The Biomedical Engineering Department (BME) is partnering with Cao Thang Eye Hospital (CTEH) in research and development of Artificial Intelligence (AI) in Ophthalmology.
As a part of this effort, the new AI for Health research group is now launched with an aim to leverage the power of AI in Healthcare. The group have been developing AI applications for skin cancer detector and eye disease identification, with major focus on the later.
The ophthalmology team at BME will utilize the de-identified ophthalmic data to conduct related research with clinical and medical consultancy from ophthalmologists at CTEH. It’s hoped that via this collaboration, BME and CTEH will build an assistive tool that effectively aids doctors in making diagnosis.
Overview of Deep Learning methods in Skin Cancer Detection [*]
The new research group will be led by Dr. Ngo Thanh Hoan, Vice-Chair of Biomedical Engineering Department. Dr. Hoan received PhD degree in Biomedical Engineering from the Duke University, North Carolina, US in 2017 under supervision of professor Tuan Vo Dinh, Director of the Fitzpatrick Institute for Photonics at Duke University.
Upon graduation, Dr. Hoan did postdoctoral research at Duke Eye Center where he focused on the development of systems and software for eye imaging and eye image analysis before returning to Vietnam to lead the Clinical Engineering Laboratory in Biomedical Engineering Department of International University.
Together with Dr. Hoan, Prof. Vo Van Toi also joins the group in the advisory role. Professor Toi is an emeritus professor of Tufts University in Massachusetts, USA and former executive director of Vietnam Education Foundation.
After many years living abroad, he returned to Vietnam to set up the first Biomedical Engineering undergraduate program in Vietnam in 2009. Professor Toi has conducted many researches related to the eye. He used to be director of an Eye Institute in Sion, Switzerland. During his time in the US, he invented many devices for eye research.
In this example, the image showed signs of diabetic retinopathy.
The ophthalmology team develops the algorithm for classifying various retinal diseases using fundus image. The interface shown is from the web-app designed for users. First, the user uploads a retinal image they want to analyze. After the user clicks “Predict” button, the AI runs and output diagnostic results.