Back to results
US 12,575,793 B2Granted
Self-learning input filter for medical devices
AssigneeB. Braun Avitum AG
InventorPeter Mandry
Life of this patent
Estimating the term…
Abstract
A method for monitoring the reliability of an input in a medical device in which at least one value is entered into the medical device, an occurrence probability density of the at least one value is calculated on the basis of an input history, an error probability of the input is calculated from an a priori error probability density stored in a first database and loaded from the first database and the calculated occurrence probability density, and the error probability is output for further handling of the input.
Claims8
A method for monitoring the reliability of input values into a medical device, comprising the steps of:
- receiving an input value;
- storing the input value;
- storing historical input data as an input history;
- storing predetermined probability distribution data and an a priori error probability density;
- when an input history is available, calculating an occurrence probability density of the input value based on the input history, and when the input history is unavailable or an insufficient quantity of input history is available, using the predetermined probability distribution data to obtain the occurrence probability density;
- calculating, using a recursive Bayes filter, an error probability of the input value from the a priori error probability density and from the calculated occurrence probability density, wherein outputs of the recursive Bayes filter adapt to changing input values;
- using the calculated error probability to check correctness of the input value;
- providing the input value to the medical device in accordance with correctness of the input value;
- dynamically and independently adjusting an alarm threshold on the medical device based on the calculated error probability when checking correctness of the input value;
- outputting an error when the input value has an erroneous value;
- recalculating the a priori error probability density using the input value;
- and
- repeating the calculating the occurrence probability density, calculating the error probability of the input value, using the calculated error probability, providing the input value, adjusting the alarm threshold, outputting the error, and recalculating the a priori error probability density steps for a subsequent input value using the recalculated a priori error probability density.
10 more elements
A system for monitoring reliability of input values into a medical device, comprising:
- a first database that stores an a priori error probability density, receives the input values, and stores historical input data as an input history;
- a second database that stores predetermined probability distribution data;
- and
- a calculation device that calculates an occurrence probability density of an input value based on the input history when the input history is available and uses the predetermined probability distribution data when the input history is unavailable or an insufficient quantity of input history is available, the calculation device including a recursive Bayes filter that calculates an error probability of the input value from the a priori error probability density stored in the first database and loaded from the first database and from the calculated occurrence probability density to adapt outputs of the recursive Bayes filter to changing input values,
- wherein the calculation device further uses the calculated error probability to check correctness of the input value and provides the input value to the medical device in accordance with correctness of the input value,
- wherein the calculation device recalculates the a priori error probability density using the input value and stores the recalculated a priori error probability density for use by the recursive Bayes filter in determining the error probability of a subsequent input value,
- wherein the calculation device provides the input value to the medical device in accordance with correctness of the input value as determined from the calculated error probability,
- wherein an alarm threshold on the medical device is dynamically and independently adjusted by the calculation device based on the calculated error probability when checking correctness of the input value,
- wherein an error is outputted by the calculation device when the input value has an erroneous value, and
- wherein the calculation devices repeatedly calculates the occurrence probability density, calculates the error probability of the input value, uses the calculated error probability, provides the input value, adjusts the alarm threshold, outputs the error, and recalculates the a priori error probability density steps for subsequent input values using the recalculated a priori error probability density.
7 more elements
Dependencies are inferred from the claim text.