How fitness trackers help Singapore treat mental illnesses
By Ming En Liew
Creighton Heaukulani, Senior AI & Data Scientist, the MOH Office for Healthcare Transformation, Singapore, shares how data is being used to treat healthcare conditions.
This hypothetical scenario may soon become a reality with the use of data monitoring devices. The Singapore Ministry of Health Office for Healthcare Transformation (MOHT) is exploring ways to use personal devices to monitor the health and behaviour of its citizens to better deliver health interventions.
Creighton Heaukulani, Senior AI & Data Scientist at MOHT, shares how data helps healthcare providers treat and monitor citizens remotely.
Using real-time data to deliver personalised treatments
MOHT is exploring ways to use personal monitoring devices such as smartphones and fitness trackers for medical intervention, says Heaukulani. This is similar to how fitness trackers can prompt an individual to get up and move when it senses that they have been sedentary for too long, he explains.
This can help patients feel more empowered and take more responsibility for their health on a daily basis, adds Heukulani.
At the moment, Singapore is using this approach to help patients with mental illnesses. A research programme tested the feasibility of this strategy by monitoring patients with schizophrenia who were recently discharged from a psychiatric hospital.
Patients had to wear a fitness tracker, and were monitored based on several factors such as location, sociability, motion, finger taps, or ambient light. These factors can clue a healthcare practitioner in on the patient’s physical and mental health.
For example, analysing how a person types can help detect if they are affected by physical or mental conditions. Someone who is fatigued may tap more slowly; while individuals suffering from diseases may have small or uncontrollable movements, according to a research paper co-authored by Heaukulani.
If anomalies are detected, automated systems can send healthcare resources to affected individuals. These systems can also notify care managers to contact patients who may have more severe cases, says Heaukulani.
“Similar strategies have the potential to revolutionise treatment across many patient groups,” says Heaukulani.
Monitoring citizens’ health through personal devices
Data collected by personal devices is also useful in monitoring the physical and mental health of the population at large.
For instance, researchers conducted a small-scale study to determine if Singapore’s lockdown in 2020 had affected the physical and mental health of the population. Their data discovered that physical activity, sleep, and sleep efficiency all decreased during the lockdown.
MOHT is also exploring the use of data in monitoring and managing chronic diseases such as heart disease and hypertension. Health practitioners can monitor patients with these diseases remotely through at-home blood pressure machines, where readings are self-reported through a mobile app.
“The key is to create a solution that helps the patient to feel more in control and able to manage their own health,” he says.
Using machine learning to safeguard patient privacy
While using data can provide valuable insights into the health of citizens and help in treating certain conditions, user privacy still needs to be protected. Machine learning can help ensure this.
MOHT designed a machine learning model that predicts the distance someone travels in a day, which can be used as a measure of sedentary behaviour or sociability, shares Heaukulani. Instead of using GPS sensor data which tracks the location of individuals, the machine learning model will only need data from the accelerometer, pedometer, and an ambient light sensor.
Other ways of preserving patient data is to avoid collecting information on what users are doing on their devices. For example, a user’s keystrokes while typing are converted into arbitrary categories such as alphabetic, numerical, or punctuation to avoid recording the content of what they are typing.
“The first principle is that users or patients always have to understand what data is being collected, how it is being used, and consent to its use,” emphasises Heaukulani. “The onus is on us to make sure we…carefully protect the data in accordance with all laws and applicable ethical principles.”
Data is opening doors for public sector spending, aid provision, and threat detection. The healthcare sector is no exception. Granting healthcare workers real-time access to patients’ lifestyle and health data allows for more timely intervention and better health outcomes for all.