PATTERN RECOGNITION SYSTEM AND METHOD FOR THE DETECTION OF STUCK FLUID DROPLETS IN A FLUID DELIVERY LINE OF AN INFUSION SYSTEM

• US 20130046508A1
• Filed: 08/17/2012
• Published: 02/21/2013
• Est. Priority Date: 08/19/2011
• Status: Active Grant

• Current Assignee: ICU Medical Incorporated
• Original Assignee: Iii, John Hicks Dumas
• Inventors: Sur, Kunal, Belkin, Anatoly S., Ruchti, Timothy L., Kotnik, Paul T., Dumas, John Hicks III
• Granted Patent: US 9,114,217 B2
• Time in Patent Office: Days
• US Class Current: 702/183
• CPC Class Codes: A61M 2205/18 with alarm
• CPC Class Codes: A61M 2205/50 with microprocessors or computers
• CPC Class Codes: A61M 5/142 Pressure infusion, e.g. using pumps
• CPC Class Codes: A61M 5/365 Air detectors A61M5/1684 takes precedence; in extracorporeal blood circuits A61M1/3626

Abstract

An infusion system includes a pump, a fluid delivery line, a sensor, a processor, and a memory. The fluid delivery line is connected to the pump for delivering fluid. The sensor is connected to the fluid delivery line for emitting and receiving signals to detect whether there is air in the fluid delivery line. The processor is in electronic communication with the pump and the sensor. The memory is in electronic communication with the processor. The memory includes programming code for execution by the processor. The programming code is configured to analyze the signals to determine if a stuck fluid droplet is stuck within the fluid delivery line at a location of the sensor.

Claims

1. An infusion system comprising:

a pump;

a fluid delivery line connected to the pump for delivering fluid;

at least one sensor connected to the fluid delivery line for emitting and receiving signals to detect whether there is air in the fluid delivery line;

a processor in electronic communication with the pump and the at least one sensor; and

a memory in electronic communication with the processor, wherein the memory comprises programming code for execution by the processor, and the programming code is configured to analyze the signals to determine if a stuck fluid droplet is stuck within the fluid delivery line at a location of the at least one sensor.

2. The infusion system of claim 1 wherein the programming code is configured to determine if the stuck fluid droplet is stuck within the fluid delivery line at the location of the at least one sensor by determining a pattern relationship over time between a plurality of digital measurements of the signals.

3. The infusion system of claim 2 wherein the programming code is configured to determine if the stuck fluid droplet is stuck within the fluid delivery line at the location of the at least one sensor by comparing the determined pattern relationship over time to a programmed pattern over time algorithm which is indicative of the presence of the stuck fluid droplet.

4. The infusion system of claim 2 wherein the programming code is configured to analyze the plurality of digital measurements over multiple differently programmed states to determine the pattern relationship.

5. The infusion system of claim 4 wherein the multiple differently programmed states, programmed in the programming code, comprise a monitoring state, a trigger transition state, a droplet detection state, and a droplet count state.

6. The infusion system of claim 4 wherein the programming code is programmed so that during a monitoring state the digital measurements of the signals are taken, average ADC values and average variance values of the signals are calculated, and the monitoring state transitions to a trigger transition state.

7. The infusion system of claim 4 wherein the programming code is programmed so that during a trigger transition state, depending on the digital measurements of the signals, on a state status, and on comparisons of the measurements of the signals to thresholds, the transition state transitions to either a monitoring state, a droplet detection state, or a droplet count state.

8. The infusion system of claim 4 wherein the programming code is programmed so that during a droplet detection state, depending on the digital measurements of the signals, and on comparisons of the measurements and average ADC values of the signals to thresholds, the droplet detection state either transitions to a monitoring state or enters a droplet count state.

9. The infusion system of claim 4 wherein the programming code is programmed so that during a droplet count state, depending on the digital measurements of the signals, and on comparisons of the measurements, average ADC values of the signals, and average variances of the signals to thresholds, the droplet count state either transitions to a monitoring state, enters a droplet detection state, or signals an alarm.

10. The infusion system of claim 2 wherein the programming code is configured to determine that the stuck fluid droplet is present if the determined pattern relationship starts at or below a known liquid signal ADC value at a first period in time, increases from at or below the known liquid signal ADC value over a primary threshold ADC value at a second period in time, decreases from over the primary threshold ADC value to between the primary threshold ADC value and a secondary threshold ADC value at a third period in time, and passes a stability test while being disposed at a fourth period in time between the primary threshold ADC value and the secondary threshold ADC value.

