Siamak Yousefi received his PhD in Electrical Engineering from the University of Texas at Dallas in 2012 where he conducted research on biomedical pattern recognition, machine learning, and medical image processing in the Signal and Image Processing (SIP) Laboratory. He was trained as a postdoctoral fellow at the Neural Signal Processing (NSP) Laboratory at the University of California Los Angeles (UCLA) from 2012 to 2013 working on Brain Computer Interface (BCI) and retinotopic mapping. He completed another postdoctoral training at the University of California San Diego (UCSD), Shiley Eye Institute, working on computational ophthalmology and retinal data mining and machine learning from 2013 to 2014. He was a research scientist at the same institute from 2014 to 2016. He also was an Adjunct Professor at San Diego State University (SDSU) from 2014 to 2017. Siamak was a Visiting Assistant Professor at the University of Tokyo, Department of Creative Informatics and Department of Ophthalmology in 2017.
His research areas include retinal data mining and machine learning and biomedical image analysis. He is on editorial board of EC Ophthalmology, a senior member of the Institute of Electrical and Electronics Engineers (IEEE), and a member of the Association for Research in Vision and Ophthalmology (ARVO).
Data Mining and Machine Learning (DM2L) Laboratory
In DM2L lab, we are working on emerging deep learning, conventional machine learning and big retinal and corneal data mining techniques to identify the onset and progression of different ocular conditions using novel imaging signatures.
Our ongoing projects could be of interest to students with background in computer science, engineering, biostatistics, or medicine. We typically work on retinal fundus photographs, visual fields, optical coherence tomography (OCT), medical records, and demographic parameters from patients with glaucoma, diabetic retinopathy, macular degeneration, keratoconus, and uveitis.
DM2L laboratory is located on the 7th floor of the Hamilton Eye Institute, Department of Ophthalmology of UTHSC in Memphis. DM2L is affiliated to the Department of Genetics, Genomics, and Informatics as well.
Keywords: Big ophthalmic data, ocular data, data mining, machine learning, deep learning, artificial intelligence, automated diagnosis, and medical expert systems.
Siamak Yousefi, PhD
Assistant Professor of Ophthalmology
Assistant Professor of Genetics, Genomics, and Informatics
Xiaoqin Huang, PhD, Postdoctoral Fellow
Rafiqul Islam, PhD, Prospective Postdoctoral Fellow
Yeganeh Madadi, PhD, Research Fellow
Jian Sun, PhD, Research Scientist
Ehsan Kazemi, MD, Prospective Clinical Research Fellow
Asma Poursoroush, MSs, Volunteer Research Assistant
Former Lab members
Krati Gupta, PhD, Research Assistant, 2020-2021
Sidharth Mahotra, MSc, Research Assistant, 2020
Eugene Rho, Medical Student, Research Fellow
Anshul Thakur, PhD, Postdoctoral Fellow, 2019
Hassan Kabiri, MD, Research Fellow, 2019
Mohammad Norouzifard, PhD Student, Research Assistant, 2019
Tarus Dukes, BSc, Research Assistant, 2019
Golnoush-Sadat Mahmoudi-Nezhad, MPH, MD, Research Assistant, 2018
Edward De Guzman,MSc, Data Analytics, 2017-2018
Siamak's paper titled "Machine-identified patterns of visual field loss and an association with rapid progression in the ocular hypertension treatment study" was accepted by Ophthalmology journal. June 28.
Siamak gave a talk on "Towards objective criteria for glaucoma based on artificial intelligence" at the Hamilton Eye Institute, University of Tennessee Health Science Center (UTHSC), June 22.
Alborz, a robot from the NAO family, officially joined DM2L lab. We will train Alborz to become a digital ophthalmologist and a digital glaucoma expert. June 15.
Paper titled "Pointwise and regionwise course of visual field loss in patients with glaucoma" was accepted by the TVST journal. June 14.
