{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,17]],"date-time":"2026-07-17T10:26:30Z","timestamp":1784283990592,"version":"3.55.0"},"reference-count":44,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2024,1,8]],"date-time":"2024-01-08T00:00:00Z","timestamp":1704672000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>Graph-based representations are becoming more common in the medical domain, where each node defines a patient, and the edges signify associations between patients, relating individuals with disease and symptoms in a node classification task. In this study, a Graph Convolutional Networks (GCN) model was utilized to capture differences in neurocognitive, genetic, and brain atrophy patterns that can predict cognitive status, ranging from Normal Cognition (NC) to Mild Cognitive Impairment (MCI) and Alzheimer's Disease (AD), on the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. Elucidating model predictions is vital in medical applications to promote clinical adoption and establish physician trust. Therefore, we introduce a decomposition-based explanation method for individual patient classification.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>Our method involves analyzing the output variations resulting from decomposing input values, which allows us to determine the degree of impact on the prediction. Through this process, we gain insight into how each feature from various modalities, both at the individual and group levels, contributes to the diagnostic result. Given that graph data contains critical information in edges, we studied relational data by silencing all the edges of a particular class, thereby obtaining explanations at the neighborhood level.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Our functional evaluation showed that the explanations remain stable with minor changes in input values, specifically for edge weights exceeding 0.80. Additionally, our comparative analysis against SHAP values yielded comparable results with significantly reduced computational time. To further validate the model's explanations, we conducted a survey study with 11 domain experts. The majority (71%) of the responses confirmed the correctness of the explanations, with a rating of above six on a 10-point scale for the understandability of the explanations.<\/jats:p><\/jats:sec><jats:sec><jats:title>Discussion<\/jats:title><jats:p>Strategies to overcome perceived limitations, such as the GCN's overreliance on demographic information, were discussed to facilitate future adoption into clinical practice and gain clinicians' trust as a diagnostic decision support system.<\/jats:p><\/jats:sec>","DOI":"10.3389\/frai.2023.1334613","type":"journal-article","created":{"date-parts":[[2024,1,8]],"date-time":"2024-01-08T06:12:03Z","timestamp":1704694323000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":20,"title":["Explaining graph convolutional network predictions for clinicians\u2014An explainable AI approach to Alzheimer's disease classification"],"prefix":"10.3389","volume":"6","author":[{"given":"Sule","family":"Tekkesinoglu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sara","family":"Pudas","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2024,1,8]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"4758","DOI":"10.3390\/s21144758","article-title":"Graph-based deep learning for medical diagnosis and analysis: past, present and future","volume":"21","author":"Ahmedt-Aristizabal","year":"2021","journal-title":"Sensors"},{"key":"B2","doi-asserted-by":"crossref","first-page":"3197","DOI":"10.1109\/ICASSP.2019.8683547","article-title":"\u201cBootstrapping graph convolutional neural networks for autism spectrum disorder classification,\u201d","volume-title":"ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)","author":"Anirudh","year":"2019"},{"key":"B3","article-title":"Explainability techniques for graph convolutional networks","author":"Baldassarre","year":"2019","journal-title":"arXiv"},{"key":"B4","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41598-022-10202-2","article-title":"In-depth insights into Alzheimer's disease by using explainable machine learning approach","volume":"12","author":"Bogdanovic","year":"2022","journal-title":"Sci. Rep"},{"key":"B5","doi-asserted-by":"crossref","DOI":"10.4135\/9781446263020","volume-title":"Psychological Assessment: Four volume Set","author":"Boyle","year":"2012"},{"key":"B6","doi-asserted-by":"crossref","first-page":"839","DOI":"10.1109\/WACV.2018.00097","article-title":"\u201cGrad-cam++: Generalized gradient-based visual explanations for deep convolutional networks,\u201d","volume-title":"2018 IEEE Winter Conference on Applications of Computer Vision (WACV)","author":"Chattopadhay","year":"2018"},{"key":"B7","doi-asserted-by":"publisher","first-page":"21","DOI":"10.3389\/fdata.2021.613047","article-title":"Developing an explainable machine learning-based personalised dementia risk prediction model: a transfer