July 7th, 2014
Dear Indiegogo Backers,
We hope you enjoyed your 4th of July weekend with fireworks, friends and family. We’ve been plugging away with our own sparks. Here’s the latest:
Over the past two weeks, we’ve hosted two Scanadu Scanathon events at our HQ in Mountain View, CA. These events were incredibly helpful in gathering new data points as we move forward with the Scanadu Scout™. Thank you to all who were able to participate!
During the calibration days, each participant completed three Scanadu Scout™ scans and were measured with five other comparison devices - including continuous monitoring, automated and manual methods. A comparison device is a device used in clinical testing that is known to be at a certain level of accuracy according to standards and regulations. These are used to compare and calibrate scans from the Scanadu Scout™.
With over 60 diverse participants throughout two days we captured 200 completed scans, which resulted in over 800 vital sign measurements.
This is just the meta data though - beneath each scan is a combination of thousands of individual raw sensor readings that are captured every two milliseconds from each device. All of the data is ingested by our algorithms - that is then used to improve the model for the measurements for each vital sign and for each individual person. Our team is knee deep in processing all this rich data (code warrior pictured below).
WORKING ON TIMELINES
During our rigorous testing at the hypoxia lab, we located an opto-mechanical (light sensor) issue in the SPO2 sensors. The aesthetics of the acrylic window causes unwanted light reflection picked up by the photodiode (part of the SPO2 sensor diagramed below) that is interfering with the signals. We have a light blocking solution that will make the signals more stable.
To test our solution for temperature, we need to collect temperature data points from a patient population with fever in a hospital/clinic setting. Our clinical team is setting up the appropriate framework for this. These solutions involve hardware and we’re working with our manufacturer in Fremont, CA to make the right decisions about timelines to be able to resume shipping.
All in all - why does this take so long? Algorithms. An algorithm is the massive combination and processing all of the data from all our sensors. If one element in our algorithm gives a wrong input, the whole algorithm ecosystem is affected. Meaning every minute change on hardware or software requires us to redo bench testing (on healthy people) and edge-case testing (on sick people) - which then must be re-calibrated again and again with every change.
There is so much going on in this small unit by putting all these sensors so close to each other that some friendly fire is inescapable. We know you want to have your hands on it - and we want that too and are working towards this every day. Sorry it takes so much time, if only our team was made of silicon - we could sample and process per millisecond, but we are slow carbon creatures. But resilient.
We value your patience. If you have any additional questions, contact us at firstname.lastname@example.org and we will work with you directly.
The Scanadu Team