TL;DR
Researchers have developed a non-invasive system using capacitive sensors to monitor the breathing and heart rates of laboratory mice. This technology aims to reduce animal stress during transport and improve research data quality by providing a more humane and accurate method for physiological monitoring.
What Problem Does This Paper Address?
Animal testing is crucial for biomedical research, but it also raises ethical concerns regarding animal welfare. Minimizing stress in laboratory animals, especially during transport, is vital for both ethical reasons and for obtaining reliable research results. Current methods for monitoring physiological signals like heart rate and breathing in mice are often invasive, requiring sedation or restraint, which can introduce confounding stress factors and complicate long-term monitoring. This paper addresses the need for non-invasive, non-restrained, and non-sedated methods to capture these vital signs, particularly in the context of animal transport, to better understand and mitigate transport-induced stress.
Methods
The core of this research lies in the application of Electric Potential Integrated Circuit (EPIC) sensors for capacitive sensing. These sensors, unlike traditional electrodes, do not require direct ohmic contact with the skin, allowing for measurements through insulating layers like fur or even air. The study explored both contactless and contact-based approaches.
- EPIC Sensors: The PS25251 EPIC capacitive sensor was chosen for its high input resistance, dry-contact capacitive coupling, and ability to operate with low power.
- Contactless Sensing (System A): Initially, mice were placed in a plastic tube with sensors positioned externally. This setup primarily captured breathing signals, likely due to the movement of the chest during respiration.
- Contact Sensing (Systems B & C): To capture heart rate signals, which are of smaller amplitude, contact-based methods were investigated. This involved placing mice on platforms with sensors (System B) or a matrix of 16 EPIC sensors (System C). Different grounding techniques, including standard ground (GND) and driven right leg (DRL), were tested to improve signal quality and reduce noise.
- Signal Processing: Acquired signals were processed using MATLAB. Standard techniques like notch filtering (to remove 50 Hz electrical noise), low-pass filtering, polynomial spline fitting, and Savitzky-Golay filtering were employed to enhance signal clarity.
- Validation: For heart rate validation, a commercial photo-plethysmography system (Visitech Systems BP-2000) was used to obtain baseline heart rate data for comparison.
Datasets
The paper does not describe the use of pre-existing datasets for training or evaluation. Instead, it presents experimental data collected from laboratory mice (C57BL6/J/LDLR-/- strain) during various test trials designed to evaluate the performance of the developed sensing systems. The study generated its own benchmark data through these experiments.
Key Insights and Ideas
- Breathing is Easier to Capture Contactlessly: The initial contactless experiments (System A) successfully captured a strong signal attributed to breathing. This highlights the potential for non-invasive breathing monitoring without direct animal contact.
- Heart Rate is a Significant Challenge: Capturing the subtle electrical signals associated with heart activity proved much more difficult, especially in a non-invasive, non-sedated manner. This necessitated a shift towards contact-based methods.
- Contact Improves Heart Rate Signal Quality: By having the mouse's paws in contact with the sensors, the system was able to detect signals with patterns resembling ECG components, allowing for heart rate estimation.
- DRL Grounding Shows Promise: The use of the Driven Right Leg (DRL) signal as a grounding method appeared more effective than standard GND in reducing electrical noise, particularly at 50 Hz.
- Dual Monitoring Potential: The research suggests that with careful sensor placement (e.g., one paw on a sensor and the chest near another), it might be possible to simultaneously capture both respiratory and heart rate signals, offering a novel, integrated monitoring approach.
Key Results
- Breathing Rate: System A (contactless) successfully captured signals correlated with mouse breathing, with frequencies consistent with normal respiratory rates.
- Heart Rate (Contact-Based):
- Test Trial 2 (System B, GND) yielded a heart rate estimate of 687 bpm, which was within the range obtained by a commercial photo-plethysmography system (average of 668 and 701 bpm).
- Test Trial 4 (System C, GND, two paws in contact) produced a regular heart rate pattern with an average of 738 bpm.
- Test Trial 5 (System C, GND, one paw in contact) also yielded a heart rate estimate around 750 bpm.
- Signal Quality: While the contactless approach was good for breathing, achieving clear heart rate signals required contact. The matrix system (System C) offered better spatial coverage and potential for signal acquisition.
Limitations and Future Work
- Contact Requirement for Heart Rate: The most significant limitation is that reliable heart rate detection currently requires some form of physical contact between the mouse and the sensors. This limits its applicability for continuous monitoring during transport where animals are housed in containers.
- Positioning Sensitivity: The effectiveness of contact-based methods depends on the mouse being positioned correctly on the sensors. This can be challenging to ensure consistently.
- Need for Reference System: The authors acknowledge the absence of a simultaneous reference system to definitively verify the cardiac signals captured by their EPIC sensors.
- Future Work: The paper suggests that further development could focus on improving contactless methods for heart rate detection, potentially by increasing sensor density or sensitivity. The potential for simultaneous respiratory and heart rate monitoring using a single system is also highlighted as a promising avenue for future research. The authors also note that commercial solutions exist but often rely on conductive gels, which present their own disadvantages.
Summary updated