B Sohn1, K-Y Park2, J Choi1,3,4, J H Koo1, K Han1, B Joo1, S Y Won1, J Cha1, H S Choi5,6, S-K Lee1. 1. From the Department of Radiology (B.S., J.C., J.H.K., K.H., B.J., S.Y.W., J.C., H.S.C., S.-K.L.), Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Yonsei University College of Medicine, Seoul, South Korea. 2. Department of Neurosurgery (K.-Y.P.), Yonsei University College of Medicine, Seoul, South Korea. 3. Department of Neurology (J.C.), Yonsei University College of Medicine, Seoul, South Korea. 4. Department of Neurology (J.C.), Seoul Medical Center, Seoul, South Korea. 5. From the Department of Radiology (B.S., J.C., J.H.K., K.H., B.J., S.Y.W., J.C., H.S.C., S.-K.L.), Research Institute of Radiological Science and Center for Clinical Imaging Data Science, Yonsei University College of Medicine, Seoul, South Korea chlgustjr1@gmail.com. 6. Department of Radiology (H.S.C.), Seoul Medical Center, Seoul, South Korea.
Abstract
BACKGROUND AND PURPOSE: The detection of cerebral aneurysms on MRA is a challenging task. Recent studies have used deep learning-based software for automated detection of aneurysms on MRA and have reported high performance. The purpose of this study was to evaluate the incremental value of using deep learning-based software for the detection of aneurysms on MRA by 2 radiologists, a neurosurgeon, and a neurologist. MATERIALS AND METHODS: TOF-MRA examinations of intracranial aneurysms were retrospectively extracted. Four physicians interpreted the MRA blindly. After a washout period, they interpreted MRA again using the software. Sensitivity and specificity per patient, sensitivity per lesion, and the number of false-positives per case were measured. Diagnostic performances, including subgroup analysis of lesions, were compared. Logistic regression with a generalized estimating equation was used. RESULTS: A total of 332 patients were evaluated; 135 patients had positive findings with 169 lesions. With software assistance, patient-based sensitivity was statistically improved after the washout period (73.5% versus 86.5%, P < .001). The neurosurgeon and neurologist showed a significant increase in patient-based sensitivity with software assistance (74.8% versus 85.2%, P = .03, and 56.3% versus 84.4%, P < .001, respectively), while the number of false-positive cases did not increase significantly (23 versus 30, P = .20, and 22 versus 24, P = .75, respectively). CONCLUSIONS: Software-aided reading showed significant incremental value in the sensitivity of clinicians in the detection of aneurysms on MRA without a significant increase in false-positive findings, especially for the neurosurgeon and neurologist. Software-aided reading showed equivocal value for the radiologist.
BACKGROUND AND PURPOSE: The detection of cerebral aneurysms on MRA is a challenging task. Recent studies have used deep learning-based software for automated detection of aneurysms on MRA and have reported high performance. The purpose of this study was to evaluate the incremental value of using deep learning-based software for the detection of aneurysms on MRA by 2 radiologists, a neurosurgeon, and a neurologist. MATERIALS AND METHODS: TOF-MRA examinations of intracranial aneurysms were retrospectively extracted. Four physicians interpreted the MRA blindly. After a washout period, they interpreted MRA again using the software. Sensitivity and specificity per patient, sensitivity per lesion, and the number of false-positives per case were measured. Diagnostic performances, including subgroup analysis of lesions, were compared. Logistic regression with a generalized estimating equation was used. RESULTS: A total of 332 patients were evaluated; 135 patients had positive findings with 169 lesions. With software assistance, patient-based sensitivity was statistically improved after the washout period (73.5% versus 86.5%, P < .001). The neurosurgeon and neurologist showed a significant increase in patient-based sensitivity with software assistance (74.8% versus 85.2%, P = .03, and 56.3% versus 84.4%, P < .001, respectively), while the number of false-positive cases did not increase significantly (23 versus 30, P = .20, and 22 versus 24, P = .75, respectively). CONCLUSIONS: Software-aided reading showed significant incremental value in the sensitivity of clinicians in the detection of aneurysms on MRA without a significant increase in false-positive findings, especially for the neurosurgeon and neurologist. Software-aided reading showed equivocal value for the radiologist.
Authors: Robert J McDonald; Kara M Schwartz; Laurence J Eckel; Felix E Diehn; Christopher H Hunt; Brian J Bartholmai; Bradley J Erickson; David F Kallmes Journal: Acad Radiol Date: 2015-07-22 Impact factor: 3.173
Authors: S Miki; N Hayashi; Y Masutani; Y Nomura; T Yoshikawa; S Hanaoka; M Nemoto; K Ohtomo Journal: AJNR Am J Neuroradiol Date: 2016-02-18 Impact factor: 3.825
Authors: Allison Park; Chris Chute; Pranav Rajpurkar; Joe Lou; Robyn L Ball; Katie Shpanskaya; Rashad Jabarkheel; Lily H Kim; Emily McKenna; Joe Tseng; Jason Ni; Fidaa Wishah; Fred Wittber; David S Hong; Thomas J Wilson; Safwan Halabi; Sanjay Basu; Bhavik N Patel; Matthew P Lungren; Andrew Y Ng; Kristen W Yeom Journal: JAMA Netw Open Date: 2019-06-05