Systems and methods for medical acquisition processing and machine learning for anatomical assessment
Systems and methods are disclosed for determining anatomy directly from raw medical acquisitions using a machine learning system. One method includes obtaining raw medical acquisition data from transmission and collection of energy and particles traveling through and originating from bodies of one or more individuals; obtaining a parameterized model associated with anatomy of each of the one or more individuals; determining one or more parameters for the parameterized model, wherein the parameters are associated with the raw medical acquisition data; training a machine learning system to predict one or more values for each of the determined parameters of the parameterized model, based on the raw medical acquisition data; acquiring a medical acquisition for a selected patient; and using the trained machine learning system to determine a parameter value for a patient-specific parameterized model of the patient.
1. A method for determining anatomy from raw medical acquisition data, the method comprising:
obtaining raw medical acquisition data for a patient, the raw medical acquisition data including data resulting from transmission and collection of energy and particles traveling through and originating from the patient's body; and
determining, via a machine learning system, at least one parameter value for a patient-specific parameterized model of the patient;
wherein the machine learning system is trained based on (i) raw medical acquisition data resulting from transmission and collection of energy and particles traveling through and originating from bodies of one or more individuals, as training data, and (ii) one or more parameter values of a respective parameterized model associated with anatomy of each of the one or more individuals, as ground truth, such that the machine learning system is configured to output at least one parameter value of a parameterized model of a given individual in response to input raw medical acquisition data for the given individual.
2. The method of claim 1 , wherein:
training the machine learning system causes the machine learning system to determine an association between raw medical acquisition data for the given individual and at least one parameter for the parameterized model of the given individual; and
the machine learning system is configured to output the at least one parameter value of the parameterized model of the given individual based on the input raw medical acquisition data and the determined association.
3. The method of claim 2 , wherein determining the association includes:
obtaining an image reconstruction function that produces an image output based on the raw medical acquisition data for the given individual; and
adjusting the image reconstruction function such that applying a predetermined image analysis function to the image output of results of the image reconstruction function in the one or more parameter values of the at least one parameter for the parameterized model of the given individual.
4. The method of claim 2 , wherein the association between the raw medical acquisition data for the given individual and the at least one parameter for the parameterized model of the given individual has no image intermediary.
5. The method of claim 1 , wherein:
the training data for the machine learning system further includes (iii) acquisition parameters associated with acquisition of the raw medical acquisition data for each of the one or more individuals; and
the output of the at least one parameter value by the machine learning system is further based on input acquisition parameters associated with the acquisition of the raw medical acquisition data for the patient.
6. The method of claim 1 , wherein,
the machine learning system is one of a plurality of machine learning systems;
each of the plurality of machine learning systems is associated with a respective data acquisition modality; and
the method further comprises:
determining a data acquisition modality of the raw medical acquisition data for the patient; and
selecting the machine learning system from amongst the plurality of machine learning systems based on the determined data acquisition modality.
7. The method of claim 1 , wherein:
the determined at least one parameter value includes an anatomical feature measurement including a vessel centerline, an anatomical point, an anatomical surface point or boundary, or a combination thereof, and
the parameterized model of the patient includes a vascular model, a disease model, a perfusion model, a motion model, a blood flow model, or a combination thereof.
8. The method of claim 1 , further comprising:
generating an image reconstruction based on the raw medical acquisition data for the patient; and
validating the image reconstruction based on the determined at least one parameter value.
9. A system for determining anatomy from raw medical acquisition data, comprising:
a memory storing instructions and a machine learning system, wherein the machine learning system is trained based on (i) raw medical acquisition data resulting from transmission and collection of energy and particles traveling through and originating from bodies of one or more individuals, as training data, and (ii) one or more parameter values of a respective parameterized model associated with anatomy of each of the one or more individuals, as ground truth, such that the machine learning system is configured to output at least one parameter value of a parameterized model of a given individual in response to input raw medical acquisition data for the given individual; and
a processor, operatively connected to the memory and configured to execute the instructions to perform acts, including:
obtaining raw medical acquisition data for a patient, the raw medical acquisition data including data resulting from transmission and collection of energy and particles traveling through and originating from the patient's body; and
determining, via the machine learning system, at least one parameter value for a patient-specific parameterized model of the patient.
