Week13

Week 13 Progress: Testing Strategy Design and Simulation

Objective:

This week’s objective was to investigate the physiological significance of ankle swelling as a symptom of heart failure and to design a testing methodology that simulates this condition using available tools. The aim was to assess whether the current smart sock prototype could reliably detect early signs of edema through the flex sensor readings and trend detection algorithm implemented in previous weeks.

Implementation:

To begin with, a thorough background study was conducted on the clinical causes and relevance of ankle swelling, or peripheral edema, especially in heart failure patients:

  • According to the UK’s NHS website, oedema occurs when excess fluid builds up in body tissues, often affecting the ankles and feet. It’s particularly common among elderly or sedentary individuals and can be a sign of more serious conditions like heart failure or kidney problems【NHS†source】.

  • HeartFailureMatters.org points out that swelling in the lower limbs is often due to fluid retention caused by poor heart pumping efficiency, and is one of the earliest warning signs of worsening heart failure【HFM†source】.

  • Dr. Macdonald's cardiology site further explains that chronic heart failure impairs the kidneys' ability to eliminate fluid, leading to progressive swelling in dependent regions like the ankles. Importantly, such swelling tends to happen slowly, often over hours or days【Macdonald†source】.

Based on this understanding, a test simulation protocol was developed:
A balloon and a manual air pump were used to simulate the gradual expansion of the ankle. The balloon was wrapped in a loop to mimic the wearable position and inflated very slowly to simulate progressive swelling. This method allowed for controlled increases in ankle radius and corresponding changes in the flex sensor’s resistance.

Initially, the long-term filtering was set to 1-minute intervals, with data sampling every 10 seconds—parameters ideal for real patient use but too slow for testing. For feasibility, they were adjusted to:

  • Long-term EMA interval: 15 seconds

  • Sampling frequency: Every 2 seconds


The test was repeated 10 times. In 4 out of 10 trials, the algorithm successfully transitioned into the STATIONARY state and detected a consistent downward trend in the filtered voltage, triggering the alarm. However, in the remaining cases, air leaks during inflation caused unstable flex values, mistakenly interpreted as movement.

Results:

  • A medically grounded testing protocol was successfully designed.

  • The revised filtering and sampling intervals improved test responsiveness.

  • The system showed partial success in detecting simulated swelling.

  • False negatives were mainly due to hardware limitations (air leakage).

  • Further iteration is needed for a more stable simulation environment.





















Comments

Post a Comment

Popular posts from this blog

Week9

Image
Week 9 Progress: State Machine Design and System Modularity Objective To modularize the ankle monitoring code by implementing a two-state state machine (MOVING / STATIONARY) to detect whether the user is currently in motion or in a stable position for trend detection. The goal is to trigger alarm conditions only when reliable stationary measurements are available. Implementation: State Machine Architecture The firmware now uses a simple finite state machine (FSM) with two primary states: MOVING : The user is moving, sensor data fluctuates significantly, and trend evaluation is temporarily disabled. STATIONARY : The user is still. Trend data is collected to determine if ankle swelling is occurring. 📌 Transition Logic: Transition from MOVING → STATIONARY happens when sensor variation over a defined window is small (low deviation). Transition from STATIONARY → MOVING happens when at least 3 significantly different readings appear within the window (indicating movement or p...
Read more

Week14

Image
Week 14 Progress: Data Analysis and Final Testing Improvements Objective: The primary goal of Week 14 was to complete a comprehensive round of performance tests on the ankle monitoring system and to conduct a thorough analysis of the collected sensor data. Emphasis was placed on evaluating the system’s ability to detect swelling through trend analysis, as well as verifying the accuracy of filtering techniques in eliminating short-term noise. Alongside empirical testing, the effectiveness of the core embedded logic written in C++ was further validated. Implementation: A key element of the test was real-time monitoring of ADC (Analog-to-Digital Converter) output values from the flex sensor under simulated swelling conditions. The image below shows the plotted  sampled ADC values  and their corresponding  filtered values  (exponentially smoothed): From this graph, we can observe two trends: The  blue line  shows raw sampled data, which fluctuates significantly...
Read more

Week1

Image
Week 1 Progress: System Testing and Redesign Planning During the first two weeks of the project, my main task was to test the previous prototype and develop improvement plans based on its performance. The system consists of two main parts: a wristband and an anklet module. As shown in the figure, the wristband worked relatively well. Once worn, it was able to accurately display heart rate and blood oxygen levels, although the SpO₂ readings occasionally showed inconsistencies. In contrast, the anklet system had several issues. It frequently produced errors during testing and often triggered false alarms when tested with a 3D-printed cone-shaped object. Additionally, the circuit design was overly complex and bulky, making it uncomfortable to wear.
Read more