Methods, systems, and computer readable media for smart image protocoling
A method for smart image protocoling includes, using a medical imaging device, obtaining, using a first medical imaging sequence, a first set of medical images of a patient. Anatomical and, if present, disease features are extracted from the first set of medical images. A machine learning trained algorithm is used to determine, in real time, and based on the extracted anatomical and/or disease features, whether a desired medical imaging goal is achieved for the patient. In response to determining that the desired medical imaging goal is achieved, at least one image from the first set of medical images is output as a final image. In response to determining that the desired medical imaging goal has not been achieved, the machine learning trained algorithm is used to select a second medical imaging sequence. A second set of medical images of the patient is obtained using the second medical imaging sequence. The above outlined procedures will be repeated until the final imaging goal is achieved for a patient.
1. A method for smart image protocoling comprising:
extracting anatomical or disease features from the first set of medical images;
using a machine learning trained algorithm to determine, based on the extracted anatomical or disease features, that a desired medical imaging goal is not achieved from the first medical imaging sequence for the patient, wherein the machine learning trained algorithm is trained using a repository of medical image sequences and corresponding clinic interpretations of the medical image sequences to select, given features extracted from the medical images, a sequence of medical images that is optimal for patient-specific diagnosis of a disease; and
in response to determining that the desired medical imaging goal is not achieved from the first medical imaging sequence, selecting, using the machine learning trained algorithm a second medical imaging sequence and obtaining a second set of medical images of the patient using the second medical imaging sequence, wherein the first and second medical imaging sequences are clinical diagnostic medical imaging sequences and wherein the second medical imaging sequence is performed using at least one different medical imaging parameter from the first medical imaging sequence.
2. The method of claim 1 using the machine learning trained algorithm comprises applying the extracted features to a decision tree.
3. The method of claim 1 comprising extracting features from the second set of medical images, using the machine learning trained algorithm to determine, in real time, and based on the extracted features from the first set of medical images and the second set of medical images, whether the desired medical imaging goal is achieved for the patient, in response to determining that the desired medical imaging goal is achieved, outputting at least one image from the first and second sets of medical images, and, in response to determining that the desired medical imaging goal has not been achieved, selecting, using the machine learning trained algorithm, a third medical imaging sequence, and obtaining, in real time, a third set of medical images of the patient using the third medical imaging sequence.
4. The method of claim 1 wherein the medical imaging device comprises a magnetic resonance imaging device, the first medical imaging sequence comprises an initial magnetic resonance imaging sequence selected based on a patient's medical history or an imaging sequence for an organ of interest and wherein the second medical imaging sequence comprises a magnetic resonance imaging scan sequence with at least one of an orientation, slice thickness, resolution, contrast, and spatial coverage selected based on features extracted from the initial magnetic resonance imaging sequence.
5. The method of claim 1 comprising recursively obtaining medical images of the patient, extracting features from the medical images, and updating a medical imaging sequence for obtaining the medical images until the desired imaging goal is achieved.
6. The method of claim 1 wherein selecting the second medical imaging sequence includes selecting an imaging modality for the second medical imaging sequence.
7. The method of claim 6 wherein the modality for the second medical imaging sequence is different from an imaging modality of the first medical imaging sequence.
8. The method of claim 1 wherein the extracting, determining, and obtaining occur in real time.
9. A system for smart image protocoling, the system comprising:
an imaging sequence controller for controlling an imaging sequence implemented by a medical imaging device, the imaging sequence controller including:
at least one processor;
a memory;
a feature extractor for extracting anatomical or disease features from a first set of medical images of a patient obtained by the medical imaging device using a first medical imaging sequence; and
an imaging sequence selector for:
using a machine learning trained algorithm to determine based on the extracted anatomical and/or disease features, that a desired medical imaging goal is not achieved from the first medical imaging sequence for the patient, wherein the machine learning trained algorithm is trained using a repository of medical image sequences and corresponding clinical interpretations of the medical image sequences to select, given features extracted from medical images, a sequence of medical images that is optimal for patient-specific diagnosis of a disease; and
in response to determining that the desired medical imaging goal is not achieved from the first medical imaging sequence, selecting, using the machine learning trained algorithm a second medical imaging sequence and obtaining a second set of medical images of the patient using the second medical imaging sequence, wherein the first and second medical imaging sequences are clinical diagnostic medical imaging sequences and wherein the second medical imaging sequence is performed using at least one different medical imaging parameter from the first medical imaging sequence.
10. The system of claim 9 wherein the medical imaging device comprises a magnetic resonance imaging device, a computed tomography imaging device, or a positron emission tomography imaging device.
11. The system of claim 10 wherein the machine learning trained algorithm is implemented as a decision tree.
12. The system of claim 10 wherein the feature extractor is configured to extract features from the first and second sets of medical images and wherein the imaging sequence selector is configured to use the machine learning trained algorithm to determine, in real time, and based on the extracted features from the first and second sets of medical images, determine whether the desired medical imaging goal is achieved for the patient, in response to determining that the desired medical imaging goal is achieved, output at least one image from the first and second sets of medical images, in response to determining that the desired medical imaging goal has not been achieved, select, using the machine learning trained algorithm a third medical imaging sequence, and obtain, in real time, a third set of medical images of the patient using the third medical imaging sequence.
13. The system of claim 10 wherein the medical imaging device comprises a magnetic resonance imaging device, the first medical imaging sequence comprises an initial magnetic resonance imaging sequence selected based on a patient's medical history or an imaging sequence for an organ of interest and wherein the second medical imaging sequence comprises a magnetic resonance imaging scan sequence with at least one of an orientation, slice thickness, resolution, contrast, physiological information, and spatial coverage selected based on features extracted from the initial magnetic resonance imaging sequence.
14. The system of claim 10 wherein the imaging controller is configured to recursively obtain medical images of the patient, extract features from the medical images, and update a medical imaging sequence for obtaining the medical images until the desired imaging goal is achieved.
15. The system of claim 10 wherein selecting the second medical imaging sequence includes selecting an imaging modality for the second medical imaging sequence.
16. The system of claim 15 wherein the modality for the second medical imaging sequence is different from an imaging modality of the first medical imaging sequence.
17. The system of claim 10 wherein the extracting, determining, and obtaining occur in real time.
18. A non-transitory computer readable medium having stored thereon executable instructions that when executed by the processor of a computer control the computer to perform steps comprising:
using a medical imaging device, obtaining, using a first medical imaging sequence, a first set of medical images of a patient;
extracting anatomical and, if present, disease features from the first set of medical images;
using a machine learning trained algorithm to determine, based on the extracted anatomical and/or disease features, that a desired medical imaging goal is not achieved from the first medical imaging sequence for the patient, wherein the machine learning trained algorithm is trained using a repository of medical image sequences and corresponding clinical interpretations of the medical image sequences to select, given features extracted from medical images, a sequence of medical images that is optimal for patient-specific diagnosis of a disease; and
in response to determining that the desired medical imaging goal is not achieved from the first medical imaging sequence, selecting, using the machine learning trained algorithm, a second medical imaging sequence and obtaining, a second set of medical images of the patient using the second medical imaging sequence, wherein the first and second medical imaging sequences are clinical diagnostic medical imaging sequences and wherein the second medical imaging sequence is performed using at least one different medical imaging parameter from the first medical imaging sequence.