When the Heart and Eyes Expose Lies

We’re familiar with them from movies in which criminal suspects are caught: lie detectors. They measure various biosignals that, when combined, can expose lies. Students in the medical technology program at Bremerhaven University of Applied Sciences have now independently developed and tested such a device as part of a course. Their findings will also be incorporated into future student projects. Artificial intelligence could also play a role in this.

When a person lies, the body reacts with altered biosignals. The heart beats faster, breathing and eye movements change, and there are tiny muscle twitches in the face. These reactions can be measured using sensors. In the Medical Technology program at Bremerhaven University of Applied Sciences, biosignal acquisition and processing are part of the curriculum. To explore this topic in greater depth, Prof. Dr. Olaf Eick, Professor of Medical Instrumentation, and Marcell Müller, a research and technical assistant, offered a special hands-on project to students for the first time last semester. The students were tasked with developing a lie detector. There were no specific guidelines on what it should look like.

The students opted for a multimodal approach that combines various biosignals: ECG, EMG (electromyography / muscle movements), EDA (electrodermal activity / sweat measurement), eye tracking, and respiration. “We asked ourselves which signals help detect lies. We then integrated these,” says student Jan Scheibel, who took on the role of project leader. These signals were measured using Bitalino, a device for recording biosignals, and a standard cell phone camera. The received signals are then analyzed; the students programmed the algorithm for this themselves. They used the Python programming language, the OpenSignals software, and the Swift app, which measures eye movements. They also gave their lie detector a name: Poly-53—a play on the term “polygraph” and the 53rd parallel, on which Bremerhaven is located.

To find out whether the lie detector works, the students first conducted a so-called “guilty-knowledge test.” To do this, they placed five objects in a box. Their test subjects were instructed to put one of them in their pocket. They were then connected to the sensors and asked about each object whether they had taken it. It was important that they always answer “no” to this question. In this way, the students were able to distinguish between truth and lies based on the biosignals. “In eight out of ten cases, we were able to determine which items had been taken based on the measurement results,” explains Jan Scheibel. The participants were then asked questions, such as whether they had ever lied to someone close to them. The measurement data was compared with that from the “Guilty-Knowledge Test” to detect lies. This was successful in about 70 percent of cases. “The most important factors in this context are the ECG and eye tracking. These two signals are the most reliable for detecting lies. Breathing was less important in our studies and prone to errors—for example, when participants laughed during the interview,” explains Jan Scheibel.

Even though “Poly-53” was able to expose a large proportion of the lies, there are no plans to use it in real-world applications. “Polygraphs are medical devices and require approval,” notes Marcell Müller. However, the students’ findings are not meant to gather dust in a desk drawer; rather, they will serve as the basis for further work. “Students studying medical technology in Bremerhaven will have the opportunity to further develop the ‘lie detector’ in the coming semesters. This could also lead to topics for thesis projects,” says Prof. Eick. Machine learning could also be incorporated into the project. “Artificial intelligence has great potential in medical technology without posing a significant threat to jobs. It’s not about replacing skilled workers, but about providing technical support—for example, in the evaluation of test results,” adds the professor.

If you’re interested in future-oriented topics such as personalized medicine, robotics, diagnostics and therapy using artificial intelligence, 3D printing (additive manufacturing), digital health systems, and wearables, the seven-semester bachelor’s program in Medical Technology at Bremerhaven University of Applied Sciences is the perfect fit for you. Medical technology is recognized worldwide as one of the key technologies of the 21st century. Higher life expectancy, growing health awareness, and technical innovations are driving growth in an industry that is on the rise, particularly in Germany. The demand for modern medical technology that helps detect, prevent, or treat diseases at an early stage will continue to rise in the future. The application period for this and other programs with no admission requirements runs through September 18, 2026. For more information, visit www.hs-bremerhaven.de/medizintechnik.