Night-Sense unique solution is Hypo-Sense, a wrist watch like unit with an array of non-invasive sensors. The unit monitors patient's physiological parameters during sleep time.
Hypo-Sense is configured and personalized per patient on delivery by a technician in order to gain superior detection ability.
Detection of a nocturnal hypoglycemic event is done by analyzing the monitored parameters by means of a learning algorithm. When a hypoglycemic event is recognized an alert to the patient and care giver is activated. The system will record and transmit the readings to aid the care givers treatment and for further analysis.
The night sense unit is comfortable and most importantly do not require active intervention in the nocturnal hypoglycemia detection process.
The sensor unit
A non-invasive night time monitor is embodied with a cleaver combination of sensors which enable it to monitor the human body. The sensor unit is small and discrete wrist watch like unit.
The sensor unit monitors breathing and heart pulse, sweating, temperature skin, motion & tremor. Unique signal processing is used in order to counter the natural interferences caused by the human body during night time.
The Night sense unit then processes the signals by means of machine learning and decides if the signal indicates a hypoglycemic event. On the detection of an event an alert is activated and the data is transmitted and stored for further analysis.
Automatic identification process utilizes advance algorithm techniques such as machine learning and previous data in order to create and exact and reliable sleep time hypoglycemic events detection system.
The learning algorithm
The question in hand is how to successfully identify a hypoglycemic event. Learning algorithm presents a powerful tool that can handle the variety of symptoms that presents.
The first step is collecting a medically approved data set which contains physiological parameters records of patients while in a hypoglycemic event and not in a hypoglycemic event. The algorithm, after processing the data set, will know to recognize if a new signal indicates that the patient is in a hypoglycemic event.
The same algorithm can be used in order to learn the specific physiological parameters that indicate a hypoglycemic event. The learning algorithm can be adapted to specific groups of population and event to a certain patient if required.