We will save 1 million lives by 2020 … · Passive Tracking for Diabetes Physically active Visits...

Preview:

Citation preview

We will save 1 million lives by 2020

Coal miners used birds for early detection of gas leaks

We use mobile technology and AI for early detection of critical disease symptoms

The most advanced sensor devices in history

HYGROMETER MAGNETOMETER GYROSCOPE ACCELEROMETER GPS BAROMETER HEART RATE

Highly precise detection of micro movements

Detecting Sleep

Detecting Meals

Detecting Work Times

Detecting Inactivity

Detecting Cycling

TEAMMatching Sensor Data with Medical Data to prevent

Hypoglycaemia through Early Detection of Crit ical Patterns

Matching with CGM Data

Gets up at night Physical active Visits risk location

Energy

Correlat ion of glucose level and dai ly act iv i t ies of a d iabet ic

Passive Tracking for Diabetes

Physically active Visits risk location

TEAMMatching Sensor Data with Medical Data to prevent

Hypoglycaemia through Early Detection of Crit ical Patterns

Matching with CGM Data

Gets up at night Physical active Visits risk location

Energy

Correlat ion of glucose level and dai ly act iv i t ies of a d iabet ic

Passive Tracking for Diabetes

Physically active Visits risk location

Matching with CGM Data

Gets up at night Physical active Visits risk location

Energy

TEAMMatching Sensor Data with Medical Data to prevent

Hypoglycaemia through Early Detection of Crit ical Patterns

Matching with CGM Data

Gets up at night Physical active Visits risk location

Energy

Correlat ion of glucose level and dai ly act iv i t ies of a d iabet ic

Passive Tracking for Diabetes

Physically active Visits risk location

Matching with CGM Data

Gets up at night Physical active Visits risk location

Energy Gets up at night Physically active Visits risk location

Assessment of Contextual Data

Our algorithm translates raw data into real life activities and assesses their risk.

xbird Machine Learning Model

Dancing Jumping Walking

Data

Segmentation

Feature Extraction

Patient’s Accelerometer Data

Data is labeled (we know what people are doing)

xbird Machine Learning Model

Training Set Training Phase Testing Set

New Case Running Model Output

Is prediction

compatible with real

labels?

+

Confusion Matrix Heatmap

Walking Cycling Sleeping StationaryGeneral Physical

Activity

Walking 91,8 4,2 0 0,5 3,6

Cycling 2,9 78,1 0 8,9 10,2

Sleeping 0 0 98,5 1,4 0,1

Stationary 3,1 8,5 1,3 85,5 1,6

General Physical

Activity 2,3 9,1 0,3 3,7 84,5

Detecting Work & Home Location

Location Detection through Activity Data

Work Home

High Risk Locations

Flagging Locations correlated with the most ups and downs in blood sugar

TEAMWe offer an SDK, allowing for many use cases

Team

Sebastian Sujka

CEO

Matteo Carli

CTO

Dr. med. Jonas Harder

Chief Medical Officer

Dr. Silvan Türkcan

Scientific Advisor

Nikolai Stosch

User Experience

Alex Roslyakov

Tech Lead iOS

Joseph Lam

ML

Klaudia KobelrauschMedical Research

Awards and Endorsements 2016

Winner of the Health-i Award

2016

Winner of the Life Sciences Track at

Pioneers Festival 2016

Winner of the TK Innovation Award, exclusive 1-

year partnership with TK

Winner of the eHealth Venture Summit

Award 2016Digital Health Innovator

Part of Bayer’s Grants4Apps Accelerator

Program

Final Remark

This is happening

This is happening

Thank you

Sebastian Sujka

CEO

Matteo Carli

CTO

Dr. med. Jonas Harder

Chief Medical Officer

Dr. Silvan Türkcan

Scientific Advisor

Nikolai Stosch

User Experience

Alex Roslyakov

Tech Lead iOS

Joseph Lam

ML

Klaudia KobelrauschMedical Research

sebastian@xbird.io

Recommended