Siamak gave a talk on "Harnessing AI for detecting ocular conditions: addressing vision-related public health challenges" at the Oracle Health. May 31.
Xiaoqin's paper, titled "An objective and easy-to-use glaucoma functional severity staging system based on artificial intelligence" was accepted by the Journal of Glaucoma. May 22. https://journals.lww.com/glaucomajournal/Abstract/9900/An_Objective_and_Easy_to_Use_Glaucoma_Functional.37.aspx
Siamak gave a talk on "Assistive and Autonomous AI models in glaucoma" at the ARVO Imaging Session, May 2.
Siamak gave a talk on "Risk of glaucoma based on imaging and genetic data: Artificial intelligence insights" at the Emory Eye Center, Emory University. April 26.
Siamak gave a talk on "Broad applications of artificial intelligence in ophthalmology" at the Hamilton Eye Institute’s Clinical Update and I. Lee Arnold meeting. April 8.
Xiaoqin received the Sarla P. Kothary Memorial Travel Grant Award from ARVO. March.
Two (out of five) abstracts were selected as oral presentations at the ARVO annual meeting. March.
Siamak gave a talk on "Applications of data mining and machine learning to ophthalmic medical images" at the 12th Iranian and the 2nd International Conference on Machine Vision & Image Processing (MVIP). February 22.
Siamak gave a talk on "Broad applications of artificial intelligence in vision and ophthalmology" at the Eye Research Center, Oakland University. January 25.
Xiaoqin Huang received a Postdoctoral Fellowship Grant Award from the Neuroscience Institute of the UTHSC. January.
Siamak gave a talk on "Applications of AI in keratoconus" at the Asia-Pacific Tele-Ophthalmology Society (APTOS) monthly meeting. January 7.
We received an R01 grant from NIH/NEI to develop AI models for predicting the risk of glaucoma based on imaging and visual field data. January.
Paper titled "A hybrid deep learning construct for detecting keratoconus from corneal maps" was accepted by TVST. November. https://doi.org/10.1167/tvst.10.14.16
We received the "2021 Roche Collaborative Research Fellowship" from ARVO Foundation to "develop device-agnostic artificial intelligence models to detect subclinical and early-stage keratoconus". November.
Siamak discussed "Current AI developments in glaucoma and future challenges" at the AI lounge of the AAO 2021 meeting that was organized by Lama Al-Aswad and Joel Schuman. November.
Siamak presented the abstract titled "Machine-identified glaucomatous patterns of visual field loss and their association with rapid glaucoma progression" at the AAO 2021 in New Orleans (oral/paper presentation). November.
Siamak gave a talk on "Artificial intelligence and the future of clinical practice" symposium in the "American Academy of Optometry" in Boston. November.
We received a "UTHSC grant award (NGS) to develop AI models to detect uveitis from fundus photographs". October.
Paper titled "Stacking Ensemble Learning in Deep Domain Adaptation for Ophthalmic Image Classification" received the Best Paper Award from OMIA workshop of the MICCAI conference. October.
Abstract titled “Single cell RNA sequencing reference mapping using deep transfer learning”, was accepted by the ISER/Bright Focus Glaucoma Symposium. August.
Siamak gave a talk on "Recent developments of artificial intelligence models in ophthalmology" at the Department of Ophthalmology, University of Colorado Medical School. August.
Paper titled "Stacking Ensemble Learning in Deep Domain Adaptation for Ophthalmic Image Classification" was accepted by MICCAI/OMIA. July.
Abstract titled "Machine-identified glaucomatous patterns of visual field loss and their association with rapid glaucoma progression" was accepted for oral/paper presentation at the AAO 2021. July.
Paper titled "Estimating the severity of visual field damage from retinal nerve fiber layer thickness measurements with artificial intelligence" was accepted by TVST. July. https://doi.org/10.1167/tvst.10.9.16
Siamak will be organizing an AI session on "Artificial intelligence in glaucoma" at the ISER 2022, Australia. June.
