ACOUSTIC PERCUSSION DEVICE FOR ABDOMINAL SCANNING:

ACADEMIC RESEARCH

Liver enlargement and inflammation, also called hepatomegaly, is a significant indicator of the presence of a disease process. For years, clinicians have been using acoustic percussion to assess hepatic size. However, percussion can be seen as subjective and misleading or inaccurate.

​

There is an opportunity to develop an assistive, portable device that allows telemetry. With this, the doctor can remotely diagnose a patient from their homes, while also ensuring a more reliable and less subjective diagnosis.


In populations with a high percentage of elderly patients requiring frequent checks-up, remote assessment is key.
 

The project aims to mimic percussion from a trained doctor through hardware and analyse the recorded acoustic signal. 

Robotic Research Project
Raspberry Pi
Python
Data Analysis
Signal Processing
Electronics
Mechanisms

 Pilar Zhang Qiu | Oliver Thompson | Jacob Tan | Ben Cobley 

Art
 
 

ONGOING PROJECT

 Estimated End Date:  August 2020

AvgFar_Scalogram.png

Example of a CWT Scalogram after denoising the audio file.

Relevant-Bands2.png

Intensity of different frequency bands during percussion.

We have developed two physical impact prototypes mimicking the arm-wrist-hand percussion motion performed by doctors.

​

We are currently testing on different abdominal areas on healthy subjects and analysing the audio signal, which I led.

​

First, the noisy signal is smoothened through the Savitsky-Golay Filtering method. The simplified model is then analysed using Continuous Wavelet Transform or Decomposition. Our goal was to find changes in signal behaviour between different abdominal areas (e.g. lungs, liver or abdomen).

​

We identified and theorised upon the dampening effect of nodule depth and are currently expanding the research.