{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T08:17:01Z","timestamp":1781597821543,"version":"3.54.5"},"reference-count":41,"publisher":"MDPI AG","issue":"12","license":[{"start":{"date-parts":[[2022,6,9]],"date-time":"2022-06-09T00:00:00Z","timestamp":1654732800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"Bio &amp; Medical Technology Development Program of the National Research Foundation (NRF) and funded by the Korean government (MSIT)","award":["NRF-2019R1G1A1011227"],"award-info":[{"award-number":["NRF-2019R1G1A1011227"]}]},{"name":"Bio &amp; Medical Technology Development Program of the National Research Foundation (NRF) and funded by the Korean government (MSIT)","award":["NRF-2019R1G1A1097707"],"award-info":[{"award-number":["NRF-2019R1G1A1097707"]}]},{"name":"Bio &amp; Medical Technology Development Program of the National Research Foundation (NRF) and funded by the Korean government (MSIT)","award":["2018-12-031"],"award-info":[{"award-number":["2018-12-031"]}]},{"name":"HK inno. N Corp","award":["NRF-2019R1G1A1011227"],"award-info":[{"award-number":["NRF-2019R1G1A1011227"]}]},{"name":"HK inno. N Corp","award":["NRF-2019R1G1A1097707"],"award-info":[{"award-number":["NRF-2019R1G1A1097707"]}]},{"name":"HK inno. N Corp","award":["2018-12-031"],"award-info":[{"award-number":["2018-12-031"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JCM"],"abstract":"<jats:p>Purpose: We investigated whether a deep learning algorithm applied to retinal fundoscopic images could predict cerebral white matter hyperintensity (WMH), as represented by a modified Fazekas scale (FS), on brain magnetic resonance imaging (MRI). Methods: Participants who had undergone brain MRI and health-screening fundus photography at Hallym University Sacred Heart Hospital between 2010 and 2020 were consecutively included. The subjects were divided based on the presence of WMH, then classified into three groups according to the FS grade (0 vs. 1 vs. 2+) using age matching. Two pre-trained convolutional neural networks were fine-tuned and evaluated for prediction performance using 10-fold cross-validation. Results: A total of 3726 fundus photographs from 1892 subjects were included, of which 905 fundus photographs from 462 subjects were included in the age-matched balanced dataset. In predicting the presence of WMH, the mean area under the receiver operating characteristic curve was 0.736 \u00b1 0.030 for DenseNet-201 and 0.724 \u00b1 0.026 for EfficientNet-B7. For the prediction of FS grade, the mean accuracies reached 41.4 \u00b1 5.7% with DenseNet-201 and 39.6 \u00b1 5.6% with EfficientNet-B7. The deep learning models focused on the macula and retinal vasculature to detect an FS of 2+. Conclusions: Cerebral WMH might be partially predicted by non-invasive fundus photography via deep learning, which may suggest an eye\u2013brain association.<\/jats:p>","DOI":"10.3390\/jcm11123309","type":"journal-article","created":{"date-parts":[[2022,6,10]],"date-time":"2022-06-10T02:25:33Z","timestamp":1654827933000},"page":"3309","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":10,"title":["Prediction of White Matter Hyperintensity in Brain MRI Using Fundus Photographs via Deep Learning"],"prefix":"10.3390","volume":"11","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-0244-388X","authenticated-orcid":false,"given":"Bum-Joo","family":"Cho","sequence":"first","affiliation":[{"name":"Department of Ophthalmology, Hallym University Sacred Heart Hospital, Anyang 14068, Korea"},{"name":"Medical Artificial Intelligence Center, Hallym University Medical Center, Anyang 14068, Korea"},{"name":"Division of Biomedical Informatics, Seoul National University Biomedical Informatics (SNUBI), Seoul National University College of Medicine, Seoul 03080, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8474-5744","authenticated-orcid":false,"given":"Minwoo","family":"Lee","sequence":"additional","affiliation":[{"name":"Department of Neurology, Hallym Neurological Institute, Hallym University Sacred Heart Hospital, Anyang 14068, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jiyong","family":"Han","sequence":"additional","affiliation":[{"name":"Medical Artificial Intelligence Center, Hallym University Medical Center, Anyang 14068, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Soonil","family":"Kwon","sequence":"additional","affiliation":[{"name":"Department of Ophthalmology, Hallym University Sacred Heart Hospital, Anyang 14068, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mi Sun","family":"Oh","sequence":"additional","affiliation":[{"name":"Department of Neurology, Hallym Neurological Institute, Hallym University