TY - JOUR
T1 - Artificial intelligence in neuropathology
T2 - deep learning-based assessment of tauopathy
AU - Signaevsky, Maxim
AU - Prastawa, Marcel
AU - Farrell, Kurt
AU - Tabish, Nabil
AU - Baldwin, Elena
AU - Han, Natalia
AU - Iida, Megan A.
AU - Koll, John
AU - Bryce, Clare
AU - Purohit, Dushyant
AU - Haroutunian, Vahram
AU - McKee, Ann C.
AU - Stein, Thor D.
AU - White, Charles L.
AU - Walker, Jamie
AU - Richardson, Timothy E.
AU - Hanson, Russell
AU - Donovan, Michael J.
AU - Cordon-Cardo, Carlos
AU - Zeineh, Jack
AU - Fernandez, Gerardo
AU - Crary, John F.
N1 - Funding Information:
Acknowledgments The authors would like to acknowledge NIH grants R01AG054008 (JFC), R01NS095252 (JFC), R01AG062348 (ACM/JFC), RF1AG060961 (JFC), F32AG056098 (KF), Department of Defense W81XWH-13-MRPRA-CSRA, the Tau Consortium (Rainwater Charitable Trust), and the Alzheimer’s Association (NIRG-15-363188). The first author was also supported by a Career Development Award funded by NIH-NOA 3P50AG005138 (MS). We thank Jill Gregory for the illustration. The authors also would like to thank Ping Shang, HT, Jeff Harris, HTL, and Chan Foong, MS, for technical assistance, and Javed and Shahnaz Iqbal Family Trust for the generous donation.
Publisher Copyright:
© 2019, United States & Canadian Academy of Pathology.
PY - 2019/7/1
Y1 - 2019/7/1
N2 - Accumulation of abnormal tau in neurofibrillary tangles (NFT) occurs in Alzheimer disease (AD) and a spectrum of tauopathies. These tauopathies have diverse and overlapping morphological phenotypes that obscure classification and quantitative assessments. Recently, powerful machine learning-based approaches have emerged, allowing the recognition and quantification of pathological changes from digital images. Here, we applied deep learning to the neuropathological assessment of NFT in postmortem human brain tissue to develop a classifier capable of recognizing and quantifying tau burden. The histopathological material was derived from 22 autopsy brains from patients with tauopathies. We used a custom web-based informatics platform integrated with an in-house information management system to manage whole slide images (WSI) and human expert annotations as ground truth. We utilized fully annotated regions to train a deep learning fully convolutional neural network (FCN) implemented in PyTorch against the human expert annotations. We found that the deep learning framework is capable of identifying and quantifying NFT with a range of staining intensities and diverse morphologies. With our FCN model, we achieved high precision and recall in naive WSI semantic segmentation, correctly identifying tangle objects using a SegNet model trained for 200 epochs. Our FCN is efficient and well suited for the practical application of WSIs with average processing times of 45 min per WSI per GPU, enabling reliable and reproducible large-scale detection of tangles. We measured performance on test data of 50 pre-annotated regions on eight naive WSI across various tauopathies, resulting in the recall, precision, and an F1 score of 0.92, 0.72, and 0.81, respectively. Machine learning is a useful tool for complex pathological assessment of AD and other tauopathies. Using deep learning classifiers, we have the potential to integrate cell- and region-specific annotations with clinical, genetic, and molecular data, providing unbiased data for clinicopathological correlations that will enhance our knowledge of the neurodegeneration.
AB - Accumulation of abnormal tau in neurofibrillary tangles (NFT) occurs in Alzheimer disease (AD) and a spectrum of tauopathies. These tauopathies have diverse and overlapping morphological phenotypes that obscure classification and quantitative assessments. Recently, powerful machine learning-based approaches have emerged, allowing the recognition and quantification of pathological changes from digital images. Here, we applied deep learning to the neuropathological assessment of NFT in postmortem human brain tissue to develop a classifier capable of recognizing and quantifying tau burden. The histopathological material was derived from 22 autopsy brains from patients with tauopathies. We used a custom web-based informatics platform integrated with an in-house information management system to manage whole slide images (WSI) and human expert annotations as ground truth. We utilized fully annotated regions to train a deep learning fully convolutional neural network (FCN) implemented in PyTorch against the human expert annotations. We found that the deep learning framework is capable of identifying and quantifying NFT with a range of staining intensities and diverse morphologies. With our FCN model, we achieved high precision and recall in naive WSI semantic segmentation, correctly identifying tangle objects using a SegNet model trained for 200 epochs. Our FCN is efficient and well suited for the practical application of WSIs with average processing times of 45 min per WSI per GPU, enabling reliable and reproducible large-scale detection of tangles. We measured performance on test data of 50 pre-annotated regions on eight naive WSI across various tauopathies, resulting in the recall, precision, and an F1 score of 0.92, 0.72, and 0.81, respectively. Machine learning is a useful tool for complex pathological assessment of AD and other tauopathies. Using deep learning classifiers, we have the potential to integrate cell- and region-specific annotations with clinical, genetic, and molecular data, providing unbiased data for clinicopathological correlations that will enhance our knowledge of the neurodegeneration.
UR - http://www.scopus.com/inward/record.url?scp=85061718942&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85061718942&partnerID=8YFLogxK
U2 - 10.1038/s41374-019-0202-4
DO - 10.1038/s41374-019-0202-4
M3 - Article
C2 - 30770886
AN - SCOPUS:85061718942
SN - 0023-6837
VL - 99
SP - 1019
EP - 1029
JO - Laboratory Investigation
JF - Laboratory Investigation
IS - 7
ER -