Artificial intelligence is able to monitor sleeping epilepsy patients

Submitted on Tue, 01/30/2018 - 12:38

Tampere University Hospital (TAYS) in Finland is the first in the world to introduce a system that monitors the sleep of epilepsy patients

Jukka Peltola/ Kuva: Jonne Renvall
The Mediuutiset newspaper named Jukka Peltola notable person in the health field. One reason was the unique epilepsy monitoring system introduced at Tampere University Hospital.

Text: Jaakko Kinnunen
Photo: Jonne Renvall

Tampere University Hospital (TAYS) in Finland is the first in the world to introduce a system that monitors the sleep of epilepsy patients. The system makes use of artificial intelligence to recognize irregular bodily movements that may be caused by epileptic seizures.

The device consists of a night-vision camera, a depth sensor and a microphone. The system produces a report of the night’s events to be used to plan the patient’s treatment.

According to Jukka Peltola, Professor of Neurology at the University of Tampere and chief physician at TAYS, the new system is able to provide information on night-time epileptic seizures that is more reliable than information produced by other methods.

“We hope to be eventually able to transmit the information in real time. We are cooperating with researchers in both Finland and internationally,” Peltola says.

The Finnish Mediuutiset newspaper issues an annual list of influential people in the health care field. Jukka Peltola was one of the experts the newspaper included in the list in 2017. “I was happily surprised by the nomination,” Peltola says.

The symptoms of an epilepsy patient may be monitored at home or in hospital. The new method provides information for medical staff and researchers on epilepsy and the clinical picture of individual patients. The method also helps to reduce the need for expensive video EEG tracking.

Night-time monitoring is important because the patients are often incapable of noticing the seizures while sleeping. Epileptic seizures always include the risk of sudden death, which means that the monitoring system is likely to save lives in the future.

“If the patient sleeps alone, he or she may not even notice the seizures. Seizures often come with impaired awareness, which may make the patient’s own observations unreliable. The patient’s spouse may not notice the seizures either and it is often difficult to describe them accurately,” Peltola explains.

An epilepsy patient may suspect having a seizure in the night if he or she wakes up with a bitten tongue or sore muscles. However, a seizure does not necessarily leave clear signs.

The system will also help nursing staff monitor the effects of medication. If the patient’s medication is changed, the monitoring system is able to detect the effects of the change.

“About thirty percent of the patients suffer from severe forms of epilepsy. The medicines used to treat the disease are heavy and their use is often reduced if possible,” Peltola says.

The use of the first monitoring system started last spring. Thus far, the method has been used to treat about thirty patients. Neuro Event Labs Oy, a technology company in Tampere, developed the system.

Peltola points out that the artificial intelligence that enables the system to recognize irregular movements during sleep is continuously evolving.

“Artificial intelligence is often regarded as some sort of a mysterious thing that will solve every problem but it is really a process that is constantly developing,” Peltola says.

Epilepsy is a chronic disease that often requires life-long treatment. Epilepsy may occur at any age.  However, it more frequently occurs in early childhood and old age. In Finland, about 56,000 people, i.e. one percent of the population, suffer from the disease. About 36,000 patients need continuous medication.

An epileptic seizure is caused by a disorder in the electrical signals of the brain. There are various causes for the disease, which may not be clear in all cases.