Skip to main content

Research Projects

women look at data on a computer

Visual Remediation in Schizophrenia 

Funded by NIMH. PI, Steven M. Silverstein, PhD.

Many people with schizophrenia experience changes in visual perception.  Two well-established changes involve reduction in contrast sensitivity (the ability to detect subtle differences in brightness and shading) and in perceptual organization (the ability to automatically organize the parts of an object or image into a single mental representation).  The purpose of this 5-year NIMH-funded project is to determine the effectiveness of three different training programs to improve visual perception in schizophrenia by targeting contrast sensitivity, perceptual organization, or both.  

These visual training packages will be compared to changes that occur after a course of standard cognitive remediation.  The study involves a comprehensive assessment of different aspects of vision and other functions that depend on vision (e.g., visual memory, reading), and a 30-40 session training program that uses an engaging computer game-like format.  Assessment sessions happen before the training; after 20, 30, and 40 sessions; and 6-months after completion of the training program.  The program is described in detail in a recent publication that can be accessed here:  The registered ID is NCT03314129 


eye scan macular cube image

Retinal Vasculopathy as a Proxy for Cerebral Vascular Health and a Marker of Cognitive Improvement in Advanced Heart Failure Patients after LVAD Implantation

Funded by the Chernowitz Medical Research Foundation

The purpose of this project is to evaluate changes in retinal microvasculature as a biomarker of changes in cerebral blood flow and cognition in heart failure patients from before to after left ventricular assist device (LVAD) implantation. The retinal vasculature measure we will use is optical coherence tomography angiography (OCTA). 

We will compare three groups of people: heart failure patients who receive LVAD, heart failure patients who do not receive LVAD, and a group of people with non-proliferative diabetic retinopathy. Groups will be compared at three timepoints corresponding to pre-LVAD implant for the LVAD group, and then three and six months later. 

The specific aims of the study are as follows:  

  1. To determine whether cognition and cerebral blood flow (CBF) improve after LVAD implantation, and whether the extent of post-LVAD-implant CBF change predicts change in cognition scores over time;  

  1. To determine the degree to which measures of retinal vasculature and neural layer thickness predict baseline data on CBF and cognition, and whether changes in retinal vasculature and neural layer thickness from before to after LVAD implant predict the extent of change on CBF and cognition during this time period; 

  1. To determine whether retinal biomarkers of brain health predict changes in patient quality of life and other aspects of cardiac disease-relevant psychological functioning (e.g., levels of depression symptoms) from pre- to post-LVAD implant; and  

  1. To determine the degree to which: measures of neural functioning in the eye (considered to be a proxy of brain neuronal health) are affected by LVAD treatment, and changes in retinal function improve the prediction of improvements in cognitive and overall psychological functioning in LVAD patients, over and above prediction achieved by retinal vascular measures.  


women looking into eye machine while male researcher views results on screen

Retinal Vasculopathy as a Predictor of Cardiovascular Disease and Risk in Individuals with Early Psychosis​

Funded by the URMC Department of Psychiatry Internal Grants Program. PI, Joy Choi, MD.

This project examines microvascular changes that impact cardiovascular health in early psychosis with the goal of developing a foundation for primary prevention of cardiovascular disease (CVD) in individuals with psychotic illness. CVD is a major cause of death in individuals with schizophrenia. The exact pathogenesis of CVD in this population is not yet understood but is thought to be multifactorial in origin including genetic predisposition, metabolic sequelae of antipsychotic medications, and unhealthy lifestyles. The inflammatory-vascular model of schizophrenia, supported by the discovery of genes common to both schizophrenia and CVD, hypothesizes that schizophrenia is a neuropsychiatric manifestation of microvascular disease within the central nervous system (CNS) with comorbid systemic microvascular abnormalities. Based on this model, we hypothesize that microvascular changes are present early in the course of schizophrenia in the CNS, and that these correlate with structural and/or functional impairments in the heart. We will use an innovative approach to examining CNS vasculopathy via optical coherence tomography angiography (OCTA), which allows for rapid, high-resolution visualization of the retina – the only part of CNS that can be examined in a non-invasive way. Specifically, we will evaluate in the early psychosis population: 1) abnormality in cardiac structure and function using transthoracic echocardiography, 2) abnormality in retinal vasculature, 3) 10-year CVD risk elevation and 4) the relationship among these measured variables. 


Woman looking into eye scan

Eyeing the Link Between Heart and Brain Health

Funded by the American Heart Association Innovative Project Award. PI, Mark Oldham, MD.

