World’s first AI urine analysis for any toilet
An Israeli startup has developed an AI-based system that can monitor a person’s health from their urine.
Olive Diagnostics says the KG system is the world’s first 100 percent passive non-invasive real-time urine analysis sensor that mounts to any toilet.
The system uses an array of 64 LEDs at different frequencies spanning from the high UV at 350nm to Near Infrared (NIR) at 2,500nm. The LED array is mounted in the front of the toilet and projects to the back of the toilet where there is an array of six photodiodes as receivers.
The system generates information on red blood cells, protein, ketones and creatinine as well as other urine characteristics such as volume, pressure and colour. The data is then processed in the cloud to identify biomarkers for conditions including kidney stones, urinary infections and even impending heart failure.
If any issues require immediate attention or the onset of symptoms is detected, the system sends an alert to a relevant caregiver, and the on-going analysis can flag emerging problems early.
“The location problem is solved by using multiple photodiodes in the back of the toilet. This way we have 80 percent coverage of the toilet surface, only if some urinates on the way-side of the toilet it won’t be captured, everywhere else there will be a straight line between the LEDs and photodiodes that pass through the urine stream,” Guy Goldman, CEO of Olive Diagnostics told eeNews Europe
The battery powered system is based on the ESP32 microcontroller from Expressif and only turns on when the motion sensor detects movement 20cm from the toilet for a few minutes a day. It is connected to the home WiFi which is used to transfer the raw data of the sensor to the cloud that does the actual analysis, and this gives the system a three month battery life.
If there isn’t WiFi connectivity in the toilet room there will be a requirement to pair a phone. When the phone is in proximity of the toilet, the app on the phone will automatically download the raw data stored on the device and send it to the server.
The system has been in third party clinical trials at the Hadassah Ein Kerem Medical Center in Jerusalem. In tests designed to mimic the traditional urine stick method, KG delivered a sensitivity rating of 92.8 percent and a specificity of 95.5 percent, compared to 80 percent and 95 percent for a stick test. However by measuring over time, the trial found that the KG system achieved a sensitivity rating of 98.7 percent and a specificity rating of 100 percent on simulated five daily urinations.
The company has so far raised $4m and plans to work with assisted living facilities, outpatient facilities, and telemedicine providers in Europe to provide the service. The company has completed its ISO 13485 certification, a critical milestone necessary for CE certification and is in the FDA pre-submission phase for the United States.
It is looking to raise Series A funding during the first half of 2022 to support its growth.