Staff Scientist 1

This position is located in the Clinical Trials Branch of the Division of Epidemiology and Clinical Applications (DECA), National Eye Institute (NEI). The candidate will be someone with advanced skills in data science that can provide support in the development of our research program and staff projects.  They will hold a PhD degree or equivalent and have significant post-doctoral experience.

They will have a proven record of publications that provide evidence of their expertise using computer programming in one or all the following languages: Python, MATLAB, C++, R, and/or Java.  The prospective candidate will have experience in image analysis (registration, segmentation, and/or tracking) in medical images, and with experience in advanced modeling techniques, with priority given to those with the ophthalmological modalities listed above.

One of the major research focuses of the Clinical Trials Branch is the development of outcome measures in clinical trials of retinal disease, which include assessments with multi-modal imaging and some functional changes.  The Clinical Trials Branch conducts prospective clinical studies in normal and abnormal functioning of the retina, particularly those with retinal degenerations – including complex retina diseases, such as age-related macular degeneration (AMD), drug toxicities, and other neurodegenerative diseases. Acquired images are from the following modalities: optical coherence tomography (OCT) and angiography, fundus color photographs, fundus autofluorescence (FAF), and adaptive optics (AO).   Multiple measures of retinal function are also acquired in parallel with retinal imaging. Accumulated datasets of longitudinal, multimodal images collected prospectively are used to answer both hypothesis-driven and hypothesis-generating research to add to the understanding of the anatomic changes occurring in retinal disease. The goal is to develop automated metrics that can be used as outcome measures in clinical trials of retinal diseases. 

The candidate will operate as an independent researcher within the Clinical Trials Branch while consulting with other investigators within DECA and the NEI and established experts in governmental, academic, and industrial research sectors.

The scientist will primarily focus his/her efforts to develop automatic segmentation algorithms and machine learning approaches to detect/segment/quantify ocular anatomical/pathological structures seen on ophthalmic imaging systems (such as Optical coherence tomography [OCT], color fundus photographs, fundus autofluorescence, and adaptive optics images). Analyses will objectively detect and evaluate the biomarkers for onset and progression of the retinal diseases studied. Computational analyses will involve combinatorial assessment of disparate data types (i.e. patient demographics, risk factors, genetic information, other relevant data).  In addition, there will be datasets that will include the development of algorithms based upon retinal imaging (color fundus photographs, fundus autofluorescence, OCT) to predict systemic diseases such as cognitive impairment, diabetes, cardiovascular disease, hypertension and others. 

As a scientist in the Clinical Trials Branch, the scientist will work with data from ongoing and future clinical studies of Age related macular degeneration (AMD) (NCT03225131), hydroxychloroquine toxicity (NCT01145196), Macular telangiectasia type 2 (NCT )and retinitis pigmentosa (NCT03845218) which will be utilized for the development and testing of these algorithms.  Some projects will be listed here but not limited to these: i) Develop registration algorithm that aligns multi-modal retinal images collected from longitudinal clinical studies to achieve accuracy and robustness required for analysis of structural changes in large-scale clinical data. The goal of aligning multimodal images is so that local changes can be investigated across the different acquisition types to synthesize and integrate the structural information.  Alignment over time will enable the development of predictive models. (ii) The development of automated detection algorithms for OCT analyses.  These will encompass both segmentation and semantic approaches. OCT features such as the ellipsoid zone (EZ) have shown to be key anatomic structures reflecting the integrity of photoreceptors.  Given the monotonic loss of photoreceptors in retinal degenerations, the EZ reflectivity band has become an attractive structural measure that has gained FDA support.  (iii) the development of predictive models to describe retinal function from structural measures as well as models for disease progression.  Data collected across several different studies will be used to develop models that have applicability across diseases. This will be used to answer both hypothesis-driven and hypothesis-generating research to add to the understanding of the anatomic changes occurring in retinal disease. The scientist will use such inferences to communicate results in peer-reviewed manuscripts and as a guide in planning subsequent research projects.

In addition to the above-mentioned studies, it should be noted that the candidate may also collaborate with other investigators within NEI or even with our extramural investigators who are our collaborators who might have other medical images, including some basic science laboratories with images of human cells, etc.  There should be an array of projects with medical images that would be available to the candidate. 

The candidate will also apply knowledge and skills in research design to establish and maintain feasible, sustainable, standardized, valid and reliable systems of data and image collection, storage, monitoring, and management.  Facilitates image extraction from clinical instruments in formats that can be utilized by custom algorithms.  Uses scientific expertise to ensure the nature of data will allow characterization of key factors and processes implicated in health and disease of the visual system.

Forecasts to prevent avoidable problems in the post-implementation phase and troubleshoots to address unavoidable problems.  Has authority to address and solve problems as they occur and to alert supervisor when deemed necessary.

Knowledge and optimization for CUDA and/or OpenCL workflows and applications is desirable.  Submits request for changes in computing and other resources when state-of –the art advances in technology allow more effective and efficient fulfillment of research needs. 

DHHS and NIH are equal opportunity employers.