Related Experiment Video
Updated: Nov 24, 2025
Author Spotlight: An Automated Method for Assessing Visual Acuity in Infants and Toddlers Using an Eye-Tracking System
Published on: March 17, 2023
Automatic eye localization for hospitalized infants and children using convolutional neural networks
Vanessa Prinsen1, Philippe Jouvet2, Sally Al Omar3
1École de technologie supérieure, 1100 Notre-Dame St W, Montréal, Québec H3C 1K3 Canada; CHU Sainte-Justine, 3175 Chemin de la Côte-Sainte-Catherine, Montréal, QC H3T 1C5.
Insights
Developing custom datasets and models significantly improves eye localization for pediatric patients in intensive care units (ICUs). This enhances clinical decision support and patient monitoring accuracy in challenging hospital environments.
Area of Science:
- Computer Vision
- Medical Imaging
- Artificial Intelligence
Background:
- Accurate eye localization is crucial for pediatric patient monitoring and clinical decision support.
- Existing eye localization models perform poorly in pediatric hospital settings due to variations in appearance and medical equipment.
Purpose of the Study:
- To develop and evaluate improved eye localization models for pediatric intensive care units (ICUs).
- To address the limitations of existing models in busy pediatric hospital environments.
Main Methods:
- Created two datasets: one for training using public data and another for testing with pediatric ICU recordings.
- Trained two eye localization models using the Faster R-CNN algorithm and a pre-trained ResNet base network.
Main Results:
- A convolutional neural network trained on combined adult and child data achieved a 79.7% eye localization rate.
- Image contrast equalization further improved the localization rate to 84%.
Conclusions:
- Convolutional neural networks show promise for eye localization and tracking in pediatric ICUs, even with limited data.
- Task-specific datasets and transfer learning are effective for developing specialized clinical computer vision applications.
Background:
Reliable localization and tracking of the eye region in the pediatric hospital environment is a significant challenge for clinical decision support and patient monitoring applications. Existing work in eye localization achieves high performance on adult datasets but performs poorly in the busy pediatric hospital environment, where face appearance varies because of age, position and the presence of medical equipment.
Methods:
We developed two new datasets: a training dataset using public image data from internet searches, and a test dataset using 59 recordings of patients in a pediatric intensive care unit. We trained two eye localization models, using the Faster R-CNN algorithm to fine-tune a pre-trained ResNet base network, and evaluated them using the images from the pediatric ICU.
Results:
The convolutional neural network trained with a combination of adult and child data achieved an 79.7% eye localization rate, significantly higher than the model trained on adult data alone. With additional pre-processing to equalize image contrast, the localization rate rises to 84%.
Conclusion:
The results demonstrate the potential of convolutional neural networks for eye localization and tracking in a pediatric ICU setting, even when training data is limited. We obtained significant performance gains by adding task-specific images to the training dataset, highlighting the need for custom models and datasets for specialized applications like pediatric patient monitoring. The moderate size of our added training dataset shows that it is feasible to develop an internal training dataset for clinical computer vision applications, and apply it with transfer learning to fine-tune existing pre-trained models.
