Abstract
Purpose: To develop and validate a deep learning algorithm that predicts the final diagnosis of Alzheimer disease (AD), mild cognitive impairment, or neither at fluorine 18 ( 18 F) fluorodeoxyglucose (FDG) PET of the brain and compare its performance to that of radiologic readers. Materials and Methods: Prospective 18 F-FDG PET brain images from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (2109 imaging studies from 2005 to 2017, 1002 patients) and retrospective independent test set (40 imaging studies from 2006 to 2016, 40 patients) were collected. Final clinical diagnosis at follow-up was recorded. Convolutional neural network of InceptionV3 architecture was trained on 90% of ADNI data set and tested on the remaining 10%, as well as the independent test set, with performance compared to radiologic readers. Model was analyzed with sensitivity, specificity, receiver operating characteristic (ROC), saliency map, and t-distributed stochastic neighbor embedding. Results: The algorithm achieved area under the ROC curve of 0.98 (95% confidence interval: 0.94, 1.00) when evaluated on predicting the final clinical diagnosis of AD in the independent test set (82% specificity at 100% sensitivity), an average of 75.8 months prior to the final diagnosis, which in ROC space outperformed reader performance (57% [four of seven] sensitivity, 91% [30 of 33] specificity; P , .05). Saliency map demonstrated attention to known areas of interest but with focus on the entire brain. Conclusion: By using fluorine 18 fluorodeoxyglucose PET of the brain, a deep learning algorithm developed for early prediction of Alzheimer disease achieved 82% specificity at 100% sensitivity, an average of 75.8 months prior to the final diagnosis.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 456-464 |
| Number of pages | 9 |
| Journal | RADIOLOGY |
| Volume | 290 |
| Issue number | 3 |
| DOIs | |
| State | Published - Mar 1 2019 |
Fingerprint
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging
Cite this
A deep learning model to predict a diagnosis of Alzheimer disease by using 18 F-FDG PET of the brain . / Ding, Yiming; Sohn, Jae Ho; Kawczynski, Michael G.; Trivedi, Hari; Harnish, Roy; Jenkins, Nathaniel W.; Lituiev, Dmytro; Copeland, Timothy P.; Aboian, Mariam S.; Aparici, Carina Mari; Behr, Spencer C.; Flavell, Robert R.; Huang, Shih Ying; Zalocusky, Kelly A.; Nardo, Lorenzo; Seo, Youngho; Hawkins, Randall A.; Pampaloni, Miguel Hernandez; Hadley, Dexter; Franc, Benjamin L.
In: RADIOLOGY, Vol. 290, No. 3, 01.03.2019, p. 456-464.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - A deep learning model to predict a diagnosis of Alzheimer disease by using 18 F-FDG PET of the brain
AU - Ding, Yiming
AU - Sohn, Jae Ho
AU - Kawczynski, Michael G.
AU - Trivedi, Hari
AU - Harnish, Roy
AU - Jenkins, Nathaniel W.
AU - Lituiev, Dmytro
AU - Copeland, Timothy P.
AU - Aboian, Mariam S.
AU - Aparici, Carina Mari
AU - Behr, Spencer C.
AU - Flavell, Robert R.
AU - Huang, Shih Ying
AU - Zalocusky, Kelly A.
AU - Nardo, Lorenzo
AU - Seo, Youngho
AU - Hawkins, Randall A.
AU - Pampaloni, Miguel Hernandez
AU - Hadley, Dexter
AU - Franc, Benjamin L.
PY - 2019/3/1
Y1 - 2019/3/1
N2 - Purpose: To develop and validate a deep learning algorithm that predicts the final diagnosis of Alzheimer disease (AD), mild cognitive impairment, or neither at fluorine 18 ( 18 F) fluorodeoxyglucose (FDG) PET of the brain and compare its performance to that of radiologic readers. Materials and Methods: Prospective 18 F-FDG PET brain images from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (2109 imaging studies from 2005 to 2017, 1002 patients) and retrospective independent test set (40 imaging studies from 2006 to 2016, 40 patients) were collected. Final clinical diagnosis at follow-up was recorded. Convolutional neural network of InceptionV3 architecture was trained on 90% of ADNI data set and tested on the remaining 10%, as well as the independent test set, with performance compared to radiologic readers. Model was analyzed with sensitivity, specificity, receiver operating characteristic (ROC), saliency map, and t-distributed stochastic neighbor embedding. Results: The algorithm achieved area under the ROC curve of 0.98 (95% confidence interval: 0.94, 1.00) when evaluated on predicting the final clinical diagnosis of AD in the independent test set (82% specificity at 100% sensitivity), an average of 75.8 months prior to the final diagnosis, which in ROC space outperformed reader performance (57% [four of seven] sensitivity, 91% [30 of 33] specificity; P , .05). Saliency map demonstrated attention to known areas of interest but with focus on the entire brain. Conclusion: By using fluorine 18 fluorodeoxyglucose PET of the brain, a deep learning algorithm developed for early prediction of Alzheimer disease achieved 82% specificity at 100% sensitivity, an average of 75.8 months prior to the final diagnosis.
AB - Purpose: To develop and validate a deep learning algorithm that predicts the final diagnosis of Alzheimer disease (AD), mild cognitive impairment, or neither at fluorine 18 ( 18 F) fluorodeoxyglucose (FDG) PET of the brain and compare its performance to that of radiologic readers. Materials and Methods: Prospective 18 F-FDG PET brain images from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (2109 imaging studies from 2005 to 2017, 1002 patients) and retrospective independent test set (40 imaging studies from 2006 to 2016, 40 patients) were collected. Final clinical diagnosis at follow-up was recorded. Convolutional neural network of InceptionV3 architecture was trained on 90% of ADNI data set and tested on the remaining 10%, as well as the independent test set, with performance compared to radiologic readers. Model was analyzed with sensitivity, specificity, receiver operating characteristic (ROC), saliency map, and t-distributed stochastic neighbor embedding. Results: The algorithm achieved area under the ROC curve of 0.98 (95% confidence interval: 0.94, 1.00) when evaluated on predicting the final clinical diagnosis of AD in the independent test set (82% specificity at 100% sensitivity), an average of 75.8 months prior to the final diagnosis, which in ROC space outperformed reader performance (57% [four of seven] sensitivity, 91% [30 of 33] specificity; P , .05). Saliency map demonstrated attention to known areas of interest but with focus on the entire brain. Conclusion: By using fluorine 18 fluorodeoxyglucose PET of the brain, a deep learning algorithm developed for early prediction of Alzheimer disease achieved 82% specificity at 100% sensitivity, an average of 75.8 months prior to the final diagnosis.
UR - http://www.scopus.com/inward/record.url?scp=85060372984&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85060372984&partnerID=8YFLogxK
U2 - 10.1148/radiol.2018180958
DO - 10.1148/radiol.2018180958
M3 - Article
VL - 290
SP - 456
EP - 464
JO - Radiology
JF - Radiology
SN - 0033-8419
IS - 3
ER -