Network for medical image analysis, decision support system, and related graphical user interface (GUI) applications

1. A network-based decision support system comprising:

a processor; and

a memory having instructions stored thereon, wherein the instructions, when executed by the processor, cause the processor to:

(i) receive and store a plurality of medical images in a database, each medical image associated with a corresponding patient;

(ii) access one or more of the medical images associated with a particular patient from the database upon user request for transmission to the user for display on a user computing device;

(iii) automatically analyze the one or more medical images using a machine learning algorithm; and

(iv) generate a radiologist report for the particular patient according to the one or more medical images for the patient,

wherein the one or more medical images comprise a composite image of the particular patient, the composite image comprising a CT scan overlaid with a nuclear medicine image obtained at a substantially same time as the CT scan and following administration to the patient of an imaging agent comprising a Prostate Specific Membrane Antigen (PSMA) binding agent comprising a radionuclide, wherein the instructions cause the processor to automatically analyze the composite image by:

(a) using the composite image to geographically identify a 3D boundary for each of one or more regions of imaged tissue within the nuclear medicine image; and

(b) computing, using the nuclear medicine image with the identified 3D boundary(ies) of the one or more region(s)

a risk map comprising a graphical denotation marking a region of risk of cancer or risk of recurrence of cancer.

2. The system of claim 1 , wherein the nuclear medicine image is a SPECT scan.

3. The system of claim 2 , wherein the imaging agent comprises a metal chelated to the PSMA binding agent, and wherein the metal is the radionuclide.

4. The system of claim 3 , wherein the imaging agent comprises 99m Tc-MIP-1404.

5. The system of claim 1 , wherein the nuclear medicine image is a PET scan.

6. The system of claim 5 , wherein the radionuclide is a radioisotope of a halogen.

7. The system of claim 5 , wherein the imaging agent comprises [18F]DCFPyL.

8. The system of claim 5 , wherein the radionuclide is a radioisotope of gallium (Ga).

9. The system of claim 1 , wherein the system is a cloud-based system.

10. The system of claim 1 , wherein the processor is a processor of one or more network or Internet host servers.

11. The method of claim 1 , wherein the cancer is prostate cancer.

12. The method of claim 11 , wherein the cancer is metastatic prostate cancer.

13. The method of claim 1 , wherein the imaged tissue comprises bone.

14. The method of claim 1 , wherein the imaged tissue comprises a prostate of the patient.

15. The system of claim 1 , wherein at step (b), the instructions cause the processor to compute a value of each of one or more risk indices, each risk index value indicative of cancer state or progression in the patient.

16. The system of claim 15 , wherein the plurality of medical images in the database comprise a series of medical images of a first patient taken over time, and wherein the instructions cause the processor to determine a value of at least a first risk index for each medical image of the series, thereby tracking determined values of at least the first risk index for the first patient over time.

17. The system of claim 15 , wherein at step (b) the instructions cause the processor to, for at least one risk index of the one or more risk indices, compute the value of the risk index by:

determining, for each of the one or more regions, a corresponding cancerous tissue level within the region based on intensity values of the nuclear medicine image within the 3D boundary of the region; and

computing the value of the risk index based on the determined cancerous tissue levels within the one or more regions.

18. A method comprising performing, by a processor of a server computing device, (i) to (v) as follows:

(i) receiving and storing, by the processor, a plurality of medical images in a database, each medical image associated with a corresponding patient;

(ii) accessing, by the processor, one or more of the medical images associated with a particular patient from the database upon user request for transmission to the user for display on a user computing device;

(iii) automatically analyzing, by the processor, the one or more medical images using a machine learning algorithm; and

(iv) generating, by the processor, a radiologist report for the particular patient according to the one or more medical images for the patient,

wherein the one or more medical images comprise a composite image of the particular patient, the composite image comprising a CT scan overlaid with a nuclear medicine image acquired at a substantially same time as the CT scan and following administration to the patient of an imaging agent comprising a Prostate Specific Membrane Antigen (PSMA) binding agent comprising a radionuclide, wherein the method comprises automatically analyzing the composite image by:

(a) using the composite image to geographically identify a 3D boundary for each of one or more regions of imaged tissue within the nuclear medicine image; and

(b) computing, using the nuclear medicine image with the identified 3D boundary(ies) of the one or more region(s),

a risk map comprising a graphical denotation marking regions of risk of cancer or risk of recurrence of cancer.

