Automatically determining a medical recommendation for a patient based on multiple medical images from multiple different medical imaging modalities
Automatically determining a medical recommendation for a patient based on multiple medical images from multiple different medical imaging modalities. In some embodiments, a method may include receiving a first and second medical images of a patient from first and second medical imaging modalities, mapping a first region of interest (ROI) on the first medical image to a second ROI on the second medical image, generating first annotation data related to the first ROI and second annotation data related to the second ROI, generating first medical clinical data related to the first ROI and second medical clinical data related to the second ROI, inputting, into a machine learning classifier, the first and second annotation data and the first and second medical clinical data, and automatically determining, by the machine learning classifier, a medical recommendation for the patient related to a medical condition of the patient.
1. A computer-implemented method for automatically determining a medical recommendation for a patient based on multiple medical images from multiple different medical imaging modalities, the computer-implemented method comprising:
receiving a first medical image of a patient from a first medical imaging modality and a second medical image of the patient from a second medical imaging modality, the first medical imaging modality being different from the second medical imaging modality;
mapping a first region of interest (ROI) on the first medical image to a second ROI on the second medical image, the first ROI and the second ROI both related to a medical condition of the patient;
generating first annotation data related to the first ROI and second annotation data related to the second ROI;
generating first medical clinical data related to the first ROI and second medical clinical data related to the second ROI;
training a machine learning classifier with annotation data and medical clinical data of medical images of one or more other patients to generate a predicted medical recommendation;
inputting, into the machine learning classifier, the first and second annotation data and the first and second medical clinical data;
in response to the inputting, automatically determining, by the machine learning classifier, a medical recommendation for the patient related to the medical condition of the patient;
receiving a third medical image of the patient from the first medical imaging modality and a fourth medical image of the patient from the second medical imaging modality;
mapping a third ROI on the third medical image to a fourth ROI on the fourth medical image, the third ROI and the fourth ROI both related to the medical condition of the patient;
generating third annotation data related to the third ROI and fourth annotation data related to the fourth ROI;
generating third medical clinical data related to the third ROI and fourth medical clinical data related to the fourth ROI;
entering, into the machine learning classifier, the third and fourth annotation data and the third and fourth medical clinical data; and
in response to the entering, automatically determining, by the machine learning classifier, an updated medical recommendation for the patient related to the medical condition of the patient.
2. The computer-implemented method of claim 1 , wherein the entering further comprises reinputting, into the machine learning classifier, the first and second annotation data and the first and second medical clinical data.
3. The computer-implemented method of claim 1 , wherein the medical recommendation includes one or more of a diagnosis recommendation for the medical condition, a treatment recommendation for the medical condition, and a research recommendation for the medical condition.
4. The computer-implemented method of claim 1 , wherein:
the first medical imaging modality comprises a computation tomography (CT) imaging modality, a magnetic resonance imaging (MRI) imaging modality, a histology pathology imaging modality, or a next generation sequenced (NGS) tissue/tumor microarray pathology imaging modality; and
the second medical imaging modality comprises a tissue microarray (TMA) imaging modality, a Fluorescence In Situ Hybridization (FISH) imaging modality, a flow cytometry pathology imaging modality, a gene or chromosomal microarray pathology imaging modality, or a myocardial perfusion radiology imaging modality.
5. The computer-implemented method of claim 1 , wherein the first and second medical clinical data comprise measurement data, blood flow data, opaqueness data, abnormality data, cellular structure data, or morphology data.
6. One or more non-transitory computer-readable media comprising one or more computer readable instructions that, when executed by one or more processors of a computing device, cause the computing device to perform a method for automatically determining a medical recommendation for a patient based on multiple medical images from multiple different medical imaging modalities, the method comprising:
receiving a first medical image of a patient from a first medical imaging modality and a second medical image of the patient from a second medical imaging modality, the first medical imaging modality being different from the second medical imaging modality;
mapping a first region of interest (ROI) on the first medical image to a second ROI on the second medical image, the first ROI and the second ROI both related to a medical condition of the patient;
generating first annotation data related to the first ROI and second annotation data related to the second ROI;
generating first medical clinical data related to the first ROI and second medical clinical data related to the second ROI;
training a machine learning classifier with annotation data and medical clinical data of medical images of one or more other patients to generate a predicted medical recommendation;
inputting, into the machine learning classifier, the first and second annotation data and the first and second medical clinical data;
in response to the inputting, automatically determining, by the machine learning classifier, a medical recommendation for the patient related to the medical condition of the patient;
receiving a third medical image of the patient from the first medical imaging modality and a fourth medical image of the patient from the second medical imaging modality;
mapping a third ROI on the third medical image to a fourth ROI on the fourth medical image, the third ROI and the fourth ROI both related to the medical condition of the patient;
generating third annotation data related to the third ROI and fourth annotation data related to the fourth ROI;
generating third medical clinical data related to the third ROI and fourth medical clinical data related to the fourth ROI;
entering, into the machine learning classifier, the third and fourth annotation data and the third and fourth medical clinical data; and
in response to the entering, automatically determining, by the machine learning classifier, an updated medical recommendation for the patient related to the medical condition of the patient.