10. The system of claim 9 , wherein:
training the machine learning system causes the machine learning system to determine an association between raw medical acquisition data for the given individual and at least one parameter for the parameterized model of the given individual; and
the machine learning system is configured to output the at least one parameter value of the parameterized model of the given individual based on the input raw medical acquisition data and the determined association.
11. The system of claim 10 , wherein determining the association includes:
obtaining an image reconstruction function that produces an image output based on the raw medical acquisition data for the given individual; and
adjusting the image reconstruction function such that applying a predetermined image analysis function to the image output of results of the image reconstruction function in the one or more parameter values of the at least one parameter for the parameterized model of the given individual.
12. The system of claim 10 , wherein the association between the raw medical acquisition data for the given individual and the at least one parameter for the parameterized model of the given individual has no image intermediary.
13. The system of claim 9 , wherein:
the training data for the machine learning system further includes (iii) acquisition parameters associated with acquisition of the raw medical acquisition data for each of the one or more individuals; and
the output of the at least one parameter value by the machine learning system is further based on input acquisition parameters associated with the acquisition of the raw medical acquisition data for the patient.
14. The system of claim 9 , wherein,
the machine learning system is one of a plurality of machine learning systems stored in the memory;
each of the plurality of machine learning systems is associated with a respective data acquisition modality; and
the acts further include:
determining a data acquisition modality of the raw medical acquisition data for the patient; and
selecting the machine learning system from amongst the plurality of machine learning systems based on the determined data acquisition modality.
15. The system of claim 9 , wherein:
the determined at least one parameter value includes an anatomical feature measurement including a vessel centerline, an anatomical point, an anatomical surface point or boundary, or a combination thereof, and
the parameterized model of the patient includes a vascular model, a disease model, a perfusion model, a motion model, a blood flow model, or a combination thereof.
16. The system of claim 9 , wherein the acts further include:
generating an image reconstruction based on the raw medical acquisition data for the patient; and
validating the image reconstruction based on the determined at least one parameter value.
17. A method for determining anatomy from raw medical acquisition data, the method comprising:
obtaining raw medical acquisition data for a patient, the raw medical acquisition data including data resulting from transmission and collection of energy and particles traveling through and originating from the patient's body;
determining, via a machine learning system, at least one parameter value for a patient-specific parameterized model of the patient, wherein:
the determined at least one parameter value includes an anatomical feature measurement including a vessel centerline, an anatomical point, an anatomical surface point or boundary, or a combination thereof,
the parameterized model of the patient includes a vascular model, a disease model, a perfusion model, a motion model, a blood flow model, or a combination thereof; and
the machine learning system is trained based on (i) training data that includes raw medical acquisition data resulting from transmission and collection of energy and particles traveling through and originating from bodies of one or more individuals, and acquisition parameters associated with acquisition of the raw medical acquisition data for each of the one or more individuals, and (ii) one or more parameter values of a respective parameterized model associated with anatomy of each of the one or more individuals as ground truth, such that the machine learning system is configured to output at least one parameter value of a parameterized model of a given individual in response to input raw medical acquisition data for the given individual and input acquisition parameters associated with the acquisition of the raw medical acquisition data for the patient.
18. The method of claim 17 , wherein:
training the machine learning system causes the machine learning system to determine an association between raw medical acquisition data for the given individual and at least one parameter for the parameterized model of the given individual; and
the machine learning system is configured to output the at least one parameter value of the parameterized model of the given individual based on the input raw medical acquisition data and the determined association.
19. The method of claim 18 , wherein determining the association includes:
obtaining an image reconstruction function that produces an image output based on the raw medical acquisition data for the given individual; and
adjusting the image reconstruction function such that applying a predetermined image analysis function to the image output of results of the image reconstruction function in the one or more parameter values of the at least one parameter for the parameterized model of the given individual.
20. The method of claim 18 , wherein the association between the raw medical acquisition data for the given individual and the at least one parameter for the parameterized model of the given individual has no image intermediary.