We are exploring the novel application of retinal biomarkers to study delirium risk as well the underlying neural and neurovascular basis for delirium and its subsequent cognitive and functional outcomes. In this study, we are enrolling patients preparing to undergo CABG surgery and evaluating whether the results of retinal exams can identify patients with cognitive impairment before CABG surgery and predict patients with postoperative delirium and cognitive decline. Specifically, we are using retinal optical coherence tomography (OCT), OCT angiography (OCTA) at rest and during vascular reactivity, and flash electroretinography (ERG), which provide direct, efficient, non-invasive measures of neural and neurovascular structure and function in the CNS. The potential application of retinal measures to cognitive health in patients with heart disease could provide a quick and scalable approach to evaluating cognitive vulnerability in this and other populations. Our specific aims are as follows:

Aim 1: To study the association between baseline cognition and ocular biomarkers.

Subjects with lower pre-CABG cognitive performance will have:
H1a. Thinner retinal nerve fiber layer adjacent to optic nerve head, on OCT.
H1b. Reduced retinal perfusion, on OCTA.
H1c. Reduced retinovascular reactivity, on OCTA using a breath-hold maneuver.
H1d. Impaired retinal ganglion cell function, on ERG.

Aim 2: To study baseline ocular biomarkers in relation to post-CABG delirium and severity.

H2a. Post-CABG delirium will be associated with the four retinal indices above.
H2b. Post-CABG delirium severity will be associated with the four retinal indices above.

Aim 3: To explore ocular biomarkers in relation to post-CABG neuropsychiatric outcomes.


Other studies focused on schizophrenia and/or bipolar disorder that are happening at the University of Rochester Medical Center can be viewed at on the Schizophrenia Treatment and Research Laboratory (STAR) website

Past Projects

The research projects listed below have concluded. 

2 eye scan macular cube imaging

Retinal Biomarkers of Concussive and Subconcussive Head Injury 

Funded by the Schmitt Program on Integrative Neuroscience [SPIN] through the Del Monte Institute for Neuroscience at the University of Rochester Medical Center

This project will evaluate college football players with a history of concussion, college football players without a history of concussion, and non-contact college athletes (e.g., on swim and track and field teams).  Measures of retinal structure and function, along with structural brain imaging and cognitive testing, will occur prior to the football season, at the end of the season, and then six months later.  Our goal is to determine whether concussions, but also the combined effects of head hits during practices and games (measured using accelerometers placed within helmets), lead to changes in cognition and brain structure that can be observed either immediately after the season or several months later.  We are also interested in determining whether retinal markers of neural structure and blood flow can predict baseline levels of brain structure and cognition, and changes from before to after the football season.

According to the Centers for Disease Control and Prevention (CDC) between 1.6 and 3.8 million concussions occur each year in the United States.  Many of these occur during sports, and up to 10% of athletes experience at least one concussion in any single season.  The majority (69%) of youth athletes with a possible concussion continue to play despite experiencing concussion symptoms, and in many such cases (40%), coaches are unaware that athletes may have had a concussion.  Importantly, a single season of football or soccer, even in the absence of a concussion, can cause changes in brain structure. Thus, there is an urgent need for rapid and mobile methods for detecting significant brain trauma in the immediate aftermath of head impact (e.g., on the field), as well as tests that can be used to screen for longer-lasting, permanent, and/or progressive changes that would require lifestyle adaptations and possibly treatment. The proposed research aims to close the gap between our current state of knowledge and the development of such assessments.   


Women having an eye exam

Feasibility Study of the Ability of ERG Components to Distinguish Schizophrenia and Bipolar Disorder Type I 

Funded by diaMentis, Inc.

The purpose of this project is to determine if patterns of retinal functioning (reflecting sensitivity of cells at the back of the eye to light) are different in schizophrenia and bipolar disorder, both from each other and from samples of people without either condition. This study is based on a growing body of work indicating that changes in brain function in schizophrenia and bipolar disorder have parallels in changes in the functioning of certain retinal cells, which, like brain cells, are neurons. The study involves four visits: one is for gathering information about psychiatric history and current symptoms, and the other three visits involve testing using electroretinography (ERG). 

ERG data will be collected on three devices.  At the end of the study, machine learning algorithms will be used to determine which variables, in which combinations, best identify people in one group distinct from the other groups.  The goal long-term goal of the study is to develop an ERG-related diagnostic aid that can be used to increase confidence in a diagnosis of bipolar disorder vs. schizophrenia in new onset cases where the diagnosis is not clear based on symptoms alone.