Tobias Elze, Ph.D.

Harvard Medical School

Instructor in Ophthalmology, 

Schepens Eye Research Institute of Massachusetts Eye and Ear


Research Summary

Center/Research Area Affiliations


Dr. Elze is a computational vision scientist. His research addresses the methodology of basic and clinical vision science, such as optimal design of experiments and clinical studies and optimal data analysis (particularly for large and high-dimensional datasets). Dr. Elze's current projects include the functional characterization of eye diseases like glaucoma, age-related macular degeneration (AMD), and diabetic retinopathy, and the relationship between retinal structure and functional vision loss. Dr. Elze also develops adaptive sampling techniques for efficient clinical function testing and investigates display technology for clinical applications.

As a former member of the research group “Complex Structures in Biology and Cognition” at Max Planck Institute for Mathematics in the Sciences, Dr. Elze focused on methodology in visual neuroscience and psychophysics. In 2011, he joined the laboratory of Dr. Peter Bex at Schepens Eye Research Institute of Mass. Eye and Ear, where he started to investigate ophthalmic diseases using psychophysical and bioinformatical methods and machine learning. In 2013, he became an Instructor in Ophthalmology at Harvard Medical School, and in 2014 an Investigator at Schepens Eye Research Institute of Mass. Eye and Ear. He leads an interdisciplinary team of research fellows and students. His group works in the intersection among mathematics, computer science, and clinical ophthalmology.

Download his CV or biosketch [PDF] for more information.


Ph.D., Computer Science, Max Planck Institute for Mathematics in the Sciences (2011)

Postgraduate Training

Postdoctoral Fellow, Schepens Eye Research Institute, Harvard Medical School (2011-2013)


2013: Best Clinical Poster, Harvard Ophthalmology Annual Meeting and Alumni Reunion

His Story

Early Detection of Eye Diseases and of their Progression over Time

Many ophthalmic diseases have a long onset period before patients become aware of them. For instance, an estimated 25% of individuals with diabetes are undiagnosed in the United States. The disease silently affects their retina, often over many years. In fact, the retinal damage may affect functional vision so subtly that it initially goes unnoticed in the patient's daily life, requiring dedicated vision tests to be detected. Another eye disease with a high prevalence—glaucoma—often starts with small scotomas (blind spots) in the periphery, which are easy to ignore. Diseases like diabetic retinopathy or glaucoma are therefore often first noticed by the patient when he/she is already in a moderate-to-advanced state. At the same time, many ophthalmic diseases are irreversible, which makes early detection extremely important to timely initiate or adjust ocular therapy before diseases progress to stages that cause severe loss of quality of life.

Dr. Elze and his team aim to better characterize the functional loss from ocular diseases with the goal of developing novel diagnostic methods and improving the prediction of how some diseases may progress in future. For instance, he adjusted and applied statistical learning algorithms to identify representative patterns of vision loss from glaucoma. These patterns help to distinguish true glaucomatous functional loss from the various and frequent measurement artifacts and from damages caused by other eye diseases. Apart from the improvement of diagnostic and prognostic methods, Dr. Elze studies the optimal scheduling of patients, i.e., the optimal time intervals for patient testing to maximize the information gain of each test.

The translational potential of Dr. Elze's work related to diagnostic improvements is not only reflected by his publications, but also by the fact that he is co-inventor of two patents—namely one patent on modeling of glaucomatous visual field loss and another patent on an algorithm to determine the optimal scheduling intervals for patient testing.

Retinal Structure and Visual Function

Many ophthalmic diseases manifest themselves on the retina before functional deficits develop. Furthermore, even in the presence of functional damage, functional testing procedures can be time consuming and exhausting for patients. Ocular imaging, on the other hand, is convenient—even for patients at advanced age—and often takes only seconds to few minutes. Therefore, Dr. Elze and his team are investigating the relationship between damages of retinal structure and their precise effects on functional vision in eye diseases. In particular, Dr. Elze studies patient measurements from fundus images and optical coherence tomography and their association with specific effects on functional vision. For instance, his team recently identified an important retinal biomarker specifically for central vision loss in glaucoma.

How Lifestyle Affects the Eye

Many lifestyle-related parameters, such as obesity, high blood pressure, unfavorable cholesterol values, the lack of physical activity, or the consumption of alcohol or tobacco, may impair retinal structure. While lifestyle-related retinal changes are often too small to be noticed by people, they may impose risk factors for severe eye diseases, including potentially blinding optic neuropathies. Dr. Elze and his team are systematically investigating how aging and lifestyle factors affect the retina. They are participating in a population-based study (with about 10 thousand participants) that includes ocular imaging and a large number of physiological and cognitive parameters.


