Anatomy modeling for tumor region of interest definition
A hot spot detection system for automatically segmenting and quantifying hot spots in functional images includes a segmentation unit ( 76 ) to segment an anatomical image representation ( 72 ) into regions corresponding to anatomical structures of a subject. A hot spot detection unit ( 90 ) detects regions of high uptake from a functional second image representation ( 74 ). The regions of high tracer uptake indicate high metabolic activity which maybe caused by potentially hazardous tumor lesions or other malignant processes. However, a number of normally functioning organs uptake high amounts of imaging tracer, particularly FDG. Therefore, a suppression unit ( 102 ) suppresses regions of high tracer uptake in the functional second image representation based on the results of a classification unit ( 101 ). The classification unit classifies the regions of high tracer uptake according to their position relative to the anatomical structures segmented from the anatomical first image representation. The remaining un-suppressed regions of high uptake are identified by an identification unit ( 106 ) as one of potential lesion and non-potential lesion.
1. A hot spot detection system, including:
a segmentation unit which segments an anatomical first image representation into regions corresponding to anatomical structures of a subject;
a hot spot detection unit which detects regions of high tracer uptake from a functional second image representation;
a classification unit which classifies the regions of high tracer uptake according to their position relative to the anatomical structures segmented from the anatomical first image representation;
a suppression unit which suppresses regions of high tracer uptake in the functional second image representation based on results of the classification unit; and
an identification unit which identifies unsuppressed regions of high uptake as one of potential lesion and non-potential lesion.
2. The hot spot detection system according to claim 1 , wherein the segmentation unit identifies and segments the anatomical structure based on a priori information according to at least one of shape, size, and geometrical location of the anatomical structure.
3. The hot spot detection system according to claim 1 , further including:
an uptake unit which examines segmented regions of first image representation for at least one of homogeneity and uptake to identify normal and abnormal regions.
4. The hot spot detection system according to claim 3 , wherein the suppression unit suppresses the regions of the functional image representation that corresponds to anatomical structures identified as normal.
5. The hot spot detection system according to claim 1 , further including:
a probability unit which determines a probability of the high uptake regions of the functional image representation corresponding to a lymph node.
6. The hot spot detection system according to claim 1 , further including:
a symmetry unit which determines symmetric pairs of the high uptake regions of the functional image representation that correspond to paired anatomical structures; and
an activity unit which compares metabolic activity between the determined symmetric pairs of high uptake regions to identify asymmetry in metabolic activity.
7. The hot spot detection system according to claim claim 1 , further comprising:
a nominal activity unit which determines a metric for the unsuppressed high intensity regions based on any one of total tumor burden, total glcolytic volume, and standardized uptake value.
8. The hot spot detection system according to claim 7 , wherein the nominal activity unit determines a nominal metabolic activity in a selected segmented region.
9. The hot spot detection system according to claim 8 , wherein the selected segmented region corresponds to a liver.
10. The hot spot detection system according to claim 1 , wherein the hot spot detection unit detects regions of high uptake using a watershed transform.
11. The hot spot detection system according to claim 1 , further including:
a workstation that outlines the detected regions of high uptake with a color overlay and displays the detected regions of high uptake on a display device.
12. The hot spot detection system according to claim 1 , wherein the regions of high uptake indicate high metabolic activity.
13. A diagnostic imaging system, comprising:
an anatomical image scanner configured to generate an anatomical first image representation;
a positron emission tomography (PET) scanner configured to generate a functional second image representation; and
a hot spot detection system according to claim 1 .
14. A method for diagnostic imaging, including:
segmenting a anatomical first image representation into regions corresponding to anatomical structures;
detecting from a functional second image representation a plurality of regions of high tracer uptake;
classifying the regions of high tracer uptake according to their position relative to the anatomical structures segmented from the anatomical first image representation;
suppressing regions of high uptake in the functional second image representation based on the classification of the high tracer uptake regions; and
identifying unsuppressed regions of high uptake as one of potential lesion and non-potential lesion.
15. The method according to claim 14 , further including:
determining symmetric pairs of the high uptake regions of the second image representation that correspond to paired anatomical structures based on anatomical information; and
comparing metabolic activity between the determined symmetric pairs of high uptake regions to identify asymmetry in metabolic activity.
16. The method according to claim 14 , further including:
normalizing metabolic activity levels of the unsuppressed regions of high uptake with a nominal intensity in a selected segmented region.
17. The method according to claim 14 , wherein segmenting the anatomical structure is based on a priori information according to at least one of shape, size, and geometrical location of the anatomical structure.
18. The method according to claim 14 , wherein the functional image representation is generated by a PET imaging scanner.
19. The method according to claim 14 , wherein the unsuppressed regions of high uptake are displayed.