Xiaoqin Huang gave a talk on "Applications of artificial intelligence in glaucoma" at the UTHSC monthly vision research seminars. June.
Paper titled "Keratoconus severity detection from elevation, topography and pachymetry raw data using a machine learning approach" was accepted by IEEE Access. June. 10.1109/ACCESS.2021.3086021
Paper titled "Association between visual field damage and corneal structural parameters" was accepted by PLoS One. May. 10.1038/s41598-021-90298-0
Siamak gave a talk on "Latest artificial intelligence developments in ophthalmology: keratoconus, cataract, uveitis, and glaucoma" in the "Ophthalmology monthly innovation meetings" of the NYU, April 16. April.
Paper titled "Glaucoma precognition based on confocal scanning laser ophthalmoscopy images of the optic disc using convolutional neural network" was accepted by CVPR workshop on Precognition. April.
Siamak gave a talk on "Emergence of artificial intelligence algorithms in identifying glaucoma" at the Spanish Glaucoma Society (SGS). April.
Our first patent in corneal data analysis was issued. Statistical learning, manifold learning, and unsupervised clustering to mine non-invasive corneal imaging data and identify keratoconus severity applicable to predicting the future need for aggressive corneal surgeries. April.
Three abstracts were selected as paper/oral presentation at the ARVO Imaging in the Eye venue. April.
Three abstracts were selected as paper/oral presentation at the ARVO Main Conference. March.
Two ARVO abstracts received Travel Grant Awards. March.
Siamak was appointed as a member of the ARVO's Ethics and Regulations in Human Research (ERHR) Committee. March.
Siamak was invited to give a talk on "Glaucoma radar" at the 5th Asia Pacific Glaucoma Congress (AGPC) in June.
Paper titled "Prediction of Visual Field Progression from OCT Structural Measures in Moderate to Advanced Glaucoma" was accepted by the American Journal of Ophthalmology. https://doi.org/10.1016/j.ajo.2021.01.023. January.
Xiaoqin Huang joined DM2L as a postdoctoral fellow. January.
10 articles published
6 abstracts in AAO and ARVO
3 oral presentations at ARVO, ARVO-Imaging, and AAO
2 funding awards from NIH/NEI and BrightFocus Foundation with total ~ $650k
4 times highlighted in the Media and News: Ophthalmology Times, Healio Ophthalmology, Review of Optometry, and UTHSC
3 ocular conditions tackled: glaucoma, keratoconus, and uveitis
4 major machine-learning approcches developed: deep learning, unsupervised deep archetypal analysis, supervised machine learning, and manifold learning
5 major ocular data types were mined: Fundus photographs, visual fields, aneterior segment OCT, posterior segment OCT, single cell RNA-sequencing data from retinal ganglion cells (RGCs)
Siamak gave an invited talk at the "Indian conference of computer vision, graphics, and image processing" on "Evolution of AI in ophthalmology".
Paper titled "Patterns of retinal nerve fiber layer loss in patients with glaucoma identified by deep archetypal analysis" was accepted by IEEE BigData Workshop on Data Science in Medicine and Healthcare (DSMH).
Paper titled "Identifying mouse autoimmune uveitis from fundus photographs using deep learning" was accepted by TVST special issue on AI.
Siamak joined the Editorial Board of the TVST journal.
Andrew Johnston, a second-year ophthalmology resident, joined DM2L lab.
Eugene Rho, a second-year medical student, joined DM2L lab.
We received an R21 NIH/NEI funding to work on a study entitled "Improved glaucoma monitoring using artificial-intelligence enabled dashboard". Using this new support, we will develop manifold learning and unsupervised clustering models to monitor visual functional worsening in patients with glaucoma and provide simple and effective clinician-friendly visualizations like a radar.
Abstract titled "Preclinical signs of visual field loss in glaucoma suspects" was accepted by AAO 2020 as paper presentation.
Paper titled "Detecting keratoconus from corneal imaging data using machine learning" was accepted by the IEEE Access.
