How Data Science can help in a pandemic situation?

by Marlon Cárdenas - Data Scientist, Sopra Steria, Spain
| minutes read

With the aim of covering current and future needs of society, data science and Artificial Intelligence (AI) are seeking to drive the creation of technological solutions that benefit users in their daily lives. Many disciplines are uniting behind this cause, with health sciences to the fore, especially given the current context of the battle against the Covid-19 pandemic. Digital tools underpinned by these technologies are tackling problems that, approached in a conventional way, would cost more time and effort to resolve.

The knowledge challenge

Machine learning is an intricate process that can help us as a society achieve the goal of assisting people facing complex issues. Nonetheless, given this ambitious scenario, new challenges are arising that must be addressed to make space for these techniques. We always want to learn from nature, but to do so we need mechanisms that enable us to collect a lot of data. In a similar context, characterising a virus and predicting society’s behavioural response to it can be a very complex task that demands experimental scenarios in which it is possible to learn from these new experiences.

The challenge becomes tricky when we need to do this learning on the fly, observing on a daily basis how regrettable situations affect society while scientists in different fields struggle to gain this knowledge that in theory nobody has. In tackling this challenge, AI is emerging as a key factor in making learning processes more agile. The various techniques underlying this technology are competing to ascertain how they can be deployed to better describe the current situation, predict what might happen tomorrow and, ultimately, help us as a society determine the best ways to proceed and act. 

Tools and current uses

To this end, digital technologies and, specifically, data and AI are fostering the emergence of functions that, applied to the medical field, deliver high value to medicine. Some of them are already being implemented to help in the fight against Covid-19.

The example that comes most quickly to mind is that of the apps that have been developed in various countries to monitor the spread of the virus. These generate an ID on your mobile phone and register the IDs of other users you have been in contact with. In this way, if anyone tests positive for Covid-19 it is registered in the app and the system reviews the contact data history generated in previous days to stop the spread of the virus. 

In summary, the app tells users they have been in contact with a certain number of people on a certain number of days and produces aggregate statistics and trend data to aid the decision-making process. To comply with privacy and data protection requirements, the information is anonymised and is not communicated in real time.

In addition, these technologies can be used to make predictions about spread and even to help detect new cases. There are already examples of this type of use. Analytical approaches to diagnosing illnesses entail determining the likelihood that a certain illness is present based on a series of symptoms and patient characteristics. This means developing the capability to measure these symptoms and characteristics, and to analyse them together, improving and increasing speed of diagnosis as the data accumulate.

To carry out this diagnosis, predictive models are used that harness datasets which must have been processed beforehand, using techniques such as natural language processing, data cleansing, pattern recognition in medical images, etc. The aim is to aggregate all the information in a ready-to-analyse dataset.

The purpose of predictive models is to assign a probability of belonging to a category of interest, for example the likelihood that a new case being examined (a new patient with their measured symptomatic variables) is suffering from the illness.

Similarly, telemedicine might be made easier because healthcare professionals working in different fields can collaborate and remote diagnosis is enabled, for instance using videoconferencing systems, augmented reality solutions, data analysis and the IoT.

All disciplines, industries and individuals have roles to play and support to offer, whether by developing solutions, helping with distribution, treating patients, cleaning and disinfecting hospitals, or staying at home. In this context, data science is a powerful ally in the battle in which we are all enlisted.



Related content

Related contents

Arkema rocks its unified end-to-end Business Processes Data into the Cloud

To boost its product innovation and customer programs, Arkema decided to reshape the management of its data challenged by numerous systems. This transformational data project allows to unify data from its CRM, SRM and ERP systems into a Master Data Management (MDM) platform.

Related contents

ADEO reshapes its Data Platform with Cloud-based technologies

To sustain its strong growth, the company has laid the extension of its business on a Platform and Data strategy. To propel this ambitious plan, ADEO has decided to migrate its entire IT system into Google Cloud, applying a datacenterless strategy.

Related contents

Encryption (1/2) - The role of encryption under GDPR: solving the confusion

Encryption is a well-known and effective security measure applied by IT professionals all over the world to ensure that personal data is not subject to unauthorised access or disclosure. However, since the adoption of the EU General Data Protection Regulation (GDPR) in May 2018, encryption can also be perceived as a multi-layered legal concept that leverages data protection and privacy.