11. The infusion system of claim 10 wherein the programming code is configured to determine during the stability test that the stuck fluid droplet is not present if the signals during the fourth period in time vary more than a set amount, drop down to or below a previously observed fluid level, drop below the secondary threshold ADC value, or increase over the primary threshold ADC value.

12. The infusion system of claim 2 wherein the programming code is configured to determine during a triggering state, representing the possible beginning of the stuck fluid droplet, whether an ADC value of one of the digital measurements exceeds a primary threshold ADC value, or is less than a trigger threshold ADC value but greater than a minimum ADC value, and if so analyzes the digital measurements of the signals further to determine if the stuck fluid droplet is present, and if not determines that a stuck fluid droplet is not present.

13. The infusion system of claim 2 wherein the programming code is configured to determine ADC mean values for the signals to determine the pattern relationship.

14. The infusion system of claim 2 wherein the programming code is configured to conduct a test for false positives, indicating that the stuck fluid droplet is not present, by calculating and comparing at different times at least one of ADC mean values, variance mean values, or F critical calculations for the digital measurements.

15. The infusion system of claim 1 wherein the programming code is configured to dynamically calibrate the at least one sensor during fluid delivery based on real-time measurements of the signals.

16. The infusion system of claim 15 wherein the programming code is configured to update a dynamic range of the real-time measurements of the signals with the dynamic range ranging between a minimum recorded measurement of the signals and a maximum recorded measurement of the signals.

17. A method for detecting a stuck fluid droplet in a fluid delivery line of an infusion system comprising the steps of:

pumping fluid through the fluid delivery line over at least one sensor;

emitting and receiving signals from the at least one sensor into and from the fluid delivery line; and

processing measurements of the signals, using a processor, to determine whether a stuck fluid droplet is stuck within the fluid delivery line at a location of the at least one sensor.

18. The method of claim 17 wherein the step of processing measurements of the signals with the processor further comprises determining whether the stuck fluid droplet is stuck within the fluid delivery line at the location of the at least one sensor by determining a pattern relationship over time between a plurality of digital measurements of the signals.

19. The method of claim 18 wherein the step of processing measurements of the signals with the processor further comprises comparing the determined pattern relationship over time to a programmed pattern over time algorithm which is indicative of the presence of the stuck fluid droplet.

20. The method of claim 18 wherein the step of processing measurements of the signals with the processor further comprises analyzing the plurality of digital measurements over multiple differently programmed states, programmed in the processor, to determine the pattern relationship.

21. The method of claim 20 wherein the multiple differently programmed states programmed in the processor comprise a monitoring state, a trigger transition state, a droplet detection state, and a droplet count state.

22. The method of claim 20 wherein the step of processing measurements of the signals with the processor further comprises, during a monitoring state, taking the digital measurements of the signals, calculating average ADC values and average variance values of the signals, and transitioning from the monitoring state to a trigger transition state.

23. The method of claim 20 wherein the step of processing measurements of the signals with the processor further comprises, during a trigger transition state, depending on the digital measurements of the signals, on a state status, and on comparisons of the measurements of the signals to thresholds, transitioning from the trigger transition state to either a monitoring state, a droplet detection state, or a droplet count state.

24. The method of claim 20 wherein the step of processing measurements of the signals with the processor further comprises, during a droplet detection state, depending on the digital measurements of the signals, and on comparisons of the measurements and average ADC values of the signals to thresholds, either transitioning from the droplet detection state to a monitoring state or entering a droplet count state.

25. The method of claim 20 wherein the step of processing measurements of the signals with the processor further comprises, during a droplet count state, depending on the digital measurements of the signals, and on comparisons of the measurements, average ADC values of the signals, and average variances of the signals to thresholds, either transitioning from the droplet count state to a monitoring state, entering a droplet detection state, or signaling an alarm.

26. The method of claim 18 wherein the step of processing measurements of the signals with the processor further comprises determining that the stuck fluid droplet is present if the determined pattern relationship starts at or below a known liquid signal ADC value at a first period in time, increases from at or below the known liquid signal ADC value over a primary threshold ADC value at a second period in time, decreases from over the primary threshold ADC value to between the primary threshold ADC value and a secondary threshold ADC value at a third period in time, and passes a stability test while being disposed at a fourth period in time between the primary threshold ADC value and the secondary threshold ADC value.

27. The method of claim 26 wherein the step of processing measurements of the signals with the processor further comprises determining during the stability test that the stuck fluid droplet is not present if the signals during the fourth period in time vary more than a set amount, drop down to or below a previously observed fluid level, drop below the secondary threshold ADC value, or increase over the primary threshold ADC value.

