Sensorama 2016

In the 2016 Sensorama project, the students investigate possible confounding variables that could influence the correlation between measured impedance and near-infrared (NIR) spectra and blood glucose. The investigated variables are gender, height and weight (and hence BMI).

The test subjects skip their breakfast and meet fasting in the morning. Their blood sugar, impedance spectrum and NIR spectrum is measured every 10 minutes and after the first 3 measurements, they are given 7 dl Coca-Cola to drink. Measurements are then done as the blood sugar level increases and then decreases again.

They are then given another 7 dl of Coca-Cola and measurements are continued while the blood sugar again increases.

After measurements have been done on eight male and eight female subjects (students), multivariate methods are used to generate the best possible prediction algorithm for the blood sugar, based on impedance and NIR. It is then studied how gender, height and weight may influence this correlation.
Ole gives a lecture on diabetes and blood sugar.
Ready for the first experiments.
Coca-Cola may be good, but not as your only breakfast.
Doing impedance measurements.
And NIR measurements.
Christian explaining about the data analysis.
A first look at the data.
Some simple correlation studies of single variables.