Latin America. Every time we send information over the internet we expose ourselves to the manipulation of our data. Developing a new tool to protect digital data in a secure and affordable way is what recent work led by researcher Tanya Koohpayeh Araghi, from the Internet Interdisciplinary Institute (IN3) of the Open University of Catalonia (UOC), has done.
The study, which focuses on medical imaging, offers significant advances in protection through an innovative digital watermarking technique, a kind of imperceptible "digital seal" that is placed on these documents. The applications of this technique range from medical images – X-rays, CT scans or MRIs – and telemedicine reports to the protection of multimedia content to prevent the spread of fake news.
Published in open access in the journal Multimedia Tools and Applications, the new method combines two mathematical tools commonly used to create digital watermarks: discrete wavelet transformation (DWT) and singular value decomposition (SVD), and adds greater depth to the latter (something like performing a second decomposition). The result is easier-to-create digital watermarks that provide a very robust defense against filtering, digital signal processing, and geometric attacks.
Prevent manipulation of medical data
"Our work is focused on developing enhanced digital watermarking techniques," explains Koohpayeh, a researcher in the K-ryptography and Information Security for Open Networks (KISON) group. "Not only do they improve perceptibility and robustness, but they also reduce the computational load, making them more efficient and accessible," he adds.
These findings, which the researchers tested on medical imaging, are especially relevant.
Whenever doctors use the internet to transfer images or make a diagnosis, the data is vulnerable to attack. Therefore, it is essential that there is some type of protection that ensures the veracity and confidentiality of the images. "It's critical to put digital watermarks on all types of medical images, resulting from X-rays, CT scans, or MRIs," Koohpayeh explains. "It's the way to ensure the authenticity and integrity of health data," he says.
A tool against fake news
The technique of water marking and digital security tools can also be used in other contexts, such as the protection of devices connected to the Internet of Things, transactions with cryptocurrencies or the dissemination of fake news.
It is precisely on the fight against disinformation that Koohpayeh and his team are now concentrating their efforts. Its objective is to offer techniques that allow tracing the origin of fake news, recognizing legitimate information and detecting its manipulation.
"It is important for the public to be aware of the risks associated with fake news, as even small manipulations of images or videos can have serious consequences on public perception and decision-making," explains Koohpayeh.
Koohpayeh and his team's research is currently focused on videos or images, but it could be used in other formats. It also includes tools such as blockchain technology or artificial intelligence, to detect the most subtle fakes.
However, digital watermarks remain their main object of study. By embedding information imperceptibly into content without adding a large computational load, they offer a versatile tool: robust marks make it possible to identify the origin of the content – for example, if it comes from recognized media – and fragile and semi-fragile marks are effective in detecting any fraudulent manipulation – for example, a retouching of an official photograph.
Becoming aware of our digital vulnerability
At a time when deepfakes – an artificial intelligence technique that allows fake videos of people to be edited that appear to be real – and fakes have taken hold in the digital world, it is also necessary for the public to become aware of the importance of these tools to protect their own well-being. Our images and data are becoming more and more vulnerable, even if we are not aware of it.
"It is important for the public to understand the risks associated with the dissemination of images online and take steps to protect their own information. By being aware of the importance of verifying the authenticity of digital content, we can work together to combat misinformation and protect the integrity of information online," says the expert.
Two projects funded by the National Institute of Cybersecurity
The UOC's KISON research group currently has two initiatives recently funded by the National Institute of Cybersecurity (INCIBE). The team, led by David Megías, professor and director of IN3, participates in the ARTEMISA chair, coordinated by Pompeu Fabra University, which among its objectives seeks to promote scientific vocations in the professional field of cybersecurity.
In addition, KISON has started, together with the Universitat Autònoma de Barcelona, the DANGER project, to generate solutions for the filtering of false or malicious content, protect Internet of Things devices and transactions with cryptocurrencies, and improve the security of 5G networks.
KISON's research supports the UN Sustainable Development Goals (SDGs) 4, quality education; 5, gender equality, and 9, industry, innovation and infrastructure.
Text published by the Open University of Catalonia.