28. The method of claim 18 wherein the step of processing measurements of the signals with the processor further comprises determining during a triggering state, representing the possible beginning of the stuck fluid droplet, whether an ADC value of one of the digital measurements exceeds a primary threshold ADC value, or is less than a trigger threshold ADC value but greater than a minimum ADC value, and if so analyzes the digital measurements of the signals further to determine if the stuck fluid droplet is present, and if not determines that a stuck fluid droplet is not present.

29. The method of claim 18 wherein the step of processing measurements of the signals with the processor further comprises determining ADC mean values for the signals to determine the pattern relationship.

30. The method of claim 18 wherein the step of processing measurements of the signals with the processor further comprises conducting a test for false positives, indicating that the stuck fluid droplet is not present, by calculating and comparing at different times at least one of ADC mean values, variance mean values, or F critical calculations for the digital measurements.

31. The method of claim 17 wherein the step of processing measurements of the signals with the processor further comprises dynamically calibrating the at least one sensor during the pumping of the fluid through the fluid delivery line based on real-time measurements of the signals.

32. The method of claim 31 wherein the step of processing measurements of the signals with the processor further comprises updating a dynamic range of the real-time measurements of the signals with the dynamic range ranging between a minimum recorded measurement of the signals and a maximum recorded measurement of the signals.

33. A method for detecting air in a fluid delivery line of an infusion system comprising the steps of:

pumping fluid through the fluid delivery line over at least one sensor;

emitting and receiving signals from the at least one sensor into and from the fluid delivery line; and

processing measurements of the signals, using a processor, to determine whether the air is in the fluid delivery line by determining a pattern relationship over time of the signals while the signals are below an air/liquid threshold ADC value boundary.

34. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises comparing the determined pattern relationship over time to a programmed pattern over time algorithm which is indicative of the presence of the air being present in the fluid delivery line.

35. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises analyzing the plurality of measurements over multiple differently programmed states, programmed in the processor, to determine the pattern relationship.

36. The method of claim 35 wherein the multiple differently programmed states programmed in the processor comprise a monitoring state, a trigger transition state, a detection state, and a count state.

37. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises, during a monitoring state, taking the measurements of the signals, calculating average ADC values and average variance values of the signals, and transitioning from the monitoring state to a trigger transition state.

38. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises, during a trigger transition state, depending on the measurements of the signals, on a state status, and on comparisons of the measurements of the signals to thresholds, transitioning from the trigger transition state to either a monitoring state, a detection state, or a count state.

39. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises, during a detection state, depending on the measurements of the signals, and on comparisons of the measurements and average ADC values of the signals to thresholds, either transitioning from the detection state to a monitoring state or entering a count state.

40. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises, during a count state, depending on the measurements of the signals, and on comparisons of the measurements, average ADC values of the signals, and average variances of the signals to thresholds, either transitioning from the count state to a monitoring state, entering a detection state, or signaling an alarm.

41. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises determining that the air is present if the determined pattern relationship starts at or below a known liquid signal ADC value at a first period in time, increases from at or below the known liquid signal ADC value over the air/liquid threshold ADC value boundary at a second period in time, decreases from over the air/liquid threshold ADC value boundary to between the air/liquid threshold ADC value boundary and a secondary threshold ADC value at a third period in time, and passes a stability test while being disposed at a fourth period in time between the air/liquid threshold ADC value boundary and the secondary threshold ADC value.

42. The method of claim 41 wherein the step of processing measurements of the signals with the processor further comprises determining during the stability test that the air is not present if the signals during the fourth period in time vary more than a set amount, drop down to or below a previously observed fluid level, drop below the secondary threshold ADC value, or increase over the air/liquid threshold ADC value boundary.

43. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises determining during a triggering state, representing the possible beginning of the air, whether an ADC value of one of the measurements exceeds the air/liquid threshold ADC boundary, or is less than a trigger threshold ADC value but greater than a minimum ADC value, and if so analyzes the measurements of the signals further to determine if the air is present, and if not determines that air is not present.

44. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises determining ADC mean values for the signals to determine the pattern relationship.

45. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises conducting a test for false positives, indicating that the air is not present, by calculating and comparing at different times at least one of ADC mean values, variance mean values, or F critical calculations for the measurements.

46. The method of claim 33 wherein the step of processing measurements of the signals with the processor further comprises dynamically calibrating the at least one sensor during the pumping of the fluid through the fluid delivery line based on real-time measurements of the signals.

47. The method of claim 46 wherein the step of processing measurements of the signals with the processor further comprises updating a dynamic range of the real-time measurements of the signals with the dynamic range ranging between a minimum recorded measurement of the signals and a maximum recorded measurement of the signals.