Dynamic adjustment of algorithms for separation and collection of blood components

Patent metadata

Field	Value
Publication number	US-12453810-B2
Application number	US-202519053529-A
Country	US
Kind code	B2
Filing date	Feb 14, 2025
Priority date	Sep 16, 2019
Publication date	Oct 28, 2025
Grant date	Oct 28, 2025

Abstract

Blood is conveyed from a source into a separator, which separates at least one target blood component from the blood. The target blood component is then conveyed out of the separator, with the procedure continuing until an initial target amount of blood to be processed has been conveyed from the source into the separator and the target blood component separated from the initial target amount of blood to be processed has been conveyed out of the separator as an actual yield of the target blood component. An adjusted target amount of blood to be processed is then determined based at least in part on the difference between a target yield of the target blood component and the actual yield. The initial target amount of blood to be processed is then replaced with the adjusted target amount of blood to be processed when next executing the procedure.

First claim

The invention claimed is:

1. A blood separation device, comprising:
a separator configured to receive a blood separation chamber;
a pump system; and
a controller programmed with a blood separation algorithm for executing a blood separation procedure in which an initial target amount of blood from a blood source is pumped into the separator by the pump system and separated so as to collect a target yield of a target blood component, wherein the controller is configured to
access or receive results for a plurality of blood separation procedures executed by the blood separation device and/or by one or more other blood separation devices, with each of said results including an actual yield of said target blood component and/or a difference between a blood characteristic before and after execution of said blood separation procedure,
determine a correlation between said target and actual yields and/or said differences between the blood characteristic before and after execution of said blood separation procedure for a subset of said results so as to determine an adjusted amount of blood to be separated when subsequently executing said blood separation procedure and/or a scaling factor to be applied to said blood separation algorithm so as to separate the adjusted amount of blood when subsequently executing said blood separation procedure; and
modify said blood separation algorithm so as to separate the adjusted amount of blood when subsequently executing said blood separation procedure, with the controller being configured to employ machine learning techniques so as to minimize a mean squared error when correlating said target and actual yields and/or said differences between the blood characteristic before and after execution of said blood separation procedure.

2. The blood separation device of claim 1 , wherein
said results are for a plurality of blood separation procedures performed on a blood source, and
the controller is configured to modify said blood separation algorithm so as to separate the adjusted amount of blood when subsequently executing said blood separation procedure for the blood source.

3. The blood separation device of claim 1 , wherein
said results are for blood separation procedures performed on a plurality of different blood sources,
the controller is further configured to access or receive data regarding a plurality of characteristics of each of said plurality of different blood sources, and
the determination of which of said results to include in said subset is based at least in part on one of said characteristics.

4. The blood separation device of claim 1 , wherein the controller is configured to calculate an average or mean when determining said correlation.

5. The blood separation device of claim 1 , wherein the controller is configured to calculate a moving average when determining said correlation.

6. The blood separation device of claim 5 , wherein the controller is configured to employ machine learning techniques to determine the number of said results to include in said subset so as to minimize a mean squared error when calculating said moving average.

7. The blood separation device of claim 1 , wherein the controller is configured to calculate a double moving average when determining said correlation.

8. The blood separation device of claim 7 , wherein the controller is configured to employ machine learning techniques to determine the number of said results to include in said subset so as to minimize a mean squared error when calculating said double moving average.

9. The blood separation device of claim 1 , wherein the controller is configured to employ an exponential smoothing technique, a double exponential smoothing technique, or a triple exponential smoothing technique when determining said correlation.

10. The blood separation device of claim 9 , wherein the controller is configured to employ machine learning techniques to determine the value to use for the speed at which older results are dampened or smoothed so as to minimize a mean squared error when employing said exponential smoothing technique, said double exponential smoothing technique, or said triple exponential smoothing technique.

11. The blood separation device of claim 9 , wherein the controller is configured to employ machine learning techniques to determine a plurality of values to use so as to minimize a mean squared error when employing said double exponential smoothing technique or said triple exponential smoothing technique.

12. A data processing system configured to communicate with a plurality of blood separation device controllers each programmed with a blood separation algorithm for executing a blood separation procedure in which an initial target amount of blood from a blood source is separated so as to collect a target yield of a target blood component, wherein the data processing system is configured to
access or receive results for a plurality of blood separation procedures executed by at least one of said plurality of blood separation device controllers, with each of said results including an actual yield of said target blood component and/or a difference between a blood characteristic before and after execution of said blood separation procedure;
determine a correlation between said target and actual yields and/or said differences between the blood characteristic before and after execution of said blood separation procedure for a subset of said results so as to determine an adjusted amount of blood to be separated when subsequently executing said blood separation procedure and/or a scaling factor to be applied to said blood separation algorithm so as to separate the adjusted amount of blood when subsequently executing said blood separation procedure; and
transmit a signal based on said correlation to at least one of said blood separation device controllers, with the signal being configured to modify said blood separation algorithm so as to separate the adjusted amount of blood when subsequently executing said blood separation procedure, with the data processing system being configured to employ machine learning techniques so as to minimize a mean squared error when correlating said target and actual yields and/or said differences between the blood characteristic before and after execution of said blood separation procedure.

13. The data processing system of claim 12 , wherein
said results are for a plurality of blood separation procedures performed on a blood source, and
the signal is configured to modify said blood separation algorithm so as to separate the adjusted amount of blood when subsequently executing said blood separation procedure for the blood source.

14. The data processing system of claim 12 , wherein
said results are for blood separation procedures performed on a plurality of different blood sources,
the data processing system is further configured to access or receive data regarding a plurality of characteristics of each of said plurality of different blood sources, and
the determination of which of said results to include in said subset is based at least in part on one of said characteristics.

15. The data processing system of claim 12 , configured to calculate an average or mean when determining said correlation.

16. The data processing system of claim 12 , configured to calculate a moving average when determining said correlation.

17. The data processing system of claim 16 , configured to employ machine learning techniques to determine the number of said results to include in said subset so as to minimize a mean squared error when ca

Assignees

• Fenwal Inc

Inventors

• Case Brian C

Classifications

• B01D63/16

  Rotary, reciprocated or vibrated modules · CPC title

• B01D21/30

  Control equipment · CPC title

• B01D21/262

  by using a centrifuge · CPC title

• A61M2205/52

  with memories providing a history of measured variating parameters of apparatus or patient · CPC title

• A61M1/3696 Primary

  with means for adding or withdrawing liquid substances during the centrifugation, e.g. continuous centrifugation · CPC title