19. The method of claim 18 , wherein the nuclear medicine image is a SPECT scan.

20. The method of claim 19 , wherein the imaging agent comprises a metal chelated to the PSMA binding agent, and wherein the metal is the radionuclide.

21. The method of claim 20 , wherein the imaging agent comprises 99m Tc-MIP-1404.

22. The method of claim 18 , wherein the nuclear medicine image is a PET scan.

23. The method of claim 22 , wherein the radionuclide is a radioisotope of a halogen.

24. The method of claim 23 , wherein the imaging agent comprises [18F]DCFPyL.

25. The method of claim 22 , wherein the radionuclide is a radioisotope of gallium (Ga).

26. The method of claim 18 , wherein the processor is a processor of a cloud-based system.

27. The method of claim 18 , wherein the processor is a processor of one or more network or Internet host servers.

28. The method of claim 18 , wherein the cancer is prostate cancer.

29. The method of claim 28 , wherein the cancer is metastatic prostate cancer.

30. The method of claim 18 , wherein the imaged tissue comprises bone.

31. The method of claim 18 , wherein the imaged tissue comprises a prostate of the patient.

32. The method of claim 18 , wherein step (b) comprises computing a value of each of one or more risk indices, each risk index value indicative of cancer state or progression in the patient.

33. The method of claim 32 , wherein the plurality of medical images in the database comprise a series of medical images of a first patient taken over time, and wherein the method comprises determining a value of at least a first risk index for each medical image of the series, thereby tracking determined values of at least the first risk index over time.

34. The method of claim 33 , wherein the receiving and storing of the plurality of medical images comprises repeatedly receiving and storing, over time, a plurality of medical images of the first patient, each obtained at a different time, to obtain the series of medical images of the first patient.

35. The method of claim 32 , wherein step (b) comprises, for at least one risk index of the one or more risk indices, computing the value of the risk index by:

determining, for each of the one or more regions, a corresponding cancerous tissue level within the region based on intensity values of the nuclear medicine image within the 3D boundary of the region; and

computing the value of the risk index based on the determined cancerous tissue levels within the one or more regions.

36. A method for tracking prostate cancer progression and treatment efficacy over time, for one or more patient(s), the method comprising:

(a) repeatedly receiving and storing in a database, over time, by a processor of a computing device, a plurality of medical images for each of the one or more patient(s) to obtain, for each of the one or more patient(s), a series of medical images taken over time;

(b) for each of the one or more patient(s), automatically analyzing, by the processor, using a machine learning algorithm, the series of medical images for the patient to determine (i) values of one or more risk indices for each medical image of the series, thereby tracking determined values of the one or more risk indices over a course prostate cancer progression and treatment for the patient and (ii) a risk map comprising a graphical denotation marking a region of risk of cancer or risk of recurrence of cancer; and

(c) for each of the one or more patient(s), storing, by the processor, (i) the determined values of the one or more risk indices, each indicative of prostate cancer state or progression, and (ii) the risk map, for the patient for further processing or causing, by the processor, display of a graphical representation of the determined values of the one or more risk indices for the patient,

wherein the series of medical images for a particular patient of the one or more patient(s) comprises a composite images of the particular patient, each composite image comprising a CT scan overlaid with a corresponding nuclear medicine image acquired at a substantially same time as the CT scan and following administration to the patient of an imaging agent comprising a Prostate Specific Membrane Antigen (PSMA) binding agent comprising a radionuclide, and wherein step (b) comprises:

using the composite image to geographically identify a 3D boundary for each of one or more regions of imaged tissue within the corresponding nuclear medicine image; and

using the corresponding nuclear medicine image with the identified 3D boundary(ies) of the one or more region(s) to compute (i) the value of each of the one or more risk indices and (ii) the risk map.

