Hoan Ngo Laboratory
Advanced Diagnostics for Health

Alzheimer’s disease (AD) diagnosis is important to treatment and care of AD patient. We aim to develop an AI system for automatic analysis of brain MRI images and predict state of memory of patient (healthy or AD). The AI system can be used as a tool to support medical doctors in diagnosis of AD through:

- Providing quantitative features of brain structure

- Providing objective evaluation of memory state of patient so that medical doctors can use as reference in diagnosis process

Skin diseases are among the most common of all human health afflictions and affect almost 900 million people in the world at any time (source: WHO). Meanwhile, many people around the world have smartphone equipped with camera. Our goal is to develop AI models for detection of skin diseases using images captured by smartphone camera.

​

Our current AI model for detection of Skin cancer based on a modified ResNet50 architecture achieves 88%. We are preparing a manuscript for the ISIC Skin Image Analysis Workshop @ CVPR 2020. Our AI model for acne grading currently achieves 85% accuracy. The model has been integrated into an iOS app named Skin Detective, which is available for download on App store.

The most important organs among five human senses are eyes, which are responsible for 80% of all impressions we perceive. Retina play a vital role in eyes’ functionality. Damage to retina can lead to permanent blindness.

​

OCTA is a functional extension of OCT that non-invasively provides images of the microvasculature of the retina and choroid without dye injection. OCTA is thus has great potential for studying and diagnosis of retinal diseases such as diabetic retinopathy, age-related macular degeneration, glaucoma, etc.

We are interested in developing both OCTA systems and OCTA image analysis algorithms for quantitative retinal vascular diseases diagnosis for infant and children (collaborate with Cynthia Toth, M.D., Vajzovic, M.D., and Tammy Hsu, medical student at Duke Eye Center).

Histopathology is still a gold standard in cancer diagnostic. However it requires extensive tissue preparation and staining. Full-field OCT provides fast non-destructive imaging of tissue at high resolution (~1 um) without staining. Furthermore, previous studies have shown that full-field OCT can differentiate malignant and benign tumor tissue. A typical full field OCT configuration is shown on the left (source: Advances in Optical Technologies 2012(1687-6393) DOI 10.1155/2012/435408).

We are interested in developing full-field OCT systems and Polarization-sensitive OCT systems for cancer diagnostic and intraoperative cancer margin delineation.

Alzheimer’s disease is the most common form of neurodegeneration with no cure and has severe impact to patient’s life quality. Previous reports have shown that OCT/OCTA has great potential for non-invasively diagnostic and monitoring of Alzheimer’s disease. Left image: normal, Right image: Alzheimer's (image source here) . 

We are interested in developing OCT/OCTA/multimodal retinal imaging systems and image analysis algorithms for Alzheimer’s disease diagnosis and monitoring.

The idea is to use light to locally activate drugs that accumulate at tumor site, thus killing tumor while leaving healthy adjacent tissues unharmed (image source here).

Microscopy with Ultraviolet Surface Excitation (MUSE) is a simple, slide-free technique that, within minutes, provides high-resolution diagnostic histological images resembling those obtained from conventional bright-field histology (see references 1 and 2, image source here).

​

We are interested in applying MUSE technique in low-resource settings in Vietnam.

Liquid biopsy is currently an active research area and has great potential in cancer diagnose, monitoring, and predicting survival without the need for invasive procedures. The technique is based on analyzing mutations and other changes in DNA released by tumor cells into the blood (known as circulating tumor DNA or ctDNA, image source here).

​

We are interested in developing liquid biopsy techniques for early cancer detection.

OCT is one of the most successful Biophotonics techniques in all over the world. However its high cost limits its use within large Eye Centers.

​

Photoacoustic tomography provide functional information over a wide range of spatial scales at high speed and is one of the fastest growing biomedical imaging technologies.

​

We are interested in developing low-cost OCT/OCTA/Photoacoustics systems for developing world.