Paper titled "An artificial intelligence approach to assess spatial patterns of retinal nerve fiber layer thickness maps in glaucoma" was accepted by Translational Vision Science and Technology (TVST).
We received funding support by UTHSC Office of Research to strengthen our AI program in the Hamilton Eye Institute (HEI).
We were awarded a two-year grant by Bright Focus Foundation to study the impact of glaucoma on retinal ganglion cells (RGCs) from Transcriptomics aspects
Paper titled "Glaucoma precognition: recognizing preclinical visual functional signs of glaucoma" was accepted by Computer Vision and Pattern Recognition (CVPR) to be held in June in Seattle, Washington
Paper titled "Predicting glaucoma before onset using deep learning" was accepted by Ophthalmology Glaucoma Journal
Paper titled "Convex representations using deep archetypal analysis for predicting glaucoma" was accepted by the IEEE Journal of Transnational Engineering in Health & Medicine (JTEHM) Journal
Our research was highlighted on the cover page of the Ophthalmology Times, March Edition, 2020, "AI-enabled radar for monitoring glaucoma". https://www.ophthalmologytimes.com/journals/ophthalmology-times-journal?year=2020
Our research was highlighted in the "Review of Optometry": AI may soon predict the need for corneal surgery. https://www.reviewofoptometry.com/news/article/ai-may-soon-predict-the-need-for-corneal-surgery
Paper titled "Monitoring glaucomatous functional loss using an artificial intelligence-enabled dashboard" was accepted by Ophthalmology Journal
We received the best paper award from the 29th Annual Congress of the Iranian Society of Ophthalmology, for the paper titled "Glaucoma imaging signatures derived from fundus photographs using an artificial intelligence construct"
We will give two oral presentations and three poster presentations at ARVO and ARVO Imaging in the Eye 2020
Siamak was invited by the "Progress in Retinal and Eye Research" journal to write an editorial paper on Evolution of AI in ophthalmology
Paper titled "Predicting the likelihood of need for future keratoplasty intervention using artificial intelligence" was accepted by The Ocular Surface Journal
Siamak was invited by the Asia-Pacific Glaucoma Congress (APGC) 2020 to speak on AI in glaucoma
Siamak was invited by "Ophthalmology Glaucoma" to write an editorial paper on AI in glaucoma
Four abstracts were accepted by ARVO 2020
Our AI research was highlighted in Ophthalmology Times news: https://www.ophthalmologytimes.com/sites/default/files/legacy/mm/digital/media/OT111519_DIGITAL.pdf
Four abstracts were submitted to ARVO 2020; AI in glaucoma, AMD and cornea, covering both posterior and anterior segments of the eye.
Our AI research is receiving more national attention. We were featured in the Healio Ocular Surgery news: https://www.healio.com/ophthalmology/glaucoma/news/print/ocular-surgery-news/%7Bed5f7787-ecdb-4480-96c5-04d0a7ae8439%7D/ai-enabled-radar-a-new-tool-for-diagnosing-monitoring-glaucoma?page=1
Siamak will be moderating/organizing a session on AI in ocular epidemiology at the International Society for Eye Research (ISER), 2020, Argentina.
Siamak will be the PDEC liaison of the ARVO in the joint ARVO-AAO effort for developing educational webinars for residents and medical students in 2020.
Our abstract "Glaucoma imaging signatures derived from fundus photographs using an artificial intelligence construct" was accepted as an oral presentation at the 29th Annual Congress of Iranian Society of Ophthalmology.
Siamak will be the PDEC liaison of the ARVO's 2020 educational course on "AI in Ophthalmology".
Our AI work is receiving local attention. We are in the Daily Memphian news: https://dailymemphian.com/section/business/article/6502/uthsc-researcher-using-ai-to-identify-glaucom?utm_source=email_edition&utm_medium=email&utm_campaign=morning_2019-07-31#
We were highlighted in the UTHSC news: https://news.uthsc.edu/researcher-siamak-yousefi-phd-of-uthsc-aims-to-use-big-data-and-artificial-intelligence-to-detect-glaucoma/
Our abstract "Monitoring visual functional worsening in patients with glaucoma using an AI-enabled radar" was accepted as an oral presentation at American Academy of Ophthalmology (AAO) to be held in San Fransisco in October.