Research Interests

  • Methodology of basic and clinical vision science
  • Functional characterization of ophthalmic diseases
  • Relationship between retinal structure and vision loss in eye diseases
  • Relationship between lifestyle and the retina

Detection of Ophthalmic Vision Loss and its Progression over Time

Many ophthalmic diseases have long onset periods during which their functional impairments are minor and unnoticed by the patient. Similarly, the diagnosis or functional worsening of an eye disease is challenging, as measurements are noisy, and the differentiation between true disease progression and random fluctuations or measurement artifacts is difficult.

For this project, Dr. Elze is applying bioinformatical methods to large sets of patient measurements. His goal is to develop novel methods to diagnose the onset and the progression of ophthalmic diseases like glaucoma, retinitis pigmentosa, or diabetic retinopathy. As an example from his previous achievements, he developed and evaluated a novel scheme of representative patterns for glaucomatous vision loss based on unsupervised machine learning.

Relationship between Retinal Structure and Visual Function in Eye Diseases

While detailed functional measurements of ophthalmic vision loss are often noisy, time consuming, and exhausting for the patient, ocular imaging of retinal structure can often be performed conveniently for the patient within minutes or even seconds.

Dr. Elze is applying image processing and statistical learning methods to large sets of paired structure-function measurements of several eye diseases (diabetic retinopathy, glaucoma, macular degeneration) to predict the details of functional vision loss from two- or three-dimensional retinal imaging, like fundus photography or optical coherence tomography. As an example from previous achievements, Dr. Elze identified a novel retinal biomarker (central retinal vessel trunk location) for central vision loss in glaucoma.

Relationship between Aging, Life Style, and the Retina

Aging and lifestyle effects are reflected in the eye, and retinal changes related to lifestyle parameters or aging may be risk factors for eye diseases or confounders for their diagnosis.

In this project, Dr. Elze is participating in a population-based study with about 10,000 participants to systematically investigate the relationship between retinal parameters obtained via fundus photography and optical coherence tomography and age, as well as various groups of lifestyle-related variables, such as cardiovascular parameters (e.g., blood pressure, cholesterol, history of strokes), anthropometric parameters (e.g., body mass index, waist-to-hip ratio), substances (e.g., alcohol and tobacco consumption), or neurological diseases (e.g., Parkinson’s disease, neuro-cognitive disorders, multiple sclerosis).

Current Research Funding

Research to Prevent Blindness: Principal Investigator (PI)
Association between retinal structure and age-related impairments
Dr. Elze is developing eye imaging methods to detect impairments associated with age-related diseases, such as glaucoma and AMD, at earlier stages than currently possible.
National Institutes of Health: Co-Investigator
Impact of peripheral islands in the visual field on functional ability in patients with retinitis pigmentosa
Dr. Elze is studying the frequency and spatial configuration of peripheral islands in patients with retinitis pigmentosa and how they affect quality of life.
BrightFocus Foundation: PI
Computational investigation of glaucoma progression
Dr. Elze is developing novel computational methods to assess the progression of glaucoma in retinal structure, visual function, and their combination. Machine-learning techniques are applied to large datasets of Humphrey visual fields and optical coherence tomography measurements of glaucoma patients from different hospitals to re-define, predict, and quantify disease progression over time.



8 (Google Scholar, as of May 2017)

Selected Publications

Dr. Elze has published more than 20 peer-reviewed articles. Below is a list of selected publications. View his publications on PubMed, Google Scholar, or ORCID.

  1. Wang M, Wang H, Pasquale LR, Baniasadi N, Shen LQ, Bex PJ, Elze T. Relationship between Central Retinal Vessel Trunk Location and Visual Field Loss in Glaucoma. Am J Ophthalmol. 2017 Apr;176:53-60.
  2. Elze T Pasquale LR, Shen LQ, Chen TC, Wiggs JL, Bex PJ. Patterns of functional vision loss in glaucoma determined with archetypal analysis. J R Soc Interface. 2015 Feb 6;12(103)
  3. Poppe S1, Benner P, Elze T. A predictive approach to nonparametric inference for adaptive sequential sampling of psychophysical experiments. J Math Psychol. 2012 Jun 1;56(3):179-195.
  4. Elze T. Achieving precise display timing in visual neuroscience experiments. J Neurosci Methods. 2010 Aug 30;191(2):171-9.
  5. Elze T, Tanner TG. Liquid crystal display response time estimation for medical applications. Med Phys. 2009 Nov;36(11):4984-90.