by Han-Ru Xie
Facing a future society where manpower is gradually decreasing in healthcare, while the demand for healthcare is on the rise, AI techniques can help share the burden of doctors and provide patients with more accurate and rapid medical services.
One minute only- retina scan for heart disease
The University of Cambridge and St George’s, University of London, joined hands in developing a quantitative analysis software called “QUARTZ” to examine retinal vessel topology and size. With AI-based programming, it takes only 60 seconds to determine the risk of a patient having heart diseases.
QUARTZ analyzes retinal images of 95,463 subjects from the UK Biobank and the European Prospective Investigation into Cancer and Nutrition (EPIC); with trained machine learning algorithms to analyze the width and tortuosity of retinal blood vessels, the readings are matched with the collected data of heart conditions from the participants. In this way, a model for predicting heart disease risk is built to compare the scanned images of the retinal vascular system with the database data to determine the health status of the subject’s heart.
There is one pity that, at present, 94.6% of the subjects in the UK Biobank are white people, making it harder to predict the risk of having heart diseases for the colored population.
Science for disease prevention in Southern Taiwan
After the outbreak of the COVID-19 pandemic, the Kaohsiung Municipal Siaogang Hospital, located in Siaogang District, Kaohsiung, shoulders the responsibility of transporting patients from the Kaohsiung International Airport to medical institutes, due to its geographic proximity. Way before the outbreak, the hospital has been making efforts in watching over the neighborhoods by developing a smart medical system for treating lung diseases. After the pandemic outbreak, the hospital established an AI lung X-ray film detection system at once, which could quickly determine whether a patient is contracted with COVID-19 in just 1 second.
Taking the number of films for preliminary screening in the emergency room as an example. There are about 900 films per month. Upon using the AI system, there have been an average of at least 60 working hours saved per month, greatly reducing the workload of doctors and making the most of manpower.
In addition, there is the multifunctional micro-spectrometer, jointly developed by the Kaohsiung Municipal Siaogang Hospital and a tech company Spectrochip. The device has a size 100 times smaller than the regular instrument in the hospital, and it only takes one drop of blood and 5 minutes to examine thoroughly a patient’s white blood cells, liver function, kidney function, blood sugar, etc. For now, this device can also be used to test the antigen and antibody for COVID-19. The precision is as high as 90% within 4 days of the onset of symptoms, and all it takes is one drop of blood.
Healthcare is indispensable in everyone’s life. It is believed that AI technology will bring a strong boost to the medical field after AI technology becomes more comprehensive in the future.