A new study from the Weizmann Institute of Science suggests that human breathing patterns can function as highly accurate biometric identifiers. Researchers found that nasal airflow measurements could distinguish individuals with 96.8 percent accuracy, placing breathing alongside other behavioral biometrics like voice and gait in terms of reliability.
Long-Term Breath Monitoring Enables Real-World Biometrics
The study, led by neuroscientists Noam Sobel and Timna Soroka, monitored 100 healthy young adults using wearable nasal airflow sensors over a full 24-hour period. This approach captured respiration patterns across sleep, activity, and rest, marking a departure from earlier biometric studies that relied on short, controlled recordings.
Using a software toolkit called BreathMetrics, the team extracted 24 distinct features from the participants’ breathing waveforms, including inhale duration, sniff frequency, and pause variability. The identification accuracy remained stable in multiple retests conducted over two years, matching or exceeding the performance of many voiceprint systems under lab conditions.
“I thought it would be really hard to identify someone because everyone is doing different things, like running, studying, or resting,” said Soroka. “But it turns out their breathing patterns were remarkably distinct.”
Security, Health Monitoring, and Privacy Implications
The study suggests breathing patterns could complement existing biometric authentication systems, particularly in applications involving passive or continuous monitoring. However, researchers also emphasized the potential for healthcare applications. Breathing signatures were found to correlate with body mass index, sleep-wake cycles, and mental health metrics such as anxiety levels. Participants with elevated anxiety scores, for instance, showed shorter inhales and more irregular pauses during sleep.
“We intuitively assume that how depressed or anxious you are changes the way you breathe,” said Sobel. “But it might be the other way around. Perhaps the way you breathe makes you anxious or depressed. If that’s true, we might be able to change the way you breathe to change those conditions.”
While the findings are promising, practical deployment remains a challenge. The current setup requires a nasal cannula, which can be uncomfortable for everyday use. The team is exploring more discreet sensor designs that could be embedded in eyeglasses or wearable masks, but such developments are still in early research stages.
Privacy considerations may also become relevant as this type of biometric data could reveal sensitive information about a person’s mental or physical health. Experts suggest that strong data governance and consent frameworks will be essential if breathing-based biometrics are to be adopted at scale.
Follow-up studies are already underway to test whether breathing-focused therapies could improve mental health outcomes, moving the research from diagnostics to potential interventions.
Sources: Courthouse News Service, New Scientist
–
By Cass Kennedy
