Embedded Files
Description
Our final sensing system integrates a 6V power supply to drive a DC motor, with system current and voltage continuously monitored via an INA219 analog sensor. Data processing and control logic are handled by a Teensy 4.0 microcontroller, which modulates power delivery and actuates an LED indicator and audible buzzer for real-time user feedback. To house these hardware elements, the 3D-printed handle was redesigned to optimize ergonomics and provide secure internal mounting for the electronics and sensor payload.
Retrofitted Handle
Flowchart Breakdown
The system operates on a state-machine architecture designed to detect clot clearance. Following an initial Warmup phase, the system enters the Monitoring state while awaiting clot engagement. A cumulative sum (CUSUM) drop transitions the system to a Provisional cleared state, which can either trigger a 're-engage' loop back to monitoring or advance to Confirmed cleared once a 20-second counter satisfies the MIN_CLEAR_DROP gate. Upon confirmation, the LED and buzzer fire to alert the user. If an anomaly occurs during monitoring or provisional phases, the system defaults to an Uncertain state, blinking the LED and initiating a 1- or 10-minute timeout.
Testbed: Experimental Procedure
To evaluate the feedback performance of our sensing system, we used a test bed obtained from the previous Capstone team. It features a continuous 4L/min fluid loop driven by a micro-diaphragm pump that simulates physiological vascular flow conditions. During a typical experimental trial, a graft with a clot of either size 1mL or 3mL is placed between the "vein" and "artery" tubes. our redesigned ergonomic handle is engaged to clear the occlusion and notify the surgeon with clot status.
Tests were also performed to mimic the viscosity of blood by using a mixture of glycerol and water. Results determined that viscous effects from glycerol were negligible and system sensing continued performing well.
Butterworth-Filtered Motor Current Profiles for Representative Clot Clearing Trials
State Transitions over Current Profile
To evaluate the efficacy and adaptability of our sensing framework, representative clot clearing trials were characterized across varying clot hardness profiles and motor states, as illustrated in the top plot. This demonstrates how motor current inherently scales with fluid resistance and clot density, mapping Butterworth-filtered current profiles for soft (4% gelatin), medium (8% gelatin), and hard (12% gelatin) clot models.
The bottom plot demonstrates the algorithm's state machine execution in real time against a dynamic baseline mean during an 8% gelatin clot trial. As the device navigates through, 'Clot Engaged' triggers a distinct current spike, followed closely by a 'Provisional Cleared' state identification as a sharp cumulative sum (CUSUM) drop occurs. Once the signal stabilizes below the baseline threshold for the required 20-second verification window, the system establishes a 'Confirmed Cleared' state, instantly triggering the integrated LED and buzzer to deliver closed-loop feedback to the user.
Page updated
Report abuse