Determining an indicator relating to injury

1. A method for determining an indicator relating to an injury of an anatomical structure of a patient, wherein the method comprises executing, on at least one processor of at least one computer, the steps of:

acquiring acceleration data describing an energy of a set of one or more signals in dependence on both time and frequency, the set of signals acquired by measuring acceleration of the anatomical structure over time;

acquiring analysis data describing an analysis rule for determining at least two of:

an overall energy level of at least one signal of the set of signals,

a correlation between at least two signals of the set of signals in a frequency domain, the at least two signals respectively measured at at least two different respective regions of the anatomical structure, or

a relationship between energies given for at least two different frequency ranges of at least one signal of the set of signals;

determining indicator data describing the indicator based on the acceleration data and the analysis data;

causing the at least one processor to render the indicator data as output based on the acquired acceleration data and the analysis data;

wherein the indicator data is determined by a trained machine learning algorithm, and

wherein the trained machine learning algorithm is trained using a plurality of classified acceleration data acquired for a first population of patients having the injury and a second population of patients not having the injury.

2. The method according to claim 1 , wherein the analysis data further describe a rule for combining at least two of:

the determined overall energy level,

the determined correlation, or

the determined relationship, in order to determine the indicator data based on the combination.

3. The method according to claim 1 , wherein the trained machine learning algorithm uses the analysis data, a plurality of acceleration data or combinations thereof, the plurality of acceleration data being associated to information on a presence or non-presence of the injury, for performing an optimization of the analysis data.

4. The method according to claim 1 , comprising executing, on the at least one processor of the at least one computer, the step of:

determining, based on the acceleration data, average energy data describing an average energy depending on frequency;

wherein determining indicator data is based on the average energy data.

5. The method according to claim 1 , further comprising executing, on the at least one processor of the at least one computer, the steps of:

acquiring heart rate data describing a heart rate of the patient;

determining, based on the heart rate data and the acceleration data, heart rate acceleration data describing the energy of the signal in dependence on frequency for different heart rates, by sorting frequency contents of the signal according to the heart rate, the frequency contents describing the energy in dependence on frequency for different times;

wherein determining the indicator data is based on the heart rate acceleration data by applying the analysis rule to the heart rate acceleration data.

6. The method according to claim 1 , comprising executing, on the at least one processor of the at least one computer, the steps of:

acquiring threshold data describing a threshold for the energy;

wherein the overall energy level is determined by comparing the threshold with the energy.

7. The method according to claim 1 , wherein the correlation is determined by computing a correlation coefficient describing a correlation between a first signal of the set of signals acquired at a first region of the anatomical structure and a second signal of the set of signals acquired at a second region of the anatomical structure.

8. The method according to claim 1 , wherein the relationship is a ratio of energy being determined by comparing a sum or an integral of the energies within a first frequency range to a sum or an integral of energies in a second frequency range.

9. The method according to claim 1 , wherein the acceleration data is acquired by at least two sensors respectively arranged at different regions of the anatomical structure or respectively receiving signals from different regions of the anatomical structure.

10. The method according to claim 1 , wherein the signal is acquired by analyzing a time dependent acceleration signal with respect to time and frequency by a joint time frequency analysis (JFTA) by applying a Fourier transformation to the acceleration signal for a plurality of time intervals.

11. A non-transitory computer-readable storage medium comprising instructions, which, when running on at least one processor of at least one computer, causes the at least one processor to perform the steps of:

acquiring acceleration data describing an energy of a set of one or more signals in dependence on both time and frequency, the set of signals acquired by measuring acceleration of the anatomical structure over time;

acquiring analysis data describing an analysis rule for determining at least two of:

an overall energy level of at least one signal of the set of signals,

a correlation between at least two signals of the set of signals in the frequency domain, the at least two signals respectively measured at at least two different respective regions of the anatomical structure, or

a relationship between energies given for at least two different frequency ranges of at least one signal of the set of signals; and

determining indicator data describing the indicator based on the acceleration data and the analysis data,

wherein the indicator data is determined by a trained machine learning algorithm, and

wherein the trained machine learning algorithm is trained using a plurality of classified acceleration data acquired for a first population of patients having the injury and a second population of patients not having the injury.

12. A system configured to determine an indicator relating to an injury of an anatomical structure of a patient, the system comprising:

at least one sensor; and

at least one computer having at least one processor and associated memory, the memory storing instructions which when executed on the at least one processor causes the at least one processor to:

acquire acceleration data describing an energy of a set of one or more signals in dependence on both time and frequency, the set of signals acquired by measuring acceleration of the anatomical structure over time;

acquire analysis data describing an analysis rule for determining at least two of:

an overall energy level of at least one signal of the set of signals,

a correlation between at least two signals of the set of signals in a frequency domain, the at least two signals respectively measured at at least two different respective regions of the anatomical structure, or

a relationship between energies given for at least two different frequency ranges of at least one signal of the set of signals; and

determine indicator data describing the indicator based on the acceleration data and the analysis data,

wherein the indicator data is determined by a trained machine learning algorithm,

wherein the trained machine learning algorithm is trained using a plurality of classified acceleration data acquired for a first population of patients having the injury and a second population of patients not having the injury, and

wherein the at least one computer is operatively associated with the at least one sensor.