learning approach with ensemble learning algorithms","volume":"4","author":"Danso","year":"2021","journal-title":"Front. Big Data"},{"key":"B8","doi-asserted-by":"publisher","first-page":"1439","DOI":"10.3233\/JAD-161135","article-title":"Severity of depression impacts imminent conversion from mild cognitive impairment to Alzheimer's disease","volume":"59","author":"Defrancesco","year":"2017","journal-title":"J. Alzheimer's Dis"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pmed.1002289","article-title":"Genetic assessment of age-associated Alzheimer disease risk: Development and validation of a polygenic hazard score","author":"Desikan","year":"2017","journal-title":"PLoS Med"},{"key":"B10","article-title":"Towards a rigorous science of interpretable machine learning","author":"Doshi-Velez","year":"2017","journal-title":"arXiv"},{"key":"B11","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41598-021-82098-3","article-title":"A multilayer multimodal detection and prediction model based on explainable artificial intelligence for Alzheimer's disease","volume":"11","author":"El-Sappagh","year":"2021","journal-title":"Sci. Rep"},{"key":"B12","doi-asserted-by":"publisher","first-page":"774","DOI":"10.1016\/j.neuroimage.2012.01.021","article-title":"Freesurfer","volume":"62","author":"Fischl","year":"2012","journal-title":"Neuroimage"},{"key":"B13","author":"Ghanvatkar","year":"2022","journal-title":"Towards a Theory-Based Evaluation of Explainable Predictions in Healthcare"},{"key":"B14","doi-asserted-by":"publisher","DOI":"10.1016\/j.nicl.2020.102199","article-title":"Modelling prognostic trajectories of cognitive decline due to Alzheimer's disease","author":"Giorgio","year":"2020","journal-title":"NeuroImage: Clini"},{"key":"B15","author":"Hill","year":"2022","journal-title":"How the Symptoms of Alzheimer's are Related to the Brain Lobe affected"},{"key":"B16","doi-asserted-by":"publisher","first-page":"28","DOI":"10.1016\/j.inffus.2021.01.008","article-title":"Towards multi-modal causability with graph neural networks enabling information fusion for explainable AI","volume":"71","author":"Holzinger","year":"2021","journal-title":"Inform. Fusion"},{"key":"B17","doi-asserted-by":"publisher","first-page":"4153","DOI":"10.1109\/JBHI.2022.3172479","article-title":"Explainable identification of dementia from transcripts using transformer networks","volume":"26","author":"Ilias","year":"2022","journal-title":"IEEE J. Biomed. Health Inform"},{"key":"B18","doi-asserted-by":"publisher","first-page":"685","DOI":"10.1002\/jmri.21049","article-title":"The Alzheimer's disease neuroimaging initiative (ADNI): MRI methods","volume":"27","author":"Jack","year":"2008","journal-title":"J.Magnet. Reson. Imag"},{"key":"B19","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TIM.2021.3107056","article-title":"Alzheimer's patient analysis using image and gene expression data and explainable-AI to present associated genes","volume":"70","author":"Kamal","year":"2021","journal-title":"IEEE Trans. Instrum. Meas"},{"key":"B20","article-title":"Semi-supervised classification with graph convolutional networks","author":"Kipf","year":"2016","journal-title":"arXiv preprint arXiv"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.1016\/j.patcog.2022.109106","article-title":"Multi-scale enhanced graph convolutional network for mild cognitive impairment detection","author":"Lei","year":"2023","journal-title":"Pattern Recognit"},{"key":"B22","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1186\/s40708-022-00165-5","article-title":"A robust framework to investigate the reliability and stability of explainable artificial intelligence markers of mild cognitive impairment and Alzheimer's disease","volume":"9","author":"Lombardi","year":"2022","journal-title":"Brain Inform"},{"key":"B23","doi-asserted-by":"publisher","first-page":"56","DOI":"10.1038\/s42256-019-0138-9","article-title":"From local explanations to global understanding with explainable AI for trees","volume":"2","author":"Lundberg","year":"2020","journal-title":"Nat. Mach. Intellig"},{"key":"B24","first-page":"4768","article-title":"A unified approach to interpreting model predictions","volume":"30","author":"Lundberg","year":"2017","journal-title":"Adv. Neural Inf. Process. Syst"},{"key":"B25","doi-asserted-by":"publisher","DOI":"10.1016\/j.neuroimage.2023.120073","article-title":"Xadlime: explainable Alzheimer's disease likelihood map estimation via clinically-guided prototype learning","author":"Mulyadi","year":"2022","journal-title":"arXiv"},{"key":"B26","doi-asserted-by":"publisher","first-page":"117","DOI":"10.1016\/j.media.2018.06.001","article-title":"Disease prediction using graph convolutional networks: application to autism spectrum disorder and Alzheimer's disease","volume":"48","author":"Parisot","year":"2018","journal-title":"Med. Image Anal"},{"key":"B27","first-page":"177","article-title":"\u201cSpectral graph convolutions for population-based disease