance compared with PSAD.
Authors: Jesse D Le; Nelly Tan; Eugene Shkolyar; David Y Lu; Lorna Kwan; Leonard S Marks; Jiaoti Huang; Daniel J A Margolis; Steven S Raman; Robert E Reiter Journal: Eur Urol Date: 2014-09-23 Impact factor: 20.096
Authors: Martijn G Schouten; Marloes van der Leest; Morgan Pokorny; Martijn Hoogenboom; Jelle O Barentsz; Les C Thompson; Jurgen J Fütterer Journal: Eur Urol Date: 2017-01-04 Impact factor: 20.096
Authors: Osamu Ukimura; Jonathan A Coleman; Alex de la Taille; Mark Emberton; Jonathan I Epstein; Stephen J Freedland; Gianluca Giannarini; Adam S Kibel; Rodolfo Montironi; Guillaume Ploussard; Monique J Roobol; Vincenzo Scattoni; J Stephen Jones Journal: Eur Urol Date: 2012-09-25 Impact factor: 20.096
Authors: Samuel Borofsky; Arvin K George; Sonia Gaur; Marcelino Bernardo; Matthew D Greer; Francesca V Mertan; Myles Taffel; Vanesa Moreno; Maria J Merino; Bradford J Wood; Peter A Pinto; Peter L Choyke; Baris Turkbey Journal: Radiology Date: 2017-10-20 Impact factor: 11.105
Authors: Peter Donato; Matthew J Roberts; Andrew Morton; Samuel Kyle; Geoff Coughlin; Rachel Esler; Nigel Dunglison; Robert A Gardiner; John Yaxley Journal: Eur J Nucl Med Mol Imaging Date: 2018-09-19 Impact factor: 9.236
Authors: Wolfgang P Fendler; Jeremie Calais; Matthias Eiber; Robert R Flavell; Ashley Mishoe; Felix Y Feng; Hao G Nguyen; Robert E Reiter; Matthew B Rettig; Shozo Okamoto; Louise Emmett; Helle D Zacho; Harun Ilhan; Axel Wetter; Christoph Rischpler; Heiko Schoder; Irene A Burger; Jeannine Gartmann; Raven Smith; Eric J Small; Roger Slavik; Peter R Carroll; Ken Herrmann; Johannes Czernin; Thomas A Hope Journal: JAMA Oncol Date: 2019-06-01 Impact factor: 31.777
Authors: Hashim U Ahmed; Ahmed El-Shater Bosaily; Louise C Brown; Rhian Gabe; Richard Kaplan; Mahesh K Parmar; Yolanda Collaco-Moraes; Katie Ward; Richard G Hindley; Alex Freeman; Alex P Kirkham; Robert Oldroyd; Chris Parker; Mark Emberton Journal: Lancet Date: 2017-01-20 Impact factor: 79.321
Authors: Indrani Bhattacharya; Yash S Khandwala; Sulaiman Vesal; Wei Shao; Qianye Yang; Simon J C Soerensen; Richard E Fan; Pejman Ghanouni; Christian A Kunder; James D Brooks; Yipeng Hu; Mirabela Rusu; Geoffrey A Sonn Journal: Ther Adv Urol Date: 2022-10-10