An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease

Wenjie Kang; Bo Li; Janne M. Papma; Lize C. Jiskoot; Peter Paul De Deyn; Geert Jan Biessels; Jurgen A.H.R. Claassen; Huub A.M. Middelkoop; Wiesje M.van der Flier; Inez H.G.B. Ramakers; Stefan Klein; Esther E. Bron; Parelsnoer Neurodegenerative Diseases study group

doi:10.1007/978-3-031-47401-9_7

An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease

Wenjie Kang^*, Bo Li, Janne M. Papma, Lize C. Jiskoot, Peter Paul De Deyn, Geert Jan Biessels, Jurgen A.H.R. Claassen, Huub A.M. Middelkoop, Wiesje M.van der Flier, Inez H.G.B. Ramakers, Stefan Klein, Esther E. Bron, Parelsnoer Neurodegenerative Diseases study group

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

Abstract

Machine learning methods have shown large potential for the automatic early diagnosis of Alzheimer’s Disease (AD). However, some machine learning methods based on imaging data have poor interpretability because it is usually unclear how they make their decisions. Explainable Boosting Machines (EBMs) are interpretable machine learning models based on the statistical framework of generalized additive modeling, but have so far only been used for tabular data. Therefore, we propose a framework that combines the strength of EBM with high-dimensional imaging data using deep learning-based feature extraction. The proposed framework is interpretable because it provides the importance of each feature. We validated the proposed framework on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset, achieving accuracy of 0.883 and area-under-the-curve (AUC) of 0.970 on AD and control classification. Furthermore, we validated the proposed framework on an external testing set, achieving accuracy of 0.778 and AUC of 0.887 on AD and subjective cognitive decline (SCD) classification. The proposed framework significantly outperformed an EBM model using volume biomarkers instead of deep learning-based features, as well as an end-to-end convolutional neural network (CNN) with optimized architecture.

Original language	English
Title of host publication	Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings
Editors	M. Emre Celebi, Md Sirajus Salekin, Hyunwoo Kim, Shadi Albarqouni
Publisher	Springer Verlag
Pages	69-78
Number of pages	10
Volume	14393
ISBN (Print)	9783031474002
DOIs	https://doi.org/10.1007/978-3-031-47401-9_7
Publication status	Published - 1 Jan 2023
Event	26th International Conference on Medical Image Computing and Computer Assisted Intervention - Vancouver, Canada Duration: 8 Oct 2023 → 12 Oct 2023 Conference number: 26 https://conferences.miccai.org/2023/en/

Publication series

Series	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14393
ISSN	0302-9743

Conference

Conference	26th International Conference on Medical Image Computing and Computer Assisted Intervention
Abbreviated title	MICCAI 2023
Country/Territory	Canada
City	Vancouver
Period	8/10/23 → 12/10/23
Internet address	https://conferences.miccai.org/2023/en/

Keywords

Alzheimer’s disease
Convolutional neural network
Explainable boosting machine
Interpretable AI
MRI

Access to Document

10.1007/978-3-031-47401-9_7

Cite this

Kang, W., Li, B., Papma, J. M., Jiskoot, L. C., Deyn, P. P. D., Biessels, G. J., Claassen, J. A. H. R., Middelkoop, H. A. M., Flier, W. M. V. D., Ramakers, I. H. G. B., Klein, S., Bron, E. E., & Parelsnoer Neurodegenerative Diseases study group (2023). An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease. In M. E. Celebi, M. S. Salekin, H. Kim, & S. Albarqouni (Eds.), Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings (Vol. 14393, pp. 69-78). Article 305659 Springer Verlag. https://doi.org/10.1007/978-3-031-47401-9_7

Kang, Wenjie ; Li, Bo ; Papma, Janne M. et al. / An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease. Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings. editor / M. Emre Celebi ; Md Sirajus Salekin ; Hyunwoo Kim ; Shadi Albarqouni. Vol. 14393 Springer Verlag, 2023. pp. 69-78 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 14393).

@inproceedings{98ceff79b7e74f1c86143cceaa31685a,

title = "An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer{\textquoteright}s Disease",

abstract = "Machine learning methods have shown large potential for the automatic early diagnosis of Alzheimer{\textquoteright}s Disease (AD). However, some machine learning methods based on imaging data have poor interpretability because it is usually unclear how they make their decisions. Explainable Boosting Machines (EBMs) are interpretable machine learning models based on the statistical framework of generalized additive modeling, but have so far only been used for tabular data. Therefore, we propose a framework that combines the strength of EBM with high-dimensional imaging data using deep learning-based feature extraction. The proposed framework is interpretable because it provides the importance of each feature. We validated the proposed framework on the Alzheimer{\textquoteright}s Disease Neuroimaging Initiative (ADNI) dataset, achieving accuracy of 0.883 and area-under-the-curve (AUC) of 0.970 on AD and control classification. Furthermore, we validated the proposed framework on an external testing set, achieving accuracy of 0.778 and AUC of 0.887 on AD and subjective cognitive decline (SCD) classification. The proposed framework significantly outperformed an EBM model using volume biomarkers instead of deep learning-based features, as well as an end-to-end convolutional neural network (CNN) with optimized architecture.",

keywords = "Alzheimer{\textquoteright}s disease, Convolutional neural network, Explainable boosting machine, Interpretable AI, MRI",

author = "Wenjie Kang and Bo Li and Papma, {Janne M.} and Jiskoot, {Lize C.} and Deyn, {Peter Paul De} and Biessels, {Geert Jan} and Claassen, {Jurgen A.H.R.} and Middelkoop, {Huub A.M.} and Flier, {Wiesje M.van der} and Ramakers, {Inez H.G.B.} and Stefan Klein and Bron, {Esther E.} and {Parelsnoer Neurodegenerative Diseases study group}",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.; 26th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2023 ; Conference date: 08-10-2023 Through 12-10-2023",