Sacred Heart Hospital, Anyang 14068, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kyung-Ho","family":"Yu","sequence":"additional","affiliation":[{"name":"Department of Neurology, Hallym Neurological Institute, Hallym University Sacred Heart Hospital, Anyang 14068, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3885-981X","authenticated-orcid":false,"given":"Byung-Chul","family":"Lee","sequence":"additional","affiliation":[{"name":"Department of Neurology, Hallym Neurological Institute, Hallym University Sacred Heart Hospital, Anyang 14068, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ju Han","family":"Kim","sequence":"additional","affiliation":[{"name":"Division of Biomedical Informatics, Seoul National University Biomedical Informatics (SNUBI), Seoul National University College of Medicine, Seoul 03080, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-8762-8340","authenticated-orcid":false,"given":"Chulho","family":"Kim","sequence":"additional","affiliation":[{"name":"Department of Neurology, Chuncheon Sacred Heart Hospital, Chuncheon 24253, Korea"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2022,6,9]]},"reference":[{"key":"ref_1","doi-asserted-by":"crossref","first-page":"111","DOI":"10.5853\/jos.2015.17.2.111","article-title":"Prevention and management of cerebral small vessel disease","volume":"17","author":"Mok","year":"2015","journal-title":"J. Stroke"},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"1146","DOI":"10.1212\/WNL.0000000000007654","article-title":"CNS small vessel disease: A clinical review","volume":"92","author":"Cannistraro","year":"2019","journal-title":"Neurology"},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"702","DOI":"10.1136\/jnnp.2006.103549","article-title":"Risk factor profile of cerebral small vessel disease and its subtypes","volume":"78","author":"Khan","year":"2007","journal-title":"J. Neurol. Neurosurg. Psychiatry"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"20130832","DOI":"10.1259\/bjr.20130832","article-title":"Retinal imaging as a source of biomarkers for diagnosis, characterization and prognosis of chronic illness or long-term conditions","volume":"87","author":"MacGillivray","year":"2014","journal-title":"Br. J. Radiol."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"319","DOI":"10.1111\/j.1469-7580.2005.00395.x","article-title":"Retinal vascular image analysis as a potential screening tool for cerebrovascular disease: A rationale based on homology between cerebral and retinal microvasculatures","volume":"206","author":"Patton","year":"2005","journal-title":"J. Anat."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"H201","DOI":"10.1152\/ajpheart.00201.2016","article-title":"Retinal microvascular network alterations: Potential biomarkers of cerebrovascular and neural diseases","volume":"312","author":"Somfai","year":"2017","journal-title":"Am. J. Physiol. Heart Circ. Physiol."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"1328","DOI":"10.1161\/CIRCULATIONAHA.116.023425","article-title":"Retinal vessel calibers in predicting long-term cardiovascular outcomes: The atherosclerosis risk in communities study","volume":"134","author":"Seidelmann","year":"2016","journal-title":"Circulation"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"59","DOI":"10.1016\/S0039-6257(01)00234-X","article-title":"Retinal microvascular abnormalities and their relationship with hypertension, cardiovascular disease, and mortality","volume":"46","author":"Wong","year":"2001","journal-title":"Surv. Ophthalmol."},{"key":"ref_9","first-page":"e56137","article-title":"Using retinal imaging to study dementia","volume":"129","author":"Chan","year":"2017","journal-title":"J. Vis. Exp."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"e1","DOI":"10.1002\/mp.13264","article-title":"Deep learning in medical imaging and radiation therapy","volume":"46","author":"Sahiner","year":"2019","journal-title":"Med. Phys."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"639","DOI":"10.1016\/j.jacr.2019.12.026","article-title":"Deep learning for natural language processing in radiology\u2014Fundamentals and a systematic review","volume":"17","author":"Sorin","year":"2020","journal-title":"J. Am. Coll. Radiol."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"277","DOI":"10.5853\/jos.2017.02054","article-title":"Deep into the brain: Artificial intelligence in stroke imaging","volume":"19","author":"Lee","year":"2017","journal-title":"J. Stroke"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"101804","DOI":"10.1016\/j.flowmeasinst.2020.101804","article-title":"Application of GMDH neural network technique to improve measuring precision of a simplified photon attenuation based two-phase flowmeter","volume":"75","author":"Roshani","year":"2020","journal-title":"Flow Meas. Instrum."