M Dougherty

SMELL-RS

A self-administered, digital test for olfactory dysfunction that measures how well people tell smells apart and how faint a smell they can detect, without asking them to name anything.

Years
2019 to 2025 (research), 2026 (preprint)
Role
Research consultant under Dr. Andreas Keller, second author
Contributions
Odor-mixture preparation, UI design, figure illustration, participant testing, device prototyping
Paper
medRxiv, Mar 31, 2026
DOI
10.64898/2026.03.28.26349316
A laptop showing the SMELL-RS welcome screen beside the white SMELL-RS device with its round sniffing port, on a table by a window overlooking a river.
The SMELL-RS device and test software. Photo: M Dougherty

Why it matters

Smell testing is important in clinical care but is used far less than it should be, because the standard tests take time, staff and materials. Many tests also depend on naming smells, which mixes up smelling with language and culture.

SMELL-RS tests two things with no naming at all. The first is olfactory resolution (SMELL-R), the ability to tell odor mixtures apart. The second is olfactory sensitivity (SMELL-S), the faintest odor a person can detect. The test is digital and fully automated, so people can take it on their own.

How the test works

Each trial presents three smells. Two are the same and one is different, and the person picks the odd one out on screen. The test adapts as it goes, and no one has to put a smell into words.

I designed the step-by-step instructions people see before they start.

How a trial works. Animation: M Dougherty.

Animation, no sound. A person holds a nozzle connected to the SMELL-RS device and sniffs three times while a screen shows the numbers 1, 2 and 3. One of the three smells is different. They tap the number that smelled different. When they are unsure, they guess, and the test moves on.

From prototype to device

Early line drawing: a person at a table with racks of odor vials.
Early concept with racks of vials
The device's parts laid out and numbered: housing, tubing, cartridge, tablet and controller.
The parts of the device
Dr. Andreas Keller leans in to sniff from the white cone-shaped port on the side of the SMELL-RS device, a laptop beside it, at a window with a city skyline behind.
Dr. Andreas Keller using the device. Photo: M Dougherty

What the study found

The authors conclude that SMELL-RS is rapid, fully automated, reliable and accurate, and suited to self-administration in clinical settings.

Participants
100 people with and without smell dysfunction, recruited at a tertiary care center in Geneva
Reliability
Test-retest ICC = 0.71 (p < 0.0001) for the composite score
Agreement
Correlated with the Sniffin’ Sticks composite score, r = 0.68 (p < 0.0001)
Time
Mean 5.9 min for SMELL-R and 5.5 min for SMELL-S, two to three times faster than the matching Sniffin’ Sticks tests
Status
Preprint, not yet peer reviewed

My part

Across six years as a research consultant I worked on SMELL-RS from the device to the screen. I ran user testing, collecting data by testing the device with people, and prototyped rapidly with Rockefeller's PIT team, 3D modeling and printing parts.

I made the odor cartridges, tested concentrations and mixtures, and tested materials and procedures for the cartridge inlays. I ran quality assurance on the device, including mini PID and airflow measurements.

I designed the front end of the test app with a focus on accessibility and produced its logo, photos, graphics, the paper's figures and the animation that explains the test. Throughout, I worked with the other teams and departments on the project.

Press

Citation

Hsieh JW, Dougherty M, Poulopoulou A, Blidariu D, Senn P, Hopper R, Patel D, Maggioni E, Obrist M, Vosshall LB, Keller A, Landis BN. SMELL-RS: A Self-administered, Digital Test for Olfactory Dysfunction that is Rapid, Reliable, and Accurate. medRxiv 2026. doi:10.64898/2026.03.28.26349316

A collaboration between Geneva University Hospitals, the Laboratory of Neurogenetics and Behavior at The Rockefeller University, and Hynt Labs. SMELL-RS is patented by The Rockefeller University.