year = "2023",

month = jan,

day = "1",

doi = "10.1007/978-3-031-47401-9_7",

language = "English",

isbn = "9783031474002",

volume = "14393",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "69--78",

editor = "Celebi, {M. Emre} and Salekin, {Md Sirajus} and Hyunwoo Kim and Shadi Albarqouni",

booktitle = "Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings",

address = "Germany",

url = "https://conferences.miccai.org/2023/en/",

}

Kang, W, Li, B, Papma, JM, Jiskoot, LC, Deyn, PPD, Biessels, GJ, Claassen, JAHR, Middelkoop, HAM, Flier, WMVD, Ramakers, IHGB, Klein, S, Bron, EE & Parelsnoer Neurodegenerative Diseases study group 2023, An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease. in ME Celebi, MS Salekin, H Kim & S Albarqouni (eds), Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings. vol. 14393, 305659, Springer Verlag, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14393, pp. 69-78, 26th International Conference on Medical Image Computing and Computer Assisted Intervention, Vancouver, British Columbia, Canada, 8/10/23. https://doi.org/10.1007/978-3-031-47401-9_7

An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease. / Kang, Wenjie; Li, Bo; Papma, Janne M. et al.
Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings. ed. / M. Emre Celebi; Md Sirajus Salekin; Hyunwoo Kim; Shadi Albarqouni. Vol. 14393 Springer Verlag, 2023. p. 69-78 305659 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 14393).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

TY - GEN

T1 - An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease

AU - Kang, Wenjie

AU - Li, Bo

AU - Papma, Janne M.

AU - Jiskoot, Lize C.

AU - Deyn, Peter Paul De

AU - Biessels, Geert Jan

AU - Claassen, Jurgen A.H.R.

AU - Middelkoop, Huub A.M.

AU - Flier, Wiesje M.van der

AU - Ramakers, Inez H.G.B.

AU - Klein, Stefan

AU - Bron, Esther E.

AU - Parelsnoer Neurodegenerative Diseases study group

N1 - Conference code: 26

PY - 2023/1/1

Y1 - 2023/1/1

N2 - Machine learning methods have shown large potential for the automatic early diagnosis of Alzheimer’s Disease (AD). However, some machine learning methods based on imaging data have poor interpretability because it is usually unclear how they make their decisions. Explainable Boosting Machines (EBMs) are interpretable machine learning models based on the statistical framework of generalized additive modeling, but have so far only been used for tabular data. Therefore, we propose a framework that combines the strength of EBM with high-dimensional imaging data using deep learning-based feature extraction. The proposed framework is interpretable because it provides the importance of each feature. We validated the proposed framework on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset, achieving accuracy of 0.883 and area-under-the-curve (AUC) of 0.970 on AD and control classification. Furthermore, we validated the proposed framework on an external testing set, achieving accuracy of 0.778 and AUC of 0.887 on AD and subjective cognitive decline (SCD) classification. The proposed framework significantly outperformed an EBM model using volume biomarkers instead of deep learning-based features, as well as an end-to-end convolutional neural network (CNN) with optimized architecture.

AB - Machine learning methods have shown large potential for the automatic early diagnosis of Alzheimer’s Disease (AD). However, some machine learning methods based on imaging data have poor interpretability because it is usually unclear how they make their decisions. Explainable Boosting Machines (EBMs) are interpretable machine learning models based on the statistical framework of generalized additive modeling, but have so far only been used for tabular data. Therefore, we propose a framework that combines the strength of EBM with high-dimensional imaging data using deep learning-based feature extraction. The proposed framework is interpretable because it provides the importance of each feature. We validated the proposed framework on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset, achieving accuracy of 0.883 and area-under-the-curve (AUC) of 0.970 on AD and control classification. Furthermore, we validated the proposed framework on an external testing set, achieving accuracy of 0.778 and AUC of 0.887 on AD and subjective cognitive decline (SCD) classification. The proposed framework significantly outperformed an EBM model using volume biomarkers instead of deep learning-based features, as well as an end-to-end convolutional neural network (CNN) with optimized architecture.

KW - Alzheimer’s disease

KW - Convolutional neural network

KW - Explainable boosting machine

KW - Interpretable AI

KW - MRI

U2 - 10.1007/978-3-031-47401-9_7

DO - 10.1007/978-3-031-47401-9_7

M3 - Conference article in proceeding

SN - 9783031474002

VL - 14393

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 69

EP - 78

BT - Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings

A2 - Celebi, M. Emre

A2 - Salekin, Md Sirajus

A2 - Kim, Hyunwoo

A2 - Albarqouni, Shadi

PB - Springer Verlag

T2 - 26th International Conference on Medical Image Computing and Computer Assisted Intervention

Y2 - 8 October 2023 through 12 October 2023

ER -

Kang W, Li B, Papma JM, Jiskoot LC, Deyn PPD, Biessels GJ et al. An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease. In Celebi ME, Salekin MS, Kim H, Albarqouni S, editors, Medical Image Computing and Computer Assisted Intervention – MICCAI 2023 Workshops - ISIC 2023, Care-AI 2023, MedAGI 2023, DeCaF 2023, Held in Conjunction with MICCAI 2023, Proceedings. Vol. 14393. Springer Verlag. 2023. p. 69-78. 305659. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 14393). doi: 10.1007/978-3-031-47401-9_7

An Interpretable Machine Learning Model with Deep Learning-Based Imaging Biomarkers for Diagnosis of Alzheimer’s Disease

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Cite this