Machine learning for automatic detection of intracranial hemorrhages with uncertainty measures from medical images
Systems and method for performing a medical imaging analysis task for making a clinical decision are provided. One or more input medical images of a patient are received. A medical imaging analysis task is performed from the one or more input medical images using a machine learning based network. The machine learning based network generates a probability score associated with the medical imaging analysis task. An uncertainty measure associated with the probability score is determined. A clinical decision is made based on the probability score and the uncertainty measure.
1. A computer-implemented method comprising:
receiving one or more input medical images of a patient;
performing a medical imaging analysis task from the one or more input medical images using a machine learning based network, the machine learning based network generating a probability score associated with the medical imaging analysis task;
determining an uncertainty measure representing an error associated with the probability score by:
selecting a calibration function comprising a fixed point defined according to a user selected threshold at which probability scores are totally uncertain;
applying the calibration function to the probability score, and
calculating an entropy of the probability score as the uncertainty measure based on results of the applied calibration function; and
making a clinical decision based on the probability score and the uncertainty measure.
2. The computer-implemented method of claim 1 , wherein the medical imaging analysis task comprises at least one of detection, subtyping, or segmentation of an intracranial hemorrhage of the patient.
3. The computer-implemented method of claim 1 , wherein making a clinical decision based on the probability score and the uncertainty measure comprises:
stratifying the patient into one of a plurality of patient groups based on the probability score and the uncertainty measure.
4. The computer-implemented method of claim 3 , wherein the medical imaging analysis task comprises detection of an intracranial hemorrhage of the patient, and the plurality of patient groups comprises a high confidence positive detection patient group, a high confidence negative detection patient group, and a low confidence patient group.
5. The computer-implemented method of claim 1 , wherein making a clinical decision based on the probability score and the uncertainty measure comprises:
determining whether to treat the patient based on the probability score and the uncertainty measure.
6. The computer-implemented method of claim 1 , wherein making a clinical decision based on the probability score and the uncertainty measure comprises:
determining whether to perform a clinical test on the patient based on the probability score and the uncertainty measure.
7. The computer-implemented method of claim 1 , wherein making a clinical decision based on the probability score and the uncertainty measure comprises:
prioritizing a worklist of a radiologist based on the probability score and the uncertainty measure.
8. An apparatus comprising:
means for receiving one or more input medical images of a patient;
means for performing a medical imaging analysis task from the one or more input medical images using a machine learning based network, the machine learning based network generating a probability score associated with the medical imaging analysis task;
means for determining an uncertainty measure representing an error associated with the probability score by:
selecting a calibration function comprising a fixed point defined according to a user selected threshold at which probability scores are totally uncertain;
applying the calibration function to the probability score, and
calculating an entropy of the probability score as the uncertainty measure based on results of the applied calibration function; and
means for making a clinical decision based on the probability score and the uncertainty measure.
9. The apparatus of claim 8 , wherein the means for making a clinical decision based on the probability score and the uncertainty measure comprises:
means for stratifying the patient into one of a plurality of patient groups based on the probability score and the uncertainty measure.
10. The apparatus of claim 9 , wherein the medical imaging analysis task comprises detection of an intracranial hemorrhage of the patient, and the plurality of patient groups comprises a high confidence positive detection patient group, a high confidence negative detection patient group, and a low confidence patient group.
11. A non-transitory computer readable medium storing computer program instructions, the computer program instructions when executed by a processor cause the processor to perform operations comprising:
receiving one or more input medical images of a patient;
performing a medical imaging analysis task from the one or more input medical images using a machine learning based network, the machine learning based network generating a probability score associated with the medical imaging analysis task;
determining an uncertainty measure representing an error associated with the probability score by:
selecting a calibration function comprising a fixed point defined according to a user selected threshold at which probability scores are totally uncertain;
applying the calibration function to the probability score, and
calculating an entropy of the probability score as the uncertainty measure based on results of the applied calibration function; and
making a clinical decision based on the probability score and the uncertainty measure.
12. The non-transitory computer readable medium of claim 11 , wherein the medical imaging analysis task comprises at least one of detection, subtyping, or segmentation of an intracranial hemorrhage of the patient.
13. The non-transitory computer readable medium of claim 11 , wherein making a clinical decision based on the probability score and the uncertainty measure comprises:
determining whether to treat the patient based on the probability score and the uncertainty measure.
14. The non-transitory computer readable medium of claim 11 , wherein making a clinical decision based on the probability score and the uncertainty measure comprises:
determining whether to perform a clinical test on the patient based on the probability score and the uncertainty measure.
15. The non-transitory computer readable medium of claim 11 , wherein making a clinical decision based on the probability score and the uncertainty measure comprises:
prioritizing a worklist of a radiologist based on the probability score and the uncertainty measure.