Intravenous pole-integrated automated urinary status-monitoring technique using image-based artificial intelligence:...

Area of Science:

• Biomedical Engineering
• Medical Informatics
• Urology

Background:

• Long-term monitoring of catheterized patients requires tools to track urine output and characteristics.
• Current methods for monitoring urine color and volume are often manual and labor-intensive.
• Automated systems are needed to improve efficiency and patient safety in urinary care.

Purpose of the Study:

• To develop and validate an automated system for monitoring urine color and void patterns in catheterized patients.
• To assess the accuracy of a deep learning-based technique for detecting urinary disease symptoms and estimating urine volume.
• To evaluate the potential of an intravenous (IV) pole-integrated system for real-time patient monitoring.

Main Methods:

• A novel urination-status monitoring technique was integrated into an IV pole.
• Deep learning was employed to detect liquid color and volume within the urine bag.
• A proof-of-concept simulation study used various simulated urine samples to test the system's performance.
• Long-term testing over 24 hours was conducted to evaluate accuracy and reliability.

Main Results:

• The system achieved low error rates in predicting in-bag liquid volume for different urine conditions (e.g., 2.00 ± 4.93% for purple urinary bag syndrome).
• The error rate for bag-flush request alarms was between 0.71-1.08%.
• During 24-hour testing, the system accurately classified urinary disease symptoms (100% accuracy) and estimated total void volume with varying error rates depending on the condition (e.g., 8.75 ± 4.61% for oliguria).

Conclusions:

• The proposed IV pole-integrated urinary monitoring technique shows promise for real-time, simplified monitoring of catheterized patients.
• This technology can potentially enhance patient safety, particularly for those with renal and urological conditions.
• Further clinical evaluations with actual urine samples are necessary to confirm these findings.