37. The method of claim 36 , wherein the series of medical images for the particular patient of the one or more patient(s) comprises:

(i) a first image subseries comprising one or more medical images obtained using a first nuclear imaging modality each following administration to the particular patient of a first radiopharmaceutical; and

(ii) a second image subseries comprising one or more medical images obtained using a second nuclear imaging modality each following administration to the particular patient of a second radiopharmaceutical,

such that the values of the one or more risk indices determined in step (b) for the particular patient comprise a first subseries of values of a first risk index determined by automated analysis of the first image subseries and a second subseries of values of a second risk index determined by automated analysis of the second image subseries.

38. The method of claim 37 , wherein the medical images of first image subseries are obtained over a first period of time, when prostate cancer of the particular patient is localized, and the medical images of the second image subseries are obtained over a second period of time, when prostate cancer of the particular patient is metastatic.

39. The method of claim 37 , wherein:

the first image subseries comprises one or more composite SPECT-CT image(s), each composite SPECT-CT image comprising a CT scan overlaid with a SPECT scan acquired at substantially the same time;

the second image subseries comprises one or more composite PET-CT image(s), each composite PET-CT image comprising a CT scan overlaid with a PET scan acquired at substantially the same time; and

step (b) comprises:

automatically analyzing each of the one or more composite SPECT-CT images by:

using the composite SPECT-CT image to geographically identify a 3D boundary of a prostate region within the SPECT scan of the composite SPECT-CT image; and

computing a value of the first risk index using the SPECT scan with the identified 3D boundary of the prostate region; and

automatically analyzing each of the one or more composite PET-CT images by:

using the composite PET-CT image to geographically identify a 3D boundary of one or more metastatic regions within the PET scan of the composite PET-CT image, the one or metastatic regions including regions corresponding to patient tissue locations outside of the prostate; and

computing a value of the second risk index using the PET scan with the identified 3D boundaries of the one or more metastatic region(s).

40. The method of claim 37 , wherein:

the first image subseries comprises one or more composite SPECT-CT image(s), each composite SPECT-CT image comprising a CT scan overlaid with a SPECT scan acquired at substantially the same time;

the second image subseries comprises one or more whole-body scan(s); and

step (b) comprises:

automatically analyzing each of the one or more composite SPECT-CT images by:

using the composite SPECT-CT image to geographically identify a 3D boundary of a prostate region within the SPECT scan of the composite SPECT-CT image; and

computing a value of the first risk index using the SPECT scan with the identified 3D boundary of the prostate region; and

automatically analyzing each of the one or more whole-body scan(s) by:

geographically identifying a boundary of one or more metastatic regions within the whole-body scan, the one or metastatic regions including regions corresponding to patient tissue locations outside of the prostate; and

computing a value of the second risk index using the PET scan with the identified 3D boundaries of the one or more metastatic region(s).

41. A system for tracking prostate cancer progression and treatment efficacy over time, for one or more patient(s), the system comprising:

a processor; and

a memory having instructions stored thereon, wherein the instructions, when executed by the processor, cause the processor to:

(a) repeatedly receive and store in a database, over time, a plurality of medical images for each of the one or more patient(s) to obtain, for each of the one or more patient(s), a series of medical images taken over time;

(b) for each of the one or more patient(s), automatically analyze the series of medical images for the patient using a machine learning algorithm to determine (i) values of one or more risk indices for each medical image of the series, thereby tracking determined values of the one or more risk indices over a course prostate cancer progression and treatment for the patient and (ii) a risk map comprising a graphical denotation marking a region of risk of cancer or risk of recurrence of cancer; and

