Seb Oliver and Peter Hurley are professors of astrophysics at the University of Sussex.  Peter Hurley suffers from cystic fibrosis.  Together they had the idea of applying a technique used to distinguish and match galaxies captured on different telescopes to the data collected about patients with cystic fibrosis (CF).

CF is a genetic disease and the gene is carried by one in 25 Europeans.  In the UK around 10,400 people have the condition and it radically reduces life expectancy.  It is currently incurable and treatment to most successfully impede its effects depends largely upon being put on the right medication, of which there are many different options.  Improving patient prognosis relies upon greater understanding of the long-term impact of the medication on different patients.  This is made more challenging due to the long time frame of CF treatment and that it is a rare disease.  The main problem though lies in the necessary anonymisation of patient data and the resultant break in threads of individual patient’s data, due to clerical errors and changes in data such as body mass index (BMI) not being identified.

What the two professors have done is to apply a computational framework that calculates the probability that a pair of celestial objects in two images are the same and parallel this with an alternate theory that they might be two different objects that just happen to be adjacent.

The analogous comparison with CF patient data is clear factors such as age and gender.  But often this is not sufficient as information that is likely to fluctuate, such as BMI, can lead to broken links in individual patient data.  The key factor in their framework is the inclusion on an analogous model taken from the fact that galaxies will appear to have different brightness levels when viewed through different telescopes, due to the wavelengths of light that they are reading.

Physics applicants can further research the mathematical models used in astrophysics to differentiate objects.  Medicine applicants can consider the importance of patient anonymity when their data is analysed as well as how maintaining the threads of data will help in their analysis.   

More Resources