100x Engineers
May 25, 2026
How Dogs Detect Cancer Through Breath Analysis
Key Takeaways
- 1Diagnoses, a Bengaluru startup, uses trained dogs to detect cancer from breath samples with 91% accuracy across 10 cancer types
- 2The process involves breathing into a mask for 10 minutes, which is then analyzed by dogs wearing AI-powered headsets that read their brain signals
- 3This method could cost 10 times less than conventional cancer screening, making it potentially transformative for countries like India
- 4The startup has already presented clinical data at ASCO (American Society of Clinical Oncology) and is running trials with major health providers
- 5AI-powered brain signal analysis translates dogs' olfactory detection into predictive cancer risk data through machine learning
Chapters
1. The Problem & The Solution
Traditional cancer screening is expensive and inaccessible. Diagnoses solves this by leveraging dogs' natural ability to detect odor biomarkers in breath.
2. How the Technology Works
Patients breathe into a mask for 10 minutes. Trained dogs sniff the mask while wearing a custom AI headset (Dog Sense) that captures brain signals in real time.
3. AI Integration & Results
Brain signal data from the dogs is processed by machine learning software to assess cancer risk across 10 different cancer types, achieving 91% accuracy.
4. Clinical Validation & Funding
The startup has presented data at ASCO and is conducting clinical trials with Narayana Health and Aster CMI. Recently funded by Accel and Process.
5. Impact & Market Opportunity
At 10 times lower cost than conventional screening, this technology could be a game-changer for early cancer detection in resource-limited settings like India.
Glossary
- Dog Sense
- A custom AI-powered headset worn by dogs that reads their brain signals in real time during the cancer detection process
- ASCO
- American Society of Clinical Oncology; one of the world's largest oncology conferences where Diagnoses presented their clinical data
- Biomarkers
- Measurable biological substances (in this case, odor compounds in breath) that indicate the presence of disease
- ML Software
- Machine learning software that processes brain signal data to output cancer risk predictions
- Deep Tech
- Technology companies that develop complex, science-based solutions addressing fundamental problems