Anshul Thakur will join DM2L as a postdoctoral fellow in July.
Hassan Kabiri joined DM2L as a research fellow in May.
Siamak was an invited speaker for the 26th annual meeting of the "The Glaucoma Foundation; TGF" on "optic nerve rescue and restoration" taking place at NY in June 2019. He will be talking about "artificial intelligence and future directions in ophthalmology".
Siamak was an invited reviewer for the NIH Emerging Imaging Technologies in Neuroscience (EITN) Study Section in July 2019.
Siamak was an invited speaker for the "Midsouth Computational Biology and Informatics Society (MCBIOS)" symposium talking about "single-cell RNA-Seq analysis of retinal ganglion cells", Birmingham, Alabama, March 2019.
Siamak gave a talk on "identification of clinically relevant biomarkers of glaucoma on fundus images using deep learning" at the "Memphis Data Summit" in March 2019
Thanks to NIH for funding our R21 proposal. We are now looking for enthusiastic postdocs to join us; computer scientists, data scientists, Biostatisticians, or those in related areas are welcome to apply
Siamak was an invited reviewer for the NIH "Biomedical Computing and Health Informatics (BCHI)" and Emerging Imaging Technologies in Neuroscience (EITN) Study Sections in February and March 2019, respectively.
Siamak will be co-organizing ARVO "Artificial intelligence in ocular medicine: Seeing into the future" event taking place June 13 to July 1, 2019. More details at https://www.arvo.org/education/ai-online-event/
Article accepted: Promise of optical coherence tomography angiography in predicting glaucoma progression, JAMA Ophthalmology, 2019
Accepted as poster in ARVO 2019: Siamak Yousefi, Tobias Elze, Louis Pasquale, and Michael Boland, Glaucoma monitoring using an artificial intelligence enabled map, Association for Research in Vision and Ophthalmology (ARVO), 2019
Accepted as oral presentation in ARVO 2019: Mohammad Norouzifard, Ali Nemati, Reinhard Klette, Hamid GholamHossieni, Kouros Nouri-Mahdavi, and Siamak Yousefi, A hybrid machine learning model to detect glaucoma using retinal nerve fiber layer thickness measurements, Association for Research in Vision and Ophthalmology (ARVO), 2019
Accepted as poster ARVO 2019: Kouros Nouri-Mahdavi, Alessandro Rabiolo, Vahid Mohammadzadeh, Joseph Caprioli, and Siamak Yousefi, Machine Learning for Prediction of Visual Field Progression, Association for Research in Vision and Ophthalmology (ARVO), 2019
Accepted as poster ARVO 2019: Y. Arai, H. Takahashi, S. Yousefi, S. Inoda, H. Tampo, S. Sakamoto, Y. Matsui, H. Kawashima, and Y. Yanagi, “Estimation of best corrected visual acuity from optical coherence tomography images using deep learning”, Association for Research in Vision and Ophthalmology (ARVO), 2019
Accepted as poster ARVO 2019: Lu Lu, M. Hook, F. Xu, S. Yousefi, J. Yue, and R. Williams, “Identification of genes and miRNA influential on early and late glaucoma pathogenesis in DBA/2J mice”, Association for Research in Vision and Ophthalmology (ARVO), 2019
Our NEI R21 grant proposal received a descent score, we hope receiving NIH-NEI funds soon. November.
Two papers were submitted to the IEEE conference on Image and Vision Computing New Zealand (IVCNZ) 2018. September.
Paper titled "Rates of visual field loss in primary open-angle glaucoma and primary angle-closure glaucoma: Asymmetric patterns" was accepted by the IOVS journal.
