AI is when your smartphone knows that you have COVID-19 | DW | 04.11.2020

The fact that smartphones know an awful lot about us is nothing new. And many people have long since accepted that their devices — or more precisely the providers behind the apps  that they use — have probably learned more about them than the users know.

Soon, a new app could be launched that will allow users to diagnose an asymptomatic coronavirus infection based on the sound of their coughing or speaking.

Though the AI technology has already proven itself to be able to identify the coughs of infected individuals, it still has to learn more to avoid misdiagnosing noninfected people. If it succeeds in doing so, the app could at some point complement Germany’s existing coronavirus tracing app,  which traces contacts with infected persons. 

Room for improvement – Germany’s contact tracing app

Coughing sounds contain biomarkers

Three computer scientists from the Massachusetts Institute of Technology (MIT)   came up with the idea of using sound analysis to detect infections. Jordi Lugarta, Ferran Hueto and Brian Subriana took sound recordings of 5,320 people in April and May – both participants infected with the coronavirus and participants without. In addition to coughing sounds, they analyzed speech. First, the researchers fed the sound data of 4,256 participants into the computer, which analyzed them with the help of a convolutional neural network.

The researchers were looking for acoustic biomarkers. Those include certain characteristic features in the sounds that they had already found in earlier studies on Alzheimer’s patients. They then tested what the machines had learned from samples on the remaining 1,064 participants of the study.

Many hits — but please do not panic in case of false positive results!

The results were quite promising. “When validated with subjects diagnosed using an official test, the model achieves COVID-19 sensitivity of 98.5%,” the researchers wrote in their study, which was submitted to the IEEE Open Journal of Engineering in Medicine and Biology  for publication.

The specificity was 94.2% in the group. This would mean that about 1 in 20 people taking such a test would have received a false positive result.

Our voices speak volumes

“For asymptomatic subjects, it achieves sensitivity of 100% with a specificity of 83.2%,” the authors report. This would mean that every otherwise potentially undetected COVID-19 case was correctly diagnosed. However, almost one in five participants received a false positive warning. 

In other words, the specificity would certainly have to be improved in order to make such an app really workable in practice. If it is used by a large number of people every day, the sheer number of false positive results could otherwise quickly lead to a run on COVID-19 test laboratories and exhaust their capacities.

It is possible, however, that the specificity values could be further improved if the computers were fed with more data and learned more about coughing sounds with deep learning technology.

No replacement but supplement to laboratory tests

In any case, the researchers concluded that “AI techniques can produce a free, non-invasive, real-time, any-time, instantly distributable, large-scale COVID-19 asymptomatic screening tool to augment current approaches in containing the spread of COVID-19.”

The MIT scientists suggest that the method could be used to test students, workers and office employees on a daily basis. Should a warning come up, the people concerned could still do a laboratory test to find out if the app was right. 

From the fingerprint to biometric data

A standard in modern forensics for 125 years

In 1891, a Croatian born, Argentine criminologist, Juan Vucetich, started building up the first modern-style fingerprint archive. Since then, fingerprints have become one of the main forms of evidence used to convict criminals. Here, a police officer spreads dust on the lock of a burglarized apartment. Fingerprints become visible.

From the fingerprint to biometric data

Archiving and comparing prints

He uses an adhesive film to capture the fingerprint. Then he glues it to a piece of paper. In the past, comparing fingerprints was a painstaking affair. Officers had to compare fingerprints found at the scene of a crime, one-by-one, with those of possible suspects. These days computers do the job.

From the fingerprint to biometric data

No more ink

Taking fingerprints used to be a messy affair – with ink and dirty hands. These days scanners have replaced the inky mess. And the data can immediately be sent to a database and turned into biometrical data.

From the fingerprint to biometric data

Fingerprints form an identity

The computer identifies typical spots within the ridge patterns of the fingerprint. These include forks in the lines, spots and the location of the center of the print. Fingerprints are never the same between two people – not even with identical twins.

From the fingerprint to biometric data

Vote early and vote often!

No chance! Here, officials use fingerprint scanners during an election in Nigeria. It’s how they make sure the people voting are registered voters and that they only vote once.

From the fingerprint to biometric data

Who entered Europe where?

This is an important question for officials who have to decide about the refugee or asylum status of applicants. In the European Union all migrants are supposed to have their fingerprints taken at the first point of entry – provided, of course, the local police officers are equipped with the scanners.

From the fingerprint to biometric data

Hands off! It’s my data!

Many smartphones now come with fingerprint recognition software, such as the iPhone’s Touch-ID. The owner of the phone unlocks it with his fingerprint. If someone else finds or steals the phone, they have no way of getting at any encrypted data within.

From the fingerprint to biometric data

Secure ATM banking

This is an Automatic Teller Machine (ATM) in the Scottish town of Dundee. Customers wanting to withdraw money need to show biometric proof of identity – in the form of a fingerprint. Not good news for pickpockets.

From the fingerprint to biometric data

Fingerprint inside the passport

Since 2005, German passports, and many other passports, contain a digital fingerprint as part of the biometric information stored on a RFID (radio-frequency controlled ID) chip. Other information on the chip includes a biometric passport photo. The facial image is similar to fingerprints: no two images are alike.

From the fingerprint to biometric data

When computers recognize faces

Facial recognition software, which uses biometrics, is well advanced. It is possible to identify suspects within large crowds, with surveillance cameras. Also internet services and private computer owners are increasingly making use of facial recognition software to sort holiday pictures and tagging them to names.

From the fingerprint to biometric data

The inventor of the genetic fingerprint

Alec Jeffreys discovered DNA-fingerprinting almost accidentally in 1984 during research at the University of Leicester. He identified a specific pattern on DNA segments, which were different for every human. He created a picture, which looks like a barcode at the supermarket.

From the fingerprint to biometric data

A barcode for every human

Germany’s Federal Criminal Police Office (BKA) started storing such barcodes in a federal database in 1998. Investigators have since solved more than 18,000 crimes, using genetic fingerprints.

From the fingerprint to biometric data

Clearing the innocent

It’s not just criminals who get identified. Many innocent people can be cleared of criminal charges through good identification. For some, technology has saved their lives. Kirk Bloodsworth spent almost nine years on death row. The US Innocence Project has proved the false incarceration of more than 100 people using DNA evidence.

From the fingerprint to biometric data

Clarity for victims’ families

The first big test for DNA-fingerprinting came with the mass murder of Srebrenica. Bodies, exhumed from mass graves, were systematically identified using DNA techniques. They were then reburied by their loved ones. Here, five year old Ema Hasanovic pays last respects to her uncle. More than 6,000 victims of the massacre – mostly men – were identified using DNA-fingerprinting.

From the fingerprint to biometric data

Biometric data on your phone and computer

You may be surprised, but there’s biometric information in sounds and other digital data. Voice recognition software can, for instance, identify people making threatening phone calls – the human voice is also unique. And don’t forget: we leave all kinds of digital traces on the internet, which hold clues to who we really are.

Author: Fabian Schmidt