Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles
Abstract Machine learning (ML) has increasingly been used to assist and expand current practices in neuropathology. However, generating large imaging datasets with quality labels is challenging in fields which demand high levels of expertise. Further complicating matters is the often seen disagreeme...
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2023-12-01
|
| Series: | Acta Neuropathologica Communications |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s40478-023-01691-x |
| _version_ | 1827399291850719232 |
|---|---|
| author | Juan C. Vizcarra Thomas M. Pearce Brittany N. Dugger Michael J. Keiser Marla Gearing John F. Crary Evan J. Kiely Meaghan Morris Bartholomew White Jonathan D. Glass Kurt Farrell David A. Gutman |
| author_facet | Juan C. Vizcarra Thomas M. Pearce Brittany N. Dugger Michael J. Keiser Marla Gearing John F. Crary Evan J. Kiely Meaghan Morris Bartholomew White Jonathan D. Glass Kurt Farrell David A. Gutman |
| author_sort | Juan C. Vizcarra |
| collection | DOAJ |
| description | Abstract Machine learning (ML) has increasingly been used to assist and expand current practices in neuropathology. However, generating large imaging datasets with quality labels is challenging in fields which demand high levels of expertise. Further complicating matters is the often seen disagreement between experts in neuropathology-related tasks, both at the case level and at a more granular level. Neurofibrillary tangles (NFTs) are a hallmark pathological feature of Alzheimer disease, and are associated with disease progression which warrants further investigation and granular quantification at a scale not currently accessible in routine human assessment. In this work, we first provide a baseline of annotator/rater agreement for the tasks of Braak NFT staging between experts and NFT detection using both experts and novices in neuropathology. We use a whole-slide-image (WSI) cohort of neuropathology cases from Emory University Hospital immunohistochemically stained for Tau. We develop a workflow for gathering annotations of the early stage formation of NFTs (Pre-NFTs) and mature intracellular (iNFTs) and show ML models can be trained to learn annotator nuances for the task of NFT detection in WSIs. We utilize a model-assisted-labeling approach and demonstrate ML models can be used to aid in labeling large datasets efficiently. We also show these models can be used to extract case-level features, which predict Braak NFT stages comparable to expert human raters, and do so at scale. This study provides a generalizable workflow for various pathology and related fields, and also provides a technique for accomplishing a high-level neuropathology task with limited human annotations. |
| first_indexed | 2024-03-08T19:43:44Z |
| format | Article |
| id | doaj.art-b2678bd5981b4644b024934250897893 |
| institution | Directory Open Access Journal |
| issn | 2051-5960 |
| language | English |
| last_indexed | 2024-03-08T19:43:44Z |
| publishDate | 2023-12-01 |
| publisher | BMC |
| record_format | Article |
| series | Acta Neuropathologica Communications |
| spelling | doaj.art-b2678bd5981b4644b0249342508978932023-12-24T12:31:54ZengBMCActa Neuropathologica Communications2051-59602023-12-0111112010.1186/s40478-023-01691-xToward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tanglesJuan C. Vizcarra0Thomas M. Pearce1Brittany N. Dugger2Michael J. Keiser3Marla Gearing4John F. Crary5Evan J. Kiely6Meaghan Morris7Bartholomew White8Jonathan D. Glass9Kurt Farrell10David A. Gutman11The Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory UniversityDepartment of Pathology, Division of Neuropathology, University of Pittsburgh Medical CenterDepartment of Pathology and Laboratory Medicine, University of California-Davis School of MedicineDepartment of Pharmaceutical Chemistry, Department of Bioengineering and Therapeutic Sciences, Institute for Neurodegenerative Diseases, Kavli Institute for Fundamental Neuroscience, and Bakar Computational Health Sciences Institute, University of CaliforniaDepartment of Neurology, Emory University School of MedicineDepartments of Pathology, Neuroscience, and Artificial Intelligence and Human Health, Icahn School of Medicine at Mount SinaiDepartment of Pathology and Laboratory Medicine, Emory University School of MedicineDepartment of Pathology, Johns Hopkins School of MedicineDepartment of Pathology, Beth Israel Deaconess Medical CenterDepartment of Neurology, Emory University School of MedicineDepartments of Pathology, Neuroscience, and Artificial Intelligence and Human Health, Icahn School of Medicine at Mount SinaiDepartment of Pathology and Laboratory Medicine, Emory University School of MedicineAbstract Machine learning (ML) has increasingly been used to assist and expand current practices in neuropathology. However, generating large imaging datasets with quality labels is challenging in fields which demand high levels of expertise. Further complicating matters is the often seen disagreement between experts in neuropathology-related tasks, both at the case level and at a more granular level. Neurofibrillary tangles (NFTs) are a hallmark pathological feature of Alzheimer disease, and are associated with disease progression which warrants further investigation and granular quantification at a scale not currently accessible in routine human assessment. In this work, we first provide a baseline of annotator/rater agreement for the tasks of Braak NFT staging between experts and NFT detection using both experts and novices in neuropathology. We use a whole-slide-image (WSI) cohort of neuropathology cases from Emory University Hospital immunohistochemically stained for Tau. We develop a workflow for gathering annotations of the early stage formation of NFTs (Pre-NFTs) and mature intracellular (iNFTs) and show ML models can be trained to learn annotator nuances for the task of NFT detection in WSIs. We utilize a model-assisted-labeling approach and demonstrate ML models can be used to aid in labeling large datasets efficiently. We also show these models can be used to extract case-level features, which predict Braak NFT stages comparable to expert human raters, and do so at scale. This study provides a generalizable workflow for various pathology and related fields, and also provides a technique for accomplishing a high-level neuropathology task with limited human annotations.https://doi.org/10.1186/s40478-023-01691-xNeuropathologyMachine learningModel-assisted-labelingAlzheimer’s diseaseNeurofibrillary tanglesBraak NFT staging |
| spellingShingle | Juan C. Vizcarra Thomas M. Pearce Brittany N. Dugger Michael J. Keiser Marla Gearing John F. Crary Evan J. Kiely Meaghan Morris Bartholomew White Jonathan D. Glass Kurt Farrell David A. Gutman Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles Acta Neuropathologica Communications Neuropathology Machine learning Model-assisted-labeling Alzheimer’s disease Neurofibrillary tangles Braak NFT staging |
| title | Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles |
| title_full | Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles |
| title_fullStr | Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles |
| title_full_unstemmed | Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles |
| title_short | Toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles |
| title_sort | toward a generalizable machine learning workflow for neurodegenerative disease staging with focus on neurofibrillary tangles |
| topic | Neuropathology Machine learning Model-assisted-labeling Alzheimer’s disease Neurofibrillary tangles Braak NFT staging |
| url | https://doi.org/10.1186/s40478-023-01691-x |
| work_keys_str_mv | AT juancvizcarra towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT thomasmpearce towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT brittanyndugger towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT michaeljkeiser towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT marlagearing towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT johnfcrary towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT evanjkiely towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT meaghanmorris towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT bartholomewwhite towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT jonathandglass towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT kurtfarrell towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles AT davidagutman towardageneralizablemachinelearningworkflowforneurodegenerativediseasestagingwithfocusonneurofibrillarytangles |