7. The one or more non-transitory computer-readable media of claim 6 , wherein the entering further comprises reinputting, into the machine learning classifier, the first and second annotation data and the first and second medical clinical data.
8. The one or more non-transitory computer-readable media of claim 6 , wherein the medical recommendation includes one or more of a diagnosis recommendation for the medical condition, a treatment recommendation for the medical condition, and a research recommendation for the medical condition.
9. The one or more non-transitory computer-readable media of claim 6 , wherein:
the first medical imaging modality comprises a computation tomography (CT) imaging modality, a magnetic resonance imaging (MRI) imaging modality, a histology pathology imaging modality, or a next generation sequenced (NGS) tissue/tumor microarray pathology imaging modality; and
the second medical imaging modality comprises a tissue microarray (TMA) imaging modality, a Fluorescence In Situ Hybridization (FISH) imaging modality, a flow cytometry pathology imaging modality, a gene or chromosomal microarray pathology imaging modality, or a myocardial perfusion radiology imaging modality.
10. The one or more non-transitory computer-readable media of claim 6 , wherein the first and second medical clinical data comprise measurement data, blood flow data, opaqueness data, abnormality data, cellular structure data, or morphology data.
11. A computing device comprising:
one or more processors; and
one or more non-transitory computer-readable media comprising one or more computer readable instructions that, when executed by the one or more processors, cause the computing device to perform a method for automatically determining a medical recommendation for a patient based on multiple medical images from multiple different medical imaging modalities, the method comprising:
receiving a first medical image of a patient from a first medical imaging modality and a second medical image of the patient from a second medical imaging modality, the first medical imaging modality being different from the second medical imaging modality;
mapping a first region of interest (ROI) on the first medical image to a second ROI on the second medical image, the first ROI and the second ROI both related to a medical condition of the patient;
generating first annotation data related to the first ROI and second annotation data related to the second ROI;
generating first medical clinical data related to the first ROI and second medical clinical data related to the second ROI;
training a machine learning classifier with annotation data and medical clinical data of medical images of one or more other patients to generate a predicted medical recommendation;
inputting, into the machine learning classifier, the first and second annotation data and the first and second medical clinical data;
in response to the inputting, automatically determining, by the machine learning classifier, a medical recommendation for the patient related to the medical condition of the patient;
receiving a third medical image of the patient from the first medical imaging modality and a fourth medical image of the patient from the second medical imaging modality;
mapping a third ROI on the third medical image to a fourth ROI on the fourth medical image, the third ROI and the fourth ROI both related to the medical condition of the patient;
generating third annotation data related to the third ROI and fourth annotation data related to the fourth ROI;
generating third medical clinical data related to the third ROI and fourth medical clinical data related to the fourth ROI;
entering, into the machine learning classifier, the third and fourth annotation data and the third and fourth medical clinical data; and
in response to the entering, automatically determining, by the machine learning classifier, an updated medical recommendation for the patient related to the medical condition of the patient.
12. The computing device of claim 11 , wherein the medical recommendation includes one or more of a diagnosis recommendation for the medical condition, a treatment recommendation for the medical condition, and a research recommendation for the medical condition.
13. The computing device of claim 11 , wherein:
the first medical imaging modality comprises a computation tomography (CT) imaging modality, a magnetic resonance imaging (MRI) imaging modality, a histology pathology imaging modality, or a next generation sequenced (NGS) tissue/tumor microarray pathology imaging modality; and
the second medical imaging modality comprises a tissue microarray (TMA) imaging modality, a Fluorescence In Situ Hybridization (FISH) imaging modality, a flow cytometry pathology imaging modality, a gene or chromosomal microarray pathology imaging modality, or a myocardial perfusion radiology imaging modality.
14. The computing device of claim 11 , wherein the first and second medical clinical data comprise measurement data, blood flow data, opaqueness data, abnormality data, cellular structure data, or morphology data.