The Future of Wearable Sensors - From Quantifying Behavior to Quantifying Health
Project Infobox 
| |
|---|---|
| Self researcher(s) | Marco Altini |
| Related tools | sensor |
| Related topics | Stress, Sports and fitness |
Builds on project(s) |
|
| Has inspired | Projects (0) |
| Show and Tell Talk Infobox | |
|---|---|
| Featured image | 
|
| Date | 2013/05/12 |
| Event name | 2013 QS Europe Conference |
| Slides | |
|
This content was automatically imported. See here how to improve it if any information is missing or out outdated. |
The Future of Wearable Sensors - From Quantifying Behavior to Quantifying Health is a Show & Tell talk by Marco Altini that has been imported from the Quantified Self Show & Tell library.The talk was given on 2013/05/12 and is about Stress, and Sports and fitness.
Description[edit | edit source]
A description of this project as introduced by Quantified Self follows:
Marco Altini works for Imec. His company develops wearable technology (i.e wearable sensors, wireless wearable sensors for EEG, stress, and physical activity monitoring). They often combined accelerometer data and physiological data to create these sensors. In this video, he talks about quantifying health using these wearable sensors.
Video and transcript[edit | edit source]
A transcript of this talk is below:
Marco Altini
The Future of Wearable Sensors - From Quantifying Behavior to Quantifying Health
So we are back to the Italian accent, I’m Marco and I’m going to talk about quantifying health using wearable sensors, so first some background. I work for imec which is a company that develops wearable technology so we make wearable sensors, wireless wearable sensors for EEG, stress, and physical activity monitoring and we often combined accelerometer data and physiological data to create these sensors and a stake candidate of Eindhoven University of Technology, where we work on machine learning mainly to track and quantify daily routines and aspects of behavior among other things. So what I do with my Ph.D. is mainly try to quantify not only our behavior, our daily activities and what we do but our actually health status using wearable sensors. I think there is a need for this because today we are already have all sorts of gadgets. So we have mobile phone applications, we have all sorts of accelerometers. We have more complicated devices which include all sorts of sensors, but what all of them do is to quantify some aspects of our daily life, for example how many steps we’ve taken, or how many hours we slept or how active we are, or how many calories we burned. And all of these systems they all do the same or at least they try to do the same. I mean we know this plots it’s very difficult to find agreement among what these systems measure, but this is really not my point today. So let’s assume for example that we do have already perfect activity trackers for example devices which have been validated against reference systems and while doing all sorts of activities and all sorts of lab tests. Then if we are in this condition then we can have quite a good snapshot of our behavior of what we do in our daily life. But my question then is how does our awareness in our behavior translate into awareness in our actual health status. So basically if we go back to our health meter, where do we end up there? So in order to understand where we are we need metrics and marker of health which is related in our behaviors in terms of physical activity, and that’s why for example we use scales. That’s the simplest way to measure our health to get some feedback to understand our behavior affects our health. But are these good metrics? So I’m going to stick to the case of body weight because that’s the most common metric. Here, I’m plotting the relation between body weight and BMI and mortality risk. So we have unfit subjects and we see there is a link between body weight and mortality risks. So we have three groups; normal weight, overweight, and obese, and the higher the BMI the higher the risk. But all of these people are unfit, so what happens if next to them we plot the fit people. So the fit people with high levels of cardiology fitness, it happens that basically there is no difference in mortality risk across all body weight ranges providing people are fit. So if it’s not about being fat but about being fit, I think that’s what wearable sensors should measure. So the feedback you should provide to people is the actually fitness level in order to allow people to change behavior if they are in a danger zone. Because I think if we keep using the wrong metrics then actually measuring our health is going to be more complicated, and the data that I just shown was 20 year olds, so it’s not big news. But what I think is changing now is that we finally have the technology to do much better than this. So what I do with my research is to work is to use available technology to work on algorithms to estimate fitness of any individual and without requiring any calibration or any specific test. And just a few words on how this works which is basically starting from the context, so we use sensors and phones to understand is where people are and what activities they are doing and at which intensity are they doing different activities. And then from this you can start looking at physiology, you can start looking at how the cardiovascular system reacts in this known context and situations. And then you can bring together this information, and when you look at physiology in known context you can estimate in behaviour and what people are doing but you can actually health, you can measure fitness which is the best marker of health. And you can gain knowledge, not only on behavior but also on actual health status. So I think this is very important because then what you can do is to finally close the feedback loop between what we do and how healthy we are, and even motivating behavioral change can be much easier.
Thank you.
About the presenter[edit | edit source]
Marco Altini gave this talk.