Researchers at the University of Hong Kong (HKU) developed an artificial intelligence (AI)-powered imaging tool that enables rapid and precise diagnosis of cancer patients.
The research was conducted by a team of scientists from HKU’s Li Ka Shing Faculty of Medicine (HKUMed) and Queen Mary Hospital.
The team was led by HKU Faculty of Engineering biomedical engineering programme director Kevin Tsia.
In the research, the team used Cyto-Morphology Adversarial Distillation (CytoMAD), their generative AI (GenAI) technology for lung cancer patients and drug tests.
CytoMAD technology, together with unique microfluidic technology, enables quick and cost-effective imaging of human cells for tumour assessment at the precision of individual cells.
The technology also uses AI to automatically correct cell imaging inconsistencies, enhance cell images, and extract previously undetectable information from cell images.
It ensures accurate and reliable downstream data analysis and diagnosis and can advance cell imaging for analysis of cell properties and related health and disease information.
CytoMAD developer, HKU department of electrical and electronic engineering postdoctoral researcher Michelle Lo said: “Our work primarily focuses on label-free imaging modalities (i.e. bright-field (BF) to quantitative phase image (QPI) translation) due to their growing significance in biomedicine in the recent year.
“A classical bright-field cell image typically looks like a vague photo full of scattered fainted blobs – nowhere close to informative for meaningful analysis of the cell properties and thus the related health and disease information.
“Nevertheless, CytoMAD, as generative AI model, can be trained to extract the information related to mechanical properties and molecular information of cells that was undetectable to the human eye in a brightfield image.”
According to the researchers, the low visibility of cell samples placed under the microscope led to the method of applying stains and labels to the samples.
The traditional approach is time-consuming and is not cost-effective, which requires patients to wait for prolonged periods before the results of their cell analysis.
The CytoMAD technology provides a significant advantage of label-free operation, which requires fewer steps to prepare patient or cell samples.
The technology also addresses the ‘batch effect’ challenge, which involves technical variations from different experimental conditions.
The deep-learning model, which is also developed by Tsia’s team and enables ultrafast optical imaging technology, is not limited to lung cancer patients.
Professor Tsia said: “Until now, there was no cost-effective technique to do single-cell analysis through imaging mainly because of the limitation in scale.
“Under the traditional methods, the imaging throughput is not fast enough, and the cell images are not clear and informative enough.
“We use Generative AI technology to render much clearer label-free images with useful information such as whether a treatment has had a positive effect.
“Our AI model doesn’t require the need for any assumption. Hence, it allows unbiased cell image analysis and diagnosis.”
