Meet the innovators using AI to ease one of the world’s biggest healthcare burdens

Back in 2012 while entrepreneur Farzad Saber and his colleagues were seeking investors for a new medtech solution, a hospital professor wrote, “I don’t want to read this application because machines can never treat humans.” 

Fast forward just a few years and artificial intelligence (AI) is having a positive impact on many areas of medicine. 

Luckily, Saber and his team at medtech company O2matic persevered. Their idea, to automate the oxygen therapy process, caught the attention of the Innovation Fund Denmark, whose grant helped them to move towards gaining a ‘CE mark’, allowing their product to be sold in the European Economic Area, which comprises 30 countries. 

Now O2matic’s solution, a portable device that continuously monitors and adjusts a patient’s oxygen saturation levels using complex algorithms, is helping patients with Covid-19 and chronic obstructive pulmonary disease (COPD) in hospitals across more than 25 countries. The technology has also just been approved for home use and is being offered to patients in Denmark. 

An estimated 328 million people suffer from COPD globally – and the number is rising. In 15 years, COPD is expected¹ to become the leading cause of death worldwide and many severe COPD patients need to visit their GPs or specialist doctors on a regular basis so that nursing staff can monitor their state of health. 

Modern oxygen therapy is based on a 100-year-old method, whereby manual adjustments are made to keep a patient’s oxygen saturation within range. Doing this manually means oxygen levels are often adjusted between 20 and 30 times a day. “Oxygen is a drug,” says Saber. “It’s not good to have too much or too little. If a patient does not receive the correct dosage, there is a risk of tissue damage, inhibited breathing or the need for respiratory treatment.”

Physicians define the oxygen saturation target and the range for how much oxygen the patients may receive, instead of having a fixed flow. The system will adjust the oxygen flow and keeps the patient’s oxygen saturation levels in the correct range 85 per cent of the time, significantly improving on the 47 per cent achieved with manual treatment.

Now the technology is ready to be moved home to patients. Built on Microsoft Azure and using IoT technologies and AI, the patient’s settings are stored in the cloud and synced with the device. If changes are necessary, they can be done remotely from the hospital.

The results are heartening. A 2018 study² found that the O2matic solution was able to effectively control oxygen saturation for patients admitted with an exacerbation of COPD and reduce time with unintended hypoxemia (low blood oxygen). Moreover, patients expressed high confidence and a sense of safety with automatic oxygen delivery.  

COPD sufferers are at a significantly elevated risk from Covid-19, and travelling to hospitals increases the chance of infection. Furthermore, the significant demand for oxygen involved in treating Covid-19 patients means every effort must be taken to conserve as much as possible. 

At the end of January, the Danish regions of Copenhagen and Zeeland began trialling the system for use with patients at home. “Because we are mostly talking about older people, the technology at home has to be very simple,” says Saber, “with all the setting changes being done remotely from the hospital.”

Elena Bonfiglioli, Microsoft's regional business leader of health and life sciences at EMEA, believes the ethos underpinning projects such as these is simply “to empower healthcare workers and their patients.” If we don’t do this, she adds, “We just become users of technology, rather than being enabled by it.”

The future applications for O2Matic’s tech are exciting, says Saber. “The flexibility of Microsoft’s platform means it’s easy to add new sensors, so we can measure things like temperature and weight. We have a lot of opportunities to start combining more data in the cloud – which trains our system continuously.” This, he believes, will help prevent hospitalisations by foreseeing potential exacerbations. “This has applications for every patient diagnosis where there is a problem oxygenating the blood – lung cancer, Covid-19, future pandemics and other diagnoses that have an impact on the patient’s respiratory areas.” 

It’s also a step on the journey towards changing the way we treat respiratory patients, Saber believes. “Would you accept it today if someone gave you a diabetes diagnosis, but told you that you were going to receive the same amount of insulin for the next three months? All these years we have been treating patients in the same way, so I hope we are going to significantly improve quality of life.”

For respiratory expert Dr. Thomas Ringbæk, a consultant at Copenhagen University’s Hvidovre Hospital, it means a reduction in unnecessary hospital visits. “Sometimes we have to bring patients in to get more information on their levels. [With the O2Matic solution] we – and they – can be sure their levels are right.” 

In addition to improving patients’ quality of life, automated oxygen eases the pressure on respiratory departments as the 20 to 30 daily checks on the patient by a nurse are no longer necessary. “There are serious projections from the World Health Organisation on the shortage of health professionals,” notes Microsoft’s Bonfiglioli. “If we don’t digitise patient journeys and bring intelligent systems that sit there as companions to liberate the time of existing health professionals, we’re going to find ourselves in yet another emergency situation.”

The potential impact of technologies such as O2matic's on the future of the medical industry is exciting, with the power to change people’s lives and transform how healthcare is delivered.