Mana Moassefi1,2, Shahriar Faghani1,2, Gian Marco Conte1,2, Roman O Kowalchuk3, Sanaz Vahdati1,2, David J Crompton4, Carlos Perez-Vega5, Ricardo A Domingo Cabreja5, Sujay A Vora6, Alfredo Quiñones-Hinojosa5, Ian F Parney7, Daniel M Trifiletti4, Bradley J Erickson8,9. 1. Artificial Intelligence Laboratory, Mayo Clinic, Rochester, MN, USA. 2. Department of Radiology, Mayo Clinic, Rochester, MN, USA. 3. Department of Radiation Oncology, Mayo Clinic, Rochester, MN, USA. 4. Department of Radiation Oncology, Mayo Clinic, Jacksonville, FL, USA. 5. Department of Neurosurgery, Mayo Clinic, Jacksonville, FL, USA. 6. Department of Radiation Oncology, Mayo Clinic, Phoenix, AZ, USA. 7. Department of Neurologic Surgery, Mayo Clinic, Rochester, MN, USA. 8. Artificial Intelligence Laboratory, Mayo Clinic, Rochester, MN, USA. bje@mayo.edu. 9. Department of Radiology, Mayo Clinic, Rochester, MN, USA. bje@mayo.edu.
Abstract
INTRODUCTION: Glioblastomas (GBMs) are highly aggressive tumors. A common clinical challenge after standard of care treatment is differentiating tumor progression from treatment-related changes, also known as pseudoprogression (PsP). Usually, PsP resolves or stabilizes without further treatment or a course of steroids, whereas true progression (TP) requires more aggressive management. Differentiating PsP from TP will affect the patient's outcome. This study investigated using deep learning to distinguish PsP MRI features from progressive disease. METHOD: We included GBM patients with a new or increasingly enhancing lesion within the original radiation field. We labeled those who subsequently were stable or improved on imaging and clinically as PsP and those with clinical and imaging deterioration as TP. A subset of subjects underwent a second resection. We labeled these subjects as PsP, or TP based on the histological diagnosis. We coregistered contrast-enhanced T1 MRIs with T2-weighted images for each patient and used them as input to a 3-D Densenet121 model and using five-fold cross-validation to predict TP vs PsP. RESULT: We included 124 patients who met the criteria, and of those, 63 were PsP and 61 were TP. We trained a deep learning model that achieved 76.4% (range 70-84%, SD 5.122) mean accuracy over the 5 folds, 0.7560 (range 0.6553-0.8535, SD 0.069) mean AUROCC, 88.72% (SD 6.86) mean sensitivity, and 62.05% (SD 9.11) mean specificity. CONCLUSION: We report the development of a deep learning model that distinguishes PsP from TP in GBM patients treated per the Stupp protocol. Further refinement and external validation are required prior to widespread adoption in clinical practice.
INTRODUCTION: Glioblastomas (GBMs) are highly aggressive tumors. A common clinical challenge after standard of care treatment is differentiating tumor progression from treatment-related changes, also known as pseudoprogression (PsP). Usually, PsP resolves or stabilizes without further treatment or a course of steroids, whereas true progression (TP) requires more aggressive management. Differentiating PsP from TP will affect the patient's outcome. This study investigated using deep learning to distinguish PsP MRI features from progressive disease. METHOD: We included GBM patients with a new or increasingly enhancing lesion within the original radiation field. We labeled those who subsequently were stable or improved on imaging and clinically as PsP and those with clinical and imaging deterioration as TP. A subset of subjects underwent a second resection. We labeled these subjects as PsP, or TP based on the histological diagnosis. We coregistered contrast-enhanced T1 MRIs with T2-weighted images for each patient and used them as input to a 3-D Densenet121 model and using five-fold cross-validation to predict TP vs PsP. RESULT: We included 124 patients who met the criteria, and of those, 63 were PsP and 61 were TP. We trained a deep learning model that achieved 76.4% (range 70-84%, SD 5.122) mean accuracy over the 5 folds, 0.7560 (range 0.6553-0.8535, SD 0.069) mean AUROCC, 88.72% (SD 6.86) mean sensitivity, and 62.05% (SD 9.11) mean specificity. CONCLUSION: We report the development of a deep learning model that distinguishes PsP from TP in GBM patients treated per the Stupp protocol. Further refinement and external validation are required prior to widespread adoption in clinical practice.
Authors: Philipp Kickingereder; Fabian Isensee; Irada Tursunova; Jens Petersen; Ulf Neuberger; David Bonekamp; Gianluca Brugnara; Marianne Schell; Tobias Kessler; Martha Foltyn; Inga Harting; Felix Sahm; Marcel Prager; Martha Nowosielski; Antje Wick; Marco Nolden; Alexander Radbruch; Jürgen Debus; Heinz-Peter Schlemmer; Sabine Heiland; Michael Platten; Andreas von Deimling; Martin J van den Bent; Thierry Gorlia; Wolfgang Wick; Martin Bendszus; Klaus H Maier-Hein Journal: Lancet Oncol Date: 2019-04-02 Impact factor: 41.316