MamaOpe Medicals · Medical devices · Biomedical signal processing
MamaOpe Smart Auscultation Device
Bridging hardware, biomedical science and software to build medical technology.
- My role
- Team member across hardware, signal processing and Android
- Status
- Research & development
- Areas
- Healthcare & Biomedical, Research & Innovation, Mobile & Web Applications
- Stack
- MATLAB · Signal processing · Java · Kotlin · Android
The problem
Pneumonia is the leading cause of death in children under five in Sub-Saharan Africa. Diagnosis often relies on a health worker listening to chest sounds and counting breaths, skills that vary with training and workload.
Where specialist support and diagnostic equipment are scarce, respiratory assessment is hard to do consistently.
Context
The device is MamaOpe's hardware product, separate from but complementary to HealthNavi. The product-level description below comes from the Royal Academy of Engineering; it is not a claim that I designed every component.
Publicly documented characteristics: four microphones that capture lung sounds simultaneously from the base and apex of each lung; suction-cup attachments; adjustable probes for different body sizes; LED status indicators; a mechanical power switch; and Wi-Fi transfer of recordings to a phone or tablet app.
The solution
Structured, simultaneous acquisition of chest sounds from multiple positions, transmitted to an Android app that estimates respiratory rate and flags abnormal sound patterns, with recordings stored in a screening history.
Architecture
- 01
Acoustic sensors
Four microphones on adjustable probes
- 02
Signal acquisition
Simultaneous multi-site recording
- 03
Wireless link
Wi-Fi transfer to phone or tablet
- 04
Android app
Java / Kotlin · screening history
- 05
Signal processing
Respiratory-rate estimation, sound analysis
- 06
Clinical information
Measurements to support assessment
What I did
Hardware
- Took part in hardware design and component-level work, prototyping, assembly and testing.
- Worked on hardware integration and device communication with the software ecosystem.
Biomedical signal processing
- Developed MATLAB-based processing of chest-sound recordings.
- Worked on respiratory-rate estimation algorithms using signals relevant to pneumonia screening.
Mobile
- Developed Android application features in Java and Kotlin, including device integration and handling of captured respiratory data.
Engineering decisions
- Multiple simultaneous listening sites
- Capturing several chest positions at once, rather than moving a single stethoscope, makes acquisition more structured and repeatable across operators.
- Phone as the processing and display hub
- Sending recordings over Wi-Fi to a phone or tablet keeps the device simple and lets analysis, storage and updates live in software.
Challenges
- Real-world signals are noisy
- Chest sounds from children are affected by movement, crying and ambient noise, so signal processing has to cope with imperfect recordings.
- Hardware and software must agree
- Device communication, data formats and the app have to be developed together. A change on one side affects the other.
Outcomes
- Recognised through MamaOpe's founder's selection for the Royal Academy of Engineering Leaders in Innovation Fellowship (LIF Advance, 2024).
- The Grand Challenges Africa grant for MamaOpeAI (2024) builds clinical decision support on top of the device's respiratory data.
Technologies
- MATLAB
- Signal processing
- Java
- Kotlin
- Android
- Embedded hardware
- Wi-Fi
What I took away
Physical products force end-to-end thinking: sensors, firmware, connectivity, software and the person holding the device are one system.