News

An online game to help diagnose tuberculosis

Researchers from the Technical University of Madrid developed a mobile phone app to help diagnose tuberculosis.

The game TuberSpot has been launched on World Tuberculosis Day, March 24, 2015.

Tuberculosis is a curable disease, but it continues to kill over a million and a half people per year around the world.

 

Every week, people around the world play over 3 billion hours of videogames. What if, instead of shooting aliens or stealing cars, playing these games helped diagnose diseases? This opportunity is the foundation of TuberSpot, an online game that has been launched on World Tuberculosis Day.

 

TuberSpot players help analyze real digitized sputum samples, however no previous medical knowledge is needed to play the game. TuberSpot has been developed by researchers from the Technical University of Madrid.

 

Almost 10 million people were infected with tuberculosis in 2013, a contagious disease that affects the lungs. Despite the fact that tuberculosis is curable if diagnosed on time and treated with antibiotics, the disease kills over a million and a half people every year. Over 95{43d78c606eaa777549217f87ed7d6a659f505d7045ec907620bbf1621907c48a} of these deaths occur in low- and middle-income countries. The countries most affected by this disease don’t have enough specialists to study all of the samples. However, thanks to new digital imaging technologies, it has become possible to virtually send blood and sputum sample images to be analyzed remotely.

 

‘Tuberspot is a proof-of-concept experiment to test the viability of collective tele-diagnosis, and establish a community of citizens scientists able to help diagnose tuberculosis online,’ says Sara Gil, Tuberspot project coordinator. ‘TuberSpot is the first worldwide crowdsourcing campaign focusing on tuberculosis samples.’

 

Detecting tuberculosis from home

 

The standard way to diagnose tuberculosis is to examine samples of sputum (a liquid expelled when coughing) with a microscope and then identify and count bacilli (bacteria that causes tuberculosis). To make a diagnosis, a specialist needs to examine an average of 100 images, a process that can take up to 20 minutes. TuberSpot distributes this time-consuming task amongst many different players. In the game, real sample images are shown to the players to ‘hunt’ the bacilli before the time is up.

 

TuberSpot’s vision goes even further: to develop a microscopy system on a mobile phone to allow telediagnosis. This is the first time the distribution of images via the internet from remote areas is possible, and this could enable the designing of a quick low-cost telediagnosis system with large-scale expansion possibilities. In fact, the sample images from the game have been digitalized and transmitted with a phone connected to a microscope, a low-cost system designed to be used in any part of the world

 

In 2013, the TuberSpot research team developed the first medical ‘crowdnosis’ game called MalariaSpot. Players identify malaria parasites in images of blood samples, leading to diagnosis. Over 12,000 MalariaSpot games have been played by volunteers from 100 different countries, showing that a combination of parasites identified in the same image by 22 players can provide a result as accurate as that of a single specialist.

 

‘Millions of hours are invested in playing videogames every day, especially by young people, and this turns them into digital experts. A tiny percentage of that time could be enough to diagnose every case of tuberculosis in the world,’ explains Dr. Miguel Luengo-Oroz, an Ashoka Fellow and principal investigator of the project. ‘TuberSpot is a simple experiment which will allow us to continue exploring the opportunities that collaborative games and social networks can offer us to tackle global health issues’,

Images

screenshot_tuberspot_tb_english
TuberSpot Screenshot.
captura_pantalla_juego_tuberspot
TuberSpot Screenshot.
Digitalización de muestras de tuberculosis con un móvil acoplado a un microscopio.
Digitazing sputum samples.
microscope
Visualization of a sputum sample.