Paper titled "Keratoconus severity identification using unsupervised machine learning" was accepted for publication in the PLoS One journal, the Spacial Issue on Machine Learning in Medicine. September.
Tarus Dukes, Computer Science Undergraduate Student, to join the lab as Data Manager and IT Specialist. August.
Paper titled "Rates of Visual Field Loss in Primary Open-Angle Glaucoma and Primary Angle-Closure Glaucoma: Asymmetric Patterns" was submitted to the IOVS journal. July.
Paper title "Keratoconus severity identification using unsupervised machine learning" was submitted to the PLOS One journal. July.
Paper titled "Detection of longitudinal visual field progression in glaucoma using machine learning" was accepted by the American Journal of Ophthalmology (AJO) journal. June.
Paper titled "Distribution and rates of visual field loss across different disease stages in primary open-angle glaucoma" was accepted by the Ophthalmology Glaucoma journal. June.
Paper titled "Estimating glaucomatous visual sensitivity from retinal thickness by using pattern-based regularization and visualization" was accepted by KDD. May.
Siamak gave a talk at the Machine Learning Interest Group of UTHSC on "progression of patterns: application to detecting glaucoma progression", Memphis, Tennessee. May.
Two research fellows will join our lab soon. April.
We will be giving a talk and presenting two posters at the ARVO/ARVO Imaging in the Eye, 2018. April.
Siamak gave a talk at the MCBIOS conference on "single-cell RNA-seq data analysis for identifying retinal ganglion cell sub-types", Starkville, Mississippi. April.
Siamak gave a talk at the Biostatistics Department of UTHSC on "high-dimensional sparse data mining; application to single-cell RNA-seq data". April.
Paper titled "Asymmetric Patterns of Visual Field Defect in Primary Open-Angle and Primary Angle-Closure Glaucoma" was accepted by the IOVS journal. March. Link to this article.
The abstract submitted to MCBIOS was selected for oral presentation. March.
Paper titled "Estimating Glaucomatous Visual Sensitivity from Retinal Thickness by Using Pattern-Based Regularization and Visualization" was submitted to KDD. February.
Submitted to MCBIOS 2018: single-cell RNA-seq glaucoma data analysis. February.
Submitted to ARVO/Imaging 2018: Corneal Data Mining. February.
To be presented at ARVO 2018: one oral and two poster presentations. February.
All ARVO and WOC abstracts were accepted. January.
Submitted to ARVO 2018: Impact of glaucoma on retinal ganglion cell subtypes: A single-cell RNA-seq analysis of the DBA/2J mouse
Submitted to ARVO 2018: Gaussian Mixture Model Expectation Maximization for Glaucoma Progression Detection: An open-source R Package
Submitted to ARVO 2018: Rates of Visual Field Loss in Primary Open-Angle Glaucoma
Submitted to WOC 2018: Patterns of Visual Field Defect in Primary Open-Angle Glaucoma
The Beta version of VFPro (Visual Field Progressor) R package was released. VFPro is an open-source software developed in our laboratory for monitoring and detecting glaucoma progression using longitudinal visual fields publicly-available on GitHub (https://github.com/DM2LL).
AAO 2017: Patterns of Visual Field Defects in Primary Open-Angle Glaucoma (POAG) and Primary Angle-Closure Glaucoma (PACG).["Highest viewed" poster among scientific posters in glaucoma, AAO 2017: https://aao.scientificposters.com/epsAbstractAAO.cfm?id=14]
Siamak presented a study on "Detection of longitudinal visual field change in glaucoma using a subspace-guided machine learning technique" at the NAPS Conference, Skaneateles, New York, 2017. November.
Co-investigator on the recently awarded P30 grant
IRB for "Ocular Data Mining" was approved
Data Mining and Machine Learning (DM2L) laboratory was established at the Hamilton Eye Institute, Department of Ophthalmology and Department of Genetics, Genomics, and Informatics, University of Tennessee Health Science Center in Memphis