- Google’s model is designed to be used for detection rather than diagnosis.
- Its main attraction is that it works for low-resource countries that do not have enough equipment or radiologists.
- The biggest drawback of tuberculosis is that it spreads too easily.
Until before the appearance of Covid-19 the tuberculosis It was the deadliest infectious disease on the planet. According to the World Health Organization (WHO) it is responsible for 1.5 million deaths a year. Although it is a curable and preventable problem, the vast majority of cases occur in low-income nations where the population does not have basic health services.
The tuberculosis is caused by Mycobacterium tuberculosis, a bacterium that almost always affects the lungs. It is an infection that is spread from person to person through the air. When a sick person coughs, sneezes or spits, they expel the bacteria and it is enough for a person to inhale a few of these bacilli to become infected.
It is estimated that a quarter of the world’s population is infected with the bacillus. tuberculosiswhich means that these people are infected with the bacillus but have not (yet) become ill and cannot transmit the infection.
a person with active tuberculosis It can infect between 5 and 15 people through close contact over a year. Without adequate treatment, an average of 45% of HIV-negative people with tuberculosis and almost all HIV-positive people with tuberculosis will die.
Technology can reduce detection times
To offer an alternative to the population, Google developed a new digital tool based on Artificial Intelligence. The most innovative thing is that it automates the entire process for its correct detection and thus speeds up treatment in communities where doctors are scarce.
Around this topic, the Radiology journal of the Radiological Society of North America (RSNA) published the results of a study. What was observed is that this tool had an efficiency equal to or even higher than that of radiologists in detecting tuberculosis through chest X-rays.
“Unlike much of the published data on AI, Google’s study was large and used different training sets, which showed that their system is robust,” said Edith Marom, director of chest imaging at Chaim Sheba Medical Center. In Israel.
Now, within the criticism, the specialist added that Google’s tests did not coincide with real-life circumstances. The data sets contained higher-than-normal rates of disease and were skewed toward patients who were younger and could withstand upright X-rays, conditions that generally make the images easier to interpret. Their performance also declined among sicker populations with more lung abnormalities, such as HIV patients and a group of miners in South Africa.
“To be implementable worldwide, it would have to be tested in populations with a low prevalence of tuberculosis that resembles typical patients. It will also need to be tested in an older population, which is typically found in the hospital setting.”
Despite the above, this Google’s new tool against tuberculosis is quite promising. Its main attraction is that in most low-income nations there are not enough radiologists. The direct consequence is that diagnoses in patients are not carried out or take too long to occur.
How does it work?
Google’s model is designed to be used for detection rather than diagnosis. Analyzes X-ray images to determine which patients should receive follow-up molecular testing to confirm the presence of the TB-causing bacteria.
Google’s tool ranks chest x-rays for suspicion on a scale of 0 to 1. The higher score equates to a higher probability that the disease is present. In the study, the company’s researchers calibrated the tool for recommending follow-up tests at a threshold of 0.45, which turned out to be the correct choice, as it proved to be very sensitive in detecting the disease without generating high false-positive rates.
The next challenge is to examine its performance in a real world environment. Google is now conducting a study at a clinic in Zambia, where the accuracy of the tools’ findings will be measured against molecular test results for each patient. The study is expected to be completed by the end of this year.
Also read:
National proud! Mexicans identify 2 new antibiotics against tuberculosis
Chart of the day: Deaths from Covid-19 exceed those from AIDS, tuberculosis and malaria combined
Will Mexico be able to eliminate tuberculosis by 2035?