prediction,\u201d","volume-title":"International Conference on Medical Image Computing and Computer-Assisted Intervention","author":"Parisot","year":"2017"},{"key":"B28","doi-asserted-by":"publisher","first-page":"2707","DOI":"10.1002\/alz.12662","volume":"18","author":"Podcasy","year":"2022"},{"key":"B29","first-page":"10772","article-title":"\u201cExplainability methods for graph convolutional neural networks,\u201d","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Pope","year":"2019"},{"key":"B30","doi-asserted-by":"publisher","first-page":"363","DOI":"10.1016\/j.inffus.2022.11.028","article-title":"Prediction of Alzheimer's progression based on multimodal deep-learning-based fusion and visual explainability of time-series data","volume":"92","author":"Rahim","year":"2022","journal-title":"Inform. Fusion"},{"key":"B31","doi-asserted-by":"publisher","first-page":"1899","DOI":"10.1212\/WNL.0b013e3181c3f293","article-title":"Age, Alzheimer disease, and brain structure","volume":"73","author":"Raji","year":"2009","journal-title":"Neurology"},{"key":"B32","doi-asserted-by":"publisher","first-page":"6001","DOI":"10.3390\/s20216001","article-title":"Population graph-based multi-model ensemble method for diagnosing autism spectrum disorder","volume":"20","author":"Rakhimberdina","year":"2020","journal-title":"Sensors"},{"key":"B33","first-page":"815","article-title":"\u201cLinear graph convolutional model for diagnosing brain disorders,\u201d","volume-title":"International Conference on Complex Networks and Their Applications","author":"Rakhimberdina","year":"2019"},{"key":"B34","doi-asserted-by":"publisher","first-page":"1402","DOI":"10.1016\/j.neuroimage.2012.02.084","article-title":"Within-subject template estimation for unbiased longitudinal image analysis","volume":"61","author":"Reuter","year":"2012","journal-title":"Neuroimage"},{"key":"B35","doi-asserted-by":"publisher","first-page":"589","DOI":"10.1109\/TKDE.2007.190734","article-title":"Explaining classifications for individual instances","volume":"20","author":"Robnik-\u0160ikonja","year":"2008","journal-title":"IEEE Trans. Knowl. Data Eng"},{"key":"B36","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D19-5308","article-title":"Layerwise relevance visualization in convolutional text graph classifiers","author":"Schwarzenberg","year":"2019","journal-title":"arXiv"},{"key":"B37","doi-asserted-by":"publisher","first-page":"e0244773","DOI":"10.1371\/journal.pone.0244773","article-title":"Random forest model for feature-based Alzheimer's disease conversion prediction from early mild cognitive impairment subjects","volume":"16","author":"Velazquez","year":"2021","journal-title":"PLoS ONE"},{"key":"B38","article-title":"Explainable artificial intelligence: a systematic review","author":"Vilone","year":"2020","journal-title":"arXiv"},{"key":"B39","doi-asserted-by":"publisher","first-page":"S527","DOI":"10.3233\/JAD-2010-100501","article-title":"Why women have more Alzheimer's disease than men: gender and mitochondrial toxicity of amyloid-\u03b2 peptide","volume":"20","author":"Vina","year":"2010","journal-title":"J. Alzheimer's Dis"},{"key":"B40","first-page":"6861","article-title":"\u201cSimplifying graph convolutional networks,\u201d","volume-title":"International Conference on Machine Learning","author":"Wu","year":"2019"},{"key":"B41","doi-asserted-by":"crossref","first-page":"70","DOI":"10.1007\/978-3-030-35817-4_9","article-title":"\u201cTriplet graph convolutional network for multi-scale analysis of functional connectivity using functional MRI,\u201d","volume-title":"International Workshop on Graph Learning in Medical Imaging","author":"Yao","year":"2019"},{"key":"B42","first-page":"9244","article-title":"Gnnexplainer: Generating explanations for graph neural networks","volume":"32","author":"Ying","year":"2019","journal-title":"Adv. Neural Inf. Process. Syst"},{"key":"B43","doi-asserted-by":"publisher","first-page":"237","DOI":"10.3390\/diagnostics12020237","article-title":"Applications of explainable artificial intelligence in diagnosis and surgery","volume":"12","author":"Zhang","year":"2022","journal-title":"Diagnostics"},{"key":"B44","doi-asserted-by":"crossref","DOI":"10.1109\/ISBI52829.2022.9761449","article-title":"\u201cInterpretable graph convolutional network of multi-modality brain imaging for Alzheimer's disease diagnosis,\u201d","volume-title":"2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI)","author":"Zhou","year":"2022"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2023.1334613\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,1,8]],"date-time":"2024-01-08T06:12:14Z","timestamp":1704694334000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2023.1334613\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,1,8]]},"references-count":44,"alternative-id":["10.3389\/frai.2023.1334613"],"URL":"https:\/\/doi.org\/10.3389\/frai.2023.1334613","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,1,8]]},"article-number":"1334613"}}