Research Projects - Optical Imaging and Visualisation

Project 1) Machine learning approaches in ocular images analysis: Automated detection and diagnosis

With the increasing prevalence of ocular diseases like glaucoma, diabetic retinopathy and age-related macular degeneration; annual screening for ocular diseases by a human expert, grading of retinal images is challenging. Automated retinal image assessment systems (ARIAS) may provide a clinically effective and cost-effective detection of ocular diseases. Recently, machine learning approaches have become increasingly successful in image-based diagnosis, disease prognosis, and risk assessment. This research project will highlight new research directions and examine the main challenges related to machine learning in ocular imaging, applying novel deep learning algorithms to automatic analysis of both digital fundus photographs and OCT images from both healthy control subjects and patients undergoing treatment for relevant ocular diseases. On top of existing clinical markers of disease, this project will have a focus on developing novel metrics for distinguishing between glaucomatous and non-glaucomatous eyes.

Project 2) 3D reconstruction and visualisation of OCT images in the virtual reality system

3D reconstruction of medical images helps in image interpretation with visualising depth and understanding the underlying pathological process in disease. Additionally, the modern technology of 3D visualisation with virtual or augmented reality makes the treatment and diagnosis process very faster and more comfortable even in a surgical setting for all clinicians. It also helps the patient to understand the state of his/her disease very evidently. In this project, the 2D slices of the anterior chamber angle of glaucoma patients will be collected from optical coherence tomography (OCT) images and reconstructed with volume rendering process in different 3D software, and afterwards, the 3D model will visualise with the virtual reality devices. Moreover, the project will also comprise some 3D animation and roaming in the visualisation environment to make it more realistic for the understanding of disease for diagnosis purpose. One of the goals of this project is to measure its impact on patient health literacy, to determine if this method can be used to enhance patient understanding of ocular disease processes.

Project 3) Visualising Tear Film Lipid Layer Using Quantum Dots

The effective treatment of ocular diseases requires accurate diagnosis and precise monitoring of ocular structures. The development of new therapeutics has been a significant focus of pharmaceutical research in the area of vision sciences. Quantum dots (QDs) are semiconductor nanocrystals that can provide a range of diagnostic and therapeutic applications in biology as well as ophthalmology. Regarding imaging of various parts of the eyes, QDs have been used for uveal melanoma, tracking of uveal flow for the treatment of glaucoma and better visualization. All these researches were focused on the posterior of the eyes for various purposes. This project is focused on one of the anterior layers of the eye that is tear film specifically lipid layer. Using the optical properties of the QDs, it is aimed to have a better visualisation of this layer whose composition is known but the structural arrangement and interfacial interactions are still under research. Various molecular and dynamic models have been proposed to illustrate the dynamic structural arrangement of lipids in this layer. These simulations need to be experimentally verified to develop authenticated structural models of lipids and their replenishment.

Recent work form OIV Lab

  1. Press conference at American Academy of Optometry: A Feature Agnostic Based Glaucoma Diagnosis from OCT Images from Deep Learning Technology - Maitreyee Roy & Nahida Akter
  2. Imaging of Tear Film Lipids Using Quantum Dots - OSA
  3. Modelling the effect of commercially available blue‐blocking lenses on visual and non‐visual functions
  5. The effect of blue-blocking lenses on photostress recovery times for low and high contrast chromatic and achromatic stimuli
  6. A Feature Agnostic Based Glaucoma Diagnosis from OCT Images with Deep Learning Technique
  7. Analysis of OCT Images to Optimise Glaucoma Diagnosis
  8. Quantum Dots in Ophthalmology: A Literature Review