A research team at the University of Hong Kong (HKU) has developed an artificial ielligence-based tool called “Cyto-MORPHology Adversarial Dist from Cytomad for quick and accurate diagnosis of cancer.
The technology, led by Professor Kevin Shatia of the University’s Faculty of Engineering, is capable of accurately analyzing single -celled analysis without the need for traditional labeling techniques. The technology has been tested in collaboration with the Lee Kuching Medical School and the HKU Hospital and has been effective in evaluating paties with lung cancer and drug screening processes.
Cytomad is capable of improving cellular imagery and automatically modifies the inconsistencies and increases the resolution of the image. This progress will analyze more credible data and better medical decision making. This technology connects to a dedicated microfluid system that allows for fast and cost -effective imaging of human cells. By offering high -resolution single -celled imaging, it helps physicians to evaluate tumor properties and examine the risk of metastasis.
Traditional imaging requires staining and labeling cell samples that are time consuming and costly. Cytomad eliminates this need and simplifies the sample preparation process and accelerates diagnostic workflows. This model of artificial ielligence turns conveional images of the bright field io more accurate displays and reveals cellular features that are usually not analyzed.
One of the problems with existing cellular imaging methods is their dependence on slow and costly processes that may delay critical therapeutic decisions. Many of the available solutions require fluorescence markers, which lead to additional steps and costs. Cytomad offers a labelless alternative that reduces these limitations and maiains accuracy. Using the manufactured artificial ielligence, this system converts low corast images io more information visuals that provide deeper insights io cell morphology without the need for chemical staining.
Another challenge in cellular imaging is the changes in the differences in the configurations of the imaging equipme and protocols, which are often known as “batch work”. These inconsistencies can disrupt accurate biological ierpretation. Many of the existing machine learning solutions depend on predetermined assumptions about data, but Cytomad remains this limit by functioning without predetermined data.
The system uses an ultra -fast optical imaging technology developed by the research team and is capable of capturing millions of cellular images daily. This high feature in the volume of data accelerates the training and optimization of the artificial ielligence model.
Beyond the diagnosis of lung cancer, Cytomad has the poteial to accelerate the process of discovering the drug. The combination of rapid imaging and artificial ielligence analysis allows the drug screening processes to do more efficiely than traditional methods.
The research team is seeking to finance a three -year clinical test on paties with lung cancer to track the results using artificial ielligence -based imaging. This study can make artificial ielligence more widely accommodated in medical diagnosis and improved efficiency and scalability of health solutions.




