All over the world, there are millions of people who do not have access to regular health services such as going to the doctor, and this has major consequences. But artificial intelligence can actually help improve global health and save lives. – Artificial intelligence can be an important tool for improving global health and closing the inequality gap, says Miriam Stankovich, who has a Ph.D. in development economics and is an expert in the regulation of new technology such as artificial intelligence. She works for DAI, one of the largest international development companies in the world, which, among other things, creates development programs for the American development fund USAID and the World Bank. Artificial intelligence can be difficult to understand because there are different types. Some types are built up by algorithms, while other types of artificial intelligence are self-learning. What they all have in common is that they are based on large amounts of data that are analysed, explains Stankovich. Artificial intelligence is going to play a big role in the healthcare system of the future. In Denmark, they already use artificial intelligence to help the employees at the alarm centre. When you call 112, artificial intelligence picks up the phone and will assess the risk of a cardiac arrest, and more is on the way – also in developing countries. – The beauty of this technology is that it makes it possible for developing countries to skip a few steps and improve the health of millions of people more quickly, says Stankovich. Artificial intelligence can save babies One of the most dangerous periods in a human life is the first 30 days – especially if you are born in a developing country. Although child mortality has fallen by 61 per cent globally since 1990, which is a big improvement, the death rate among newborns has not fallen as much. In 2020, 2.4 million children lost their lives in the first month, and around 80 percent of the cases were in Central and South Asia and sub-Saharan Africa. One of those working to include artificial intelligence in the healthcare system, which can benefit developing countries, is Mozziyar Estemani. He is a researcher at North Western University in Chicago, and works with an ultrasound scanner which, together with artificial intelligence, can predict possible complications in connection with a pregnancy. – Today, both heavy equipment and trained radiologists are needed for ultrasound examination of pregnant women. And it is a problem in many developing countries, where there is both a lack of equipment and doctors, he says. That is why he is working on a small hand-held scanner, which connects to a smartphone, so that pregnant women can scan themselves throughout their pregnancy. – If there is something that does not look right, the scanners and the smartphone will notify the pregnant woman to seek medical attention, says Estemandi. And the new scanner is much cheaper than the large ultrasound scanners in the hospital. – When I started medical studies over twenty years ago, a scanner cost between NOK 140,000 and 200,000. The hand-held scanner costs around NOK 7,000 today, and a smartphone is also needed, he says. So, firstly, the equipment is cheaper, and secondly, it can be used by the pregnant woman herself, even in areas far away from the hospitals. Right now, researchers are collecting data in the US, Zambia and South Africa to improve the technology. Researchers collect amateur images scanned by women themselves during pregnancy. The amateur images can be used to train the artificial intelligence to recognize possible complications in the fetus. Processes data in a few seconds But artificial intelligence can be used for far more than ultrasound examinations. It can, for example, help to diagnose patients and help doctors to find out what the various symptoms may be an expression of. – In Gambia and Zambia, artificial intelligence is used to identify diseases such as tuberculosis, which are not visible to the human eye, because we humans cannot process such large amounts of data within a few seconds, says Stankovich. Artificial intelligence can help make doctors’ work more efficient, which can be decisive because there is a global shortage of doctors. In fact, 40 percent of all the world’s countries have less than 10 doctors per 10,000 inhabitants. There are therefore too few doctors to treat many people, and this means that many do not get the treatment they need. Nevertheless, clinical experience and knowledge of the specific patient will be decisive. Artificial intelligence cannot replace doctors, explains Stankovich. – Artificial intelligence is a fantastic tool, but it must be treated as a tool, because it cannot work alone, she says. Must build on local data Data collection is an important part of the development of artificial intelligence. Without data, there is nothing to analyze. Data are the building blocks of artificial intelligence, and the data must be collected locally in the country where it will be used. – If you take an artificial intelligence tool developed in Silicon Valley and start using it in India or Zambia, it will not work, says Stankovich. Because there is a big difference between people, and when it comes to image recognition, for example, it is crucial that images of white people are not used to diagnose people with a different skin color or vice versa, she explains. The same applies when diagnosing women, because available data often come from examinations of men. – Artificial intelligence is only as intelligent as the data it is built from. It is therefore important to be aware of which data is used – and which bias it has, says Stankovich. It precisely emphasizes that artificial intelligence is a tool that should be used as an aid to the healthcare staff, i.e. it should not work on its own and decide diagnosis or treatment alone. Electricity and internet When we talk about using artificial intelligence in developing countries, there are some very specific conditions that must be met, which primarily concern access to electricity. – If we try to roll out artificial intelligence in countries where the government and people struggle with access to electricity, it will have no effect, says Stankovich. The latest figures from 2021 show that 675 million people do not have access to electricity, mainly in Africa and Asia. With access to electricity, it is often also possible to access the internet, and this is crucial as there is a lot of data to be collected and stored. – There are some basic things that must be in place before artificial intelligence can seriously strengthen global health, says Stankovich. But we are moving towards a world where there is more equality in health.
ttn-70
Artificial intelligence can improve global health and save lives
