Investigative Ophthalmology & Visual Science Cover Image for Volume 65, Issue 7
June 2024
Volume 65, Issue 7
Open Access
ARVO Annual Meeting Abstract  |   June 2024
Improving laser targeting accuracy with Augmented Reality guidance in retinal laser therapy
Author Affiliations & Notes
  • Sangjun (Sarah) Eom
    Electrical and Computer Engineering, Duke University, Durham, North Carolina, United States
  • Miroslav Pajic
    Electrical and Computer Engineering, Duke University, Durham, North Carolina, United States
  • Maria Gorlatova
    Electrical and Computer Engineering, Duke University, Durham, North Carolina, United States
  • Majda Hadziahmetovic
    Department of Ophthalmology, Duke University, Durham, North Carolina, United States
  • Footnotes
    Commercial Relationships   Sangjun (Sarah) Eom None; Miroslav Pajic None; Maria Gorlatova None; Majda Hadziahmetovic None
  • Footnotes
    Support  None
Investigative Ophthalmology & Visual Science June 2024, Vol.65, 793. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Sangjun (Sarah) Eom, Miroslav Pajic, Maria Gorlatova, Majda Hadziahmetovic; Improving laser targeting accuracy with Augmented Reality guidance in retinal laser therapy. Invest. Ophthalmol. Vis. Sci. 2024;65(7):793.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : Safe and efficacious use of laser indirect ophthalmoscope (LIO) in retinal laser therapy is challenging and requires prolonged training. To ease the training and improve safety we propose an Augmented Reality (AR) guidance system with a custom-built indirect-ophthalmoscope-integrated AR headset to assist the laser targeting tasks.

Methods : We developed an AR guidance system using the Microsoft HoloLens 2 and Raspberry Pi with a camera sensor for feature matching. We integrated an indirect ophthalmoscope into an AR headset to overlay AR holograms of magnified retinal landmarks on the retinal images using matching feature points. Employing thresholding-based image processing, we detected the laser-targeting locations and provided images (e.g., target sign or color) and text guidance in AR on the regions that needed to be treated or avoided.

Results : A total of 14 non-experts participated in the user study for the simulation of retinal laser therapy on a 6x retina phantom model (25.4±5.32 years, 78.6% male). The AR guidance helped users improve their mean laser targeting accuracy from 7.13mm±4.22mm to 5.42mm±3.06mm p=0.0057. The total completion time in targeting five simulated points on retinal images increased from 179.93s±100.19s to 377.57s±199.08s. Our survey results show that users agreed that AR guidance was helpful in improving laser targeting accuracy (64.3%; mean score, 3.86; 95% CI, 0.40).

Conclusions : The overlay of magnified retinal landmarks and guidance on laser detection demonstrated the improvements in laser targeting tasks on non-expert users. We plan to conduct further expert (e.g., retina specialists) user studies and work on improving the prototype and robustness of the image registration. The AR-guided LIO might help improve accuracy and safety and make new workforce training easier.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

 

The hardware setup of our AR guidance system includes a custom-built indirect-ophthalmoscope-integrated HoloLens 2 to visualize retinal landmarks and textual guidance in holograms; we use a 6x retinal phantom with a laser pointer to simulate the laser treatments.

The hardware setup of our AR guidance system includes a custom-built indirect-ophthalmoscope-integrated HoloLens 2 to visualize retinal landmarks and textual guidance in holograms; we use a 6x retinal phantom with a laser pointer to simulate the laser treatments.

 

Five target points given to participants for aiming with the laser pointer are visualized in the participant’s AR view. The mean laser targeting accuracy improved from 7.13mm to 5.42mm shown with p-value notation as follows: * p < 0.05, ** p < 0.01, and *** p < 0.001.

Five target points given to participants for aiming with the laser pointer are visualized in the participant’s AR view. The mean laser targeting accuracy improved from 7.13mm to 5.42mm shown with p-value notation as follows: * p < 0.05, ** p < 0.01, and *** p < 0.001.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×