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"565","DOI":"10.1016\/j.ophtha.2018.11.015","article-title":"DeepSeeNet: A deep learning model for automated classification of patient-based age-related macular degeneration severity from color fundus photographs","volume":"126","author":"Peng","year":"2019","journal-title":"Ophthalmology"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"1353","DOI":"10.1001\/jamaophthalmol.2019.3501","article-title":"Development and validation of a deep learning system to detect glaucomatous optic neuropathy using fundus photographs","volume":"137","author":"Liu","year":"2019","journal-title":"JAMA Ophthalmol."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"98","DOI":"10.1002\/acn3.688","article-title":"Retinal image analytics detects white matter hyperintensities in healthy adults","volume":"6","author":"Lau","year":"2019","journal-title":"Ann. Clin. Transl. Neurol."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"171","DOI":"10.1007\/s00401-011-0851-x","article-title":"Heterogeneity in age-related white matter changes","volume":"122","author":"Schmidt","year":"2011","journal-title":"Acta. Neuropathol."},{"key":"ref_18","doi-asserted-by":"crossref","first-page":"1497","DOI":"10.1136\/jnnp.2005.064998","article-title":"Medial temporal lobe atrophy and white matter hyperintensities are associated with mild cognitive deficits in non-disabled elderly people: The LADIS study","volume":"76","author":"Barkhof","year":"2005","journal-title":"J. Neurol. Neurosurg. Psychiatry"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"e002274","DOI":"10.1136\/bmjdrc-2021-002274","article-title":"Cerebral small-vessel disease is associated with the severity of diabetic retinopathy in type 1 diabetes","volume":"9","author":"Eriksson","year":"2021","journal-title":"BMJ Open Diab. Res. Care"},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"54","DOI":"10.3174\/ajnr.A3146","article-title":"FLAIR and diffusion MRI signals are independent predictors of white matter hyperintensities","volume":"34","author":"Maillard","year":"2013","journal-title":"Am. J. Neuroradiol."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"351","DOI":"10.2214\/ajr.149.2.351","article-title":"MR signal abnormalities at 1.5 T in Alzheimer\u2019s dementia and normal aging","volume":"149","author":"Fazekas","year":"1987","journal-title":"Am. J. Roentgenol."},{"key":"ref_22","unstructured":"Pizer, S.M., Johnston, R.E., Ericksen, J.P., Yankaskas, B.C., and Muller, K.E. (1990, January 22\u201325). Contrast-limited adaptive histogram equalization: Speed and effectiveness. Proceedings of the First Conference on Visualization in Biomedical Computing, Atlanta, GA, USA."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Huang, G., Liu, Z., van Der Maaten, L., and Weinberger, K.Q. (2017, January 21\u201326). Densely connected convolutional networks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, HI, USA.","DOI":"10.1109\/CVPR.2017.243"},{"key":"ref_24","unstructured":"Tan, M., and Le, Q.V. (2019, January 10\u201315). EfficientNet: Rethinking model scaling for convolutional neural networks. Proceedings of the International Conference of Machine Learning, Long Beach, CA, USA."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Yao, H., Zhu, D.L., Jiang, B., and Yu, P. (2019, January 25\u201326). Negative log likelihood ratio loss for deep neural network classification. Proceedings of the 2019 Future Technologies Conference, San Francisco, CA, USA.","DOI":"10.1007\/978-3-030-32520-6_22"},{"key":"ref_26","doi-asserted-by":"crossref","unstructured":"Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., and Batra, D. (2017, January 22\u201329). Grad-cam: Visual explanations from deep networks via gradient-based localization. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.","DOI":"10.1109\/ICCV.2017.74"},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"670","DOI":"10.1093\/aje\/kwj063","article-title":"The inconsistency of \u201coptimal\u201d cutpoints obtained using two criteria based on the receiver operating characteristic curve","volume":"163","author":"Perkins","year":"2006","journal-title":"Am. J. Epidemiol."