According to the United Nations, the age group of 65 years is growing faster than others and the number of elderly people will nearly triple by the year 2050. Several factors contribute to the world’s ageing population, including improved healthcare and the related increases in longevity that come with it. Low birth rates, particularly in countries in East Asia and Europe, also affect the average age of the population.
In the coming years, providing quality and affordable healthcare for this huge population of elderly will become a growing priority for countries throughout the world, and this work could be aided by recent developments in artificial intelligence, according to Fakherddine Karray, Professor of Machine Learning at Mohamed bin Zayed University of Artificial Intelligence (MBZUAI).
Karray is developing novel algorithms based on deep learning programs that can be used for all sorts of human activity recognition, which are the methods by which machines interpret people’s movements within a certain environment. These tools can be applied to monitor the health and safety of older people and provide them with support in moments of need.
Remote patient monitoring can be used in healthcare clinics or even in the home. When applied in a clinical setting, it may reduce healthcare costs by helping doctors and nurses be more efficient with their time, dedicating their attention to the patients who need it most. When applied in the home, it can provide patients with the comfort that comes with living in a familiar space, close to friends and family, while potentially reducing the frequency of trips to the hospital.
“The idea for these applications is to help those who may require attention from time to time throughout the day, or to provide assistance to patients who do not need to be under the continuous supervision of a doctor or a nurse,” Karray said. “This would allow them to stay at home and easily communicate with a healthcare provider when necessary.”
Related: Personalised, next-gen healthcare experiences are within grasp
Integration of a complex network
A system to conduct remote patient monitoring and assistance would include the integration of several innovative kinds of smart devices, secure communication systems and artificial intelligence tools.
For example, sensory devices located in the home would have the capability to monitor the patient’s movement and activity. Analysing a patient’s posture and gait could identify early warning signs of a health emergency. “The sophisticated systems we have developed analyse the person, their stature, whether they are standing, sitting, engaged in certain kinds of activities, or have perhaps fallen,” Karray said.
In addition to movement, another algorithm would examine a patient’s face and interpret it to gain a sense of how the patient was feeling emotionally, whether they were in pain or were depressed. “With this approach, we can get a sense of the overall behaviour of the person, what we call behaviour recognition, and also an understanding of their emotional status,” Karray said.
A wearable such as a smartwatch could be used to measure the patient’s blood pressure, body temperature, blood sugar level and other health metrics. “Together these devices, also known as Internet of Medical Things, would coexist and provide us with an excellent description, or semantic representation, of what a person was experiencing at a given time,” Karray said. “And if there was an anomaly in the measurements, a notification would be sent to a healthcare provider to inform them that the patient may be in danger.”
The system would also utilise the latest tools of natural language processing to supply the healthcare provider with interpretations of the data that was being collected by the system and guidance about potential next steps regarding care or requesting additional help for the patient. ChatGPT-like systems would provide excellent capabilities for three-way interaction between patients, healthcare providers and the huge medical repositories that are stored in dedicated servers.
“And then, of course, there is the cloud system that processes in real-time all the information coming from the various channels and provides the adequate outcome or decision signal to the corresponding node of the designed platform,” Karray said.
Privacy and security on the edge
As with any device that is connected to the internet, privacy and security are concerns, particularly with devices that monitor the home. Karray and his collaborators are addressing these challenges from the start: “We are working with a group of people who are on the leading edge of security for the Internet of things, a concept that describes the network of smart, responsive devices.”
To provide them control over their data, patients could determine when and where in their home the system monitored their status. Patients could also determine who the information was shared with. For example, only certain physicians or nurses who were approved by the patient could receive access to the patient’s data.
To increase responsiveness and security, the team is using a concept called edge computing. Instead of sending all data that is collected in a clinic or hospital back to one remote server, edge devices, essentially computers that are installed in the same place where they are collecting data, will do much of the data crunching. “We wouldn’t store all information we gathered in the cloud,” said Thanh Cong Ho, a research assistant at MBZUAI who is working with Karray on the initiative. “Much of it would be processed on edge devices. This will decrease the time it would take to send information to the cloud and would increase security.”
Related: Cloud-based digital solutions streamline MENA healthcare operations
Adding life to years
Karray and his collaborators have already made significant progress and have developed several hardware prototypes that will provide key functionality for the system. They will continue to test and optimise these prototypes and are seeking to establish a collaboration with a healthcare clinic to run a study to determine the effectiveness of the system.
Within a matter of a few years, Karray hopes to partner with hardware companies and commercialise the system in a “range of hospitals, clinics or even in next-generation smart houses for the elderly.”
