[3-I-5-04] Performance Evaluation of Portable Medical Devices:
A Comparative Study of Hemoglobin Sensors and Uric Acid Sensors
Portable Medical Devices, Medical Sensors, Medical Sensor Evaluation, Digital Health, Telehealth
Purpose: The accurate measurement of the vital information of the patients by portable sensors is critical in telehealth systems to facilitate correct diagnoses, and ensure patient safety. This study evaluates the performance of commonly used portable sensors for hemoglobin and uric acid measurements, and recommends the sensors with acceptable performance.
Methods: This experiment was conducted within a local internal medicine clinic in Japan and blood samples were collected from the regular patients who needed hemoglobin and uric acid tests. The minimum sample size for each test was 50 and the target subjects were both male and female within the age range of 20-70 years. Then the same samples were tested by sending to a clinical pathology laboratory, which is considered as the gold standard, and also by using the strip-based portable sensors in the clinic. A Normalized Root Mean Square Error (NRMSE) below 0.15 was the threshold for deeming the portable sensor's outcome as acceptable.
Results: A total of eight hemoglobin and five uric acid sensors from markets in Asia, Africa, and the USA were tested. Our findings reveal that while a few hemoglobin sensors exhibited satisfactory performance, many sensors, particularly from Asian developing countries and the USA, produced inaccurate results. Uric acid sensors fared even worse, with only one sensor for the USA meeting acceptable standards.
Conclusion: Given the rising global demand for portable sensors, there exists an urgent need to enhance product quality, thereby improving healthcare delivery and patient outcomes.
Methods: This experiment was conducted within a local internal medicine clinic in Japan and blood samples were collected from the regular patients who needed hemoglobin and uric acid tests. The minimum sample size for each test was 50 and the target subjects were both male and female within the age range of 20-70 years. Then the same samples were tested by sending to a clinical pathology laboratory, which is considered as the gold standard, and also by using the strip-based portable sensors in the clinic. A Normalized Root Mean Square Error (NRMSE) below 0.15 was the threshold for deeming the portable sensor's outcome as acceptable.
Results: A total of eight hemoglobin and five uric acid sensors from markets in Asia, Africa, and the USA were tested. Our findings reveal that while a few hemoglobin sensors exhibited satisfactory performance, many sensors, particularly from Asian developing countries and the USA, produced inaccurate results. Uric acid sensors fared even worse, with only one sensor for the USA meeting acceptable standards.
Conclusion: Given the rising global demand for portable sensors, there exists an urgent need to enhance product quality, thereby improving healthcare delivery and patient outcomes.