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"434","DOI":"10.1111\/j.1747-4949.2010.00483.x","article-title":"Retinal arteriolar geometry is associated with cerebral white matter hyperintensities on magnetic resonance imaging","volume":"5","author":"Doubal","year":"2010","journal-title":"Int. J. Stroke"},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"1003","DOI":"10.1212\/WNL.0000000000003080","article-title":"Retinal microvasculature and white matter microstructure: The Rotterdam study","volume":"87","author":"Mutlu","year":"2016","journal-title":"Neurology"},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"e142","DOI":"10.1212\/WNL.0000000000004792","article-title":"Differential associations between retinal signs and CMBs by location: The AGES-Reykjavik study","volume":"90","author":"Qiu","year":"2018","journal-title":"Neurology"},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"1102","DOI":"10.1212\/WNL.0b013e3181d7d8b4","article-title":"Fractal analysis of retinal vessels suggests that a distinct vasculopathy causes lacunar stroke","volume":"74","author":"Doubal","year":"2010","journal-title":"Neurology"},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"389","DOI":"10.1161\/STROKEAHA.108.529388","article-title":"Retinopathy in ischemic stroke subtypes","volume":"40","author":"Doubal","year":"2009","journal-title":"Stroke"},{"key":"ref_33","doi-asserted-by":"crossref","first-page":"158","DOI":"10.1038\/s41551-018-0195-0","article-title":"Prediction of cardiovascular risk factors from retinal fundus photographs via deep learning","volume":"2","author":"Poplin","year":"2018","journal-title":"Nat. Biomed. Eng."},{"key":"ref_34","doi-asserted-by":"crossref","unstructured":"Adedigba, A.P., Adeshininat, S.A., and Aibinu, A.M. (2019, January 10\u201312). Deep learning-based mammogram classification using small dataset. Proceedings of the 2019 15th International Conference on Electronics, Computer and Computation (ICECCO), Abuja, Nigeria.","DOI":"10.1109\/ICECCO48375.2019.9043186"},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"1391","DOI":"10.1161\/01.STR.0000221308.94473.14","article-title":"White matter lesions in an unselected cohort of the elderly: Molecular pathology suggests origin from chronic hypoperfusion injury","volume":"37","author":"Fernando","year":"2006","journal-title":"Stroke"},{"key":"ref_36","doi-asserted-by":"crossref","first-page":"385","DOI":"10.1111\/j.1365-2990.2004.00550.x","article-title":"Comparison of the pathology of cerebral white matter with post-mortem magnetic resonance imaging (MRI) in the elderly brain","volume":"30","author":"Fernando","year":"2004","journal-title":"Neuropathol. Appl. Neurobiol."},{"key":"ref_37","doi-asserted-by":"crossref","first-page":"223","DOI":"10.1177\/1479164118758581","article-title":"Long-term changes in retinal vascular diameter and cognitive impairment in type 1 diabetes","volume":"15","author":"Nunley","year":"2018","journal-title":"Diab. Vasc. Dis. Res."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"106","DOI":"10.1159\/000486269","article-title":"Novel retinal imaging in assessment of cardiovascular risk factors and systemic vascular diseases","volume":"Volume 27","author":"Sabanayagam","year":"2019","journal-title":"Diabetic Retinopathy and Cardiovascular Disease"},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"2153","DOI":"10.1007\/s10792-019-01074-z","article-title":"Accuracy of ultrawide-field fundus ophthalmoscopy-assisted deep learning for detecting treatment-na\u00efve proliferative diabetic retinopathy","volume":"39","author":"Nagasawa","year":"2019","journal-title":"Int. Ophthalmol."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"1269","DOI":"10.1007\/s10792-018-0940-0","article-title":"Accuracy of ultra-wide-field fundus ophthalmoscopy-assisted deep learning, a machine-learning technology, for detecting age-related macular degeneration","volume":"39","author":"Matsuba","year":"2019","journal-title":"Int. Ophthalmol."},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"1875431","DOI":"10.1155\/2018\/1875431","article-title":"Deep neural network-based method for detecting central retinal vein occlusion using ultrawide-field fundus ophthalmoscopy","volume":"2018","author":"Nagasato","year":"2018","journal-title":"J. Ophthalmol."}],"container-title":["Journal of Clinical Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2077-0383\/11\/12\/3309\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T23:27:03Z","timestamp":1760138823000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2077-0383\/11\/12\/3309"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,6,9]]},"references-count":41,"journal-issue":{"issue":"12","published-online":{"date-parts":[[2022,6]]}},"alternative-id":["jcm11123309"],"URL":"https:\/\/doi.org\/10.3390\/jcm11123309","relation":{},"ISSN":["2077-0383"],"issn-type":[{"value":"2077-0383","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,6,9]]}}}