Will big data revolutionise medicine?

Publié le 5 Nov, 2014

Le Monde is publishing a supplement devoted entirely to big data, i.e. to the  “technological breakthrough”, which potentially allows a vast amount of information to be collected and correlated. This heralds a revolution in all sectors, including the health sector.

 

Big data would pave the way to “preventive and personalised medicine”. Scientists hope to be able to link available health-related data (scans, prescriptions, test reports, etc.), following in the footsteps of the team at Boston Children’s Hospital which “has linked data from World Health Organisation (WHO) databases and GeoSentinel or Eurosurveillance sites as well as American search engines Google News or the Chinese search engine, Baidu News”. Thanks to this method, the Boston team detected the HIN1 virus epidemic before it spread.

 

Furthermore, several hospitals in the United States use the “Watson super computer developed by IBM” to help doctors propose a therapeutic strategy for cancer patients. “In fact, the computer is an additional team member. It directs doctors to the key facts in a patient’s medical record. It suggests treatment and confirms the reliability of its response,” explained Patrice Poiraud, Director of big data and analytics at IBM.

 

“ Cancer is also the initial focus for big data in France”. Unicancer in conjunction with the Roche Laboratory has launched a big data project – ESME (épidémio-stratégie médico-économique – medical-economic epidemic strategy), which between now and 2018 intends to analyse the data recorded for 40,000 breast cancer patients treated with a Roche molecule.

 

“ In the long run, big data will be used to examine the DNA analysis of a cancer patient, genetic anomalies and data relating to thousands of “similar” patients in order to define “the” most suitable molecule and dosage for individual patients. This will be the personalised medicine era”.

Le Monde (Sophy Caulier) 06/11/2014

Share this post

[supsystic-social-sharing id='1']

For further