(c) for each of the one or more patient(s), store (i) the determined values of the one or more risk indices, each indicative of prostate cancer state or progression, and (ii) the risk map, for the patient for further processing or cause display of a graphical representation of the determined values of the one or more risk indices for the patient,

wherein the series of medical images for a particular patient of the one or more patient(s) comprises a composite images of the particular patient, each composite image comprising a CT scan overlaid with a corresponding nuclear medicine image acquired at a substantially same time as the CT scan and following administration to the particular patient of an imaging agent comprising a Prostate Specific Membrane Antigen (PSMA) binding agent comprising a radionuclide, and wherein step (b) comprises:

using the composite image to geographically identify a 3D boundary for each of one or more regions of imaged tissue within the corresponding nuclear medicine image; and

using the corresponding nuclear medicine image with the identified 3D boundary(ies) of the one or more region(s) to compute (i) the value of each of the one or more risk indices and (ii) the risk map.

42. The system of claim 41 , wherein the series of medical images for the particular patient of the one or more patient(s) comprises:

(i) a first image subseries comprising one or more medical images obtained using a first nuclear imaging modality each following administration to the particular patient of a first radiopharmaceutical; and

(ii) a second image subseries comprising one or more medical images obtained using a second nuclear imaging modality each following administration to the particular patient of a second radiopharmaceutical,

such that the values of the one or more risk indices determined in step (b) for the particular patient comprise a first subseries of values of a first risk index determined by automated analysis of the first image subseries and a second subseries of values of a second risk index determined by automated analysis of the second image subseries.

43. The system of claim 42 , wherein the medical images of first image subseries are obtained over a first period of time, when prostate cancer of the particular patient is localized, and the medical images of the second image subseries are obtained over a second period of time, when prostate cancer of the particular patient is metastatic.

44. The system of claim 42 , wherein:

the first image subseries comprises one or more composite SPECT-CT image(s), each composite SPECT-CT image comprising a CT scan overlaid with a SPECT scan acquired at substantially the same time;

the second image subseries comprises one or more composite PET-CT image(s), each composite PET-CT image comprising a CT scan overlaid with a PET scan acquired at substantially the same time; and

the instructions cause the processor to, at step (b):

automatically analyze each of the one or more composite SPECT-CT images by:

using the composite SPECT-CT image to geographically identify a 3D boundary of a prostate region within the SPECT scan of the composite SPECT-CT image; and

computing a value of the first risk index using the SPECT scan with the identified 3D boundary of the prostate region; and

automatically analyze each of the one or more composite PET-CT images by:

using the composite PET-CT image to geographically identify a 3D boundary of one or more metastatic regions within the PET scan of the composite PET-CT image, the one or metastatic regions including regions corresponding to patient tissue locations outside of the prostate; and

computing a value of the second risk index using the PET scan with the identified 3D boundaries of the one or more metastatic region(s).

45. The system of claim 42 , wherein:

the first image subseries comprises one or more composite SPECT-CT image(s), each composite SPECT-CT image comprising a CT scan overlaid with a SPECT scan acquired at substantially the same time;

the second image subseries comprises one or more whole-body scan(s); and

the instructions cause the processor to, at step (b):

automatically analyze each of the one or more composite SPECT-CT images by:

using the composite SPECT-CT image to geographically identify a 3D boundary of a prostate region within the SPECT scan of the composite SPECT-CT image; and

computing a value of the first risk index using the SPECT scan with the identified 3D boundary of the prostate region; and

automatically analyze each of the one or more whole-body scan(s) by:

geographically identifying a boundary of one or more metastatic regions within the whole-body scan, the one or metastatic regions including regions corresponding to patient tissue locations outside of the prostate; and

computing a value of the second risk index using the PET scan with the identified 3D boundaries of the one or more metastatic region(s).