Displaying and navigating computer-aided detection results on a review workstation
View Patent ↗Display and navigation for multiple computer-aided detection (CAD) detections is described. A medical image is displayed to a viewer, and a request is received to instantiate CAD-assisted viewing. A timewise presentation sequence for the CAD detections is automatically computed according to a predetermined sequencing criterion. For each CAD detection, an expanded presentation window is displayed for its associated location in the medical image, the expanded presentation windows being displayed according to the timewise presentation sequence. Also described is a navigational tool comprising a plot of an operating curve onto which a computed feature associated with the CAD algorithm can be mapped, the operating curve characterizing an application of the CAD algorithm to a reference database of mammographic cases having known diagnoses. The navigational tool further comprises a plurality of CAD pointer icons spatially distributed therealong according to mappings of the computed feature for the CAD detections onto the operating curve.
1. A computer-implemented method for displaying a plurality of computer-aided detection (CAD) detections associated with a tomosynthesis case, the tomosynthesis case comprising a plurality of images, the tomosynthesis case being processed by a CAD algorithm that identifies CAD detections therein based at least in part upon respective computed features associated with respective CAD detections: displaying the plurality of tomosynthesis images to a viewer, respective tomosynthesis images comprising respective CAD markers corresponding to respective locations of respective CAD detections in respective tomosynthesis images; mapping a computed feature onto an operating curve of a navigational tool, the navigational tool comprising a plot of the operating curve and a plurality of CAD pointer icons, respective CAD pointer icons corresponding to respective CAD detections, the plurality of CAD pointer icons being spatially distributed along the navigational tool according to the mapping, each CAD pointer icon representing a CAD detection and visually communicating an aspect of the CAD detection to the viewer by the position of the graphical symbol on the plot; and displaying the navigational tool in proximity to at least one tomosynthesis image to the viewer.
2. The method of claim 1 , wherein the plurality of CAD pointer icons having respective designations corresponding to respective designations of respective CAD detections for visual correspondence between the plurality of CAD pointer icons and respective CAD detections.
3. The method of claim 1 , visual correspondence being based at least in part upon a characteristic selected from the group consisting of a color, a shape and a size, and a flashing characteristic.
4. The method of claim 1 , wherein the plurality of CAD markers and respective CAD pointer icons have two-way rollover correspondence therebetween.
5. The method of claim 4 , two-way rollover correspondence comprising:
a mouse cursor being rolled over a CAD marker and the CAD pointer icon corresponding thereto brightening, or
a mouse cursor being rolled over a CAD pointer icon and the CAD marker corresponding thereto brightening.
6. The method of claim 1 , the operating curve plot characterizing an application of the CAD algorithm to a reference database of tomosynthesis cases having known diagnoses.
7. The method of claim 1 , the plurality of CAD pointer icons being spatially distributed along the navigational tool according to mappings of the particular computed feature onto the operating curve.
8. The method of claim 1 , wherein the computed feature is selected from the group consisting of: (a) a scalar overall probability metric thresholded by the CAD algorithm for qualifying candidate detections as becoming “detections” versus “non-detections”; (b) a mass area metric; (c) a mass edge roughness metric; (d) a number of calcifications in cluster metric; (e) a spiculatedness metric; and (f) a combination of other computed features.
9. The method of claim 1 , the computed feature comprising a scalar overall probability metric thresholded by the CAD algorithm for qualifying candidate detections as becoming “detections” versus “non-detections”, an abscissa of the operating curve comprising a false marks per case metric, and an ordinate of the operating curve comprising a metric proportional to a number of cancers detected by the CAD algorithm in the reference database when operating at the false marks per case metric.
10. The method of claim 1 , an abscissa of the operating curve comprising the computed feature, and an ordinate of the operating curve comprising a metric for number of cancers detected by the CAD algorithm in a reference database of tomosynthesis cases having known diagnoses and having the computed feature value.
11. The method of claim 1 , the operating curve being represented by one of a two-dimensional graph, a linear color-coded graph, and a linear intensity-coded graph.
12. The method of claim 1 , the operating curve comprising a histogram of a number of cancers detected in the reference database by the CAD algorithm at each possible value of the first computed feature.
13. The method of claim 1 , the operating curve comprising a receiver operating characteristic for the CAD algorithm as applied to a reference database of tomosynthesis cases having known diagnoses.
14. The method of claim 1 , each CAD detection being associated with a location in a tomosynthesis image, the method further comprising: receiving a viewer selection of one of the CAD pointer icons, responsive to the viewer selection, displaying an expanded presentation window for the location in the tomosynthesis image associated with the CAD detection corresponding to the selected CAD pointer icon.
15. The method of claim 1 , further comprising
receiving a request from the viewer to instantiate CAD-assisted viewing of the tomosynthesis case;
automatically computing a timewise presentation sequence for the CAD detections according to a predetermined sequencing criterion comprising one of a plurality of computed features of the CAD detections as computed by a CAD algorithm; and
for each CAD detection, displaying an expanded presentation window for its associated location in the associated tomosynthesis image,
the expanded presentation windows being sequentially displayed according to the computed timewise presentation sequence for the CAD detections.
16. The method of claim 1 , the CAD detections being sequenced according to one of a plurality of computed features thereof as computed by a CAD algorithm associated with the CAD detections.
17. The method of claim 16 , the computed feature used to sequence the CAD detections being viewer-selectable.
18. The method of claim 17 , each expanded presentation window being superimposed upon the associated tomosynthesis image near the associated location.
19. The method of claim 15 , each expanded presentation windows allowing for at least one of (i) viewer manipulation of scale, (ii) viewer manipulation of tone, (iii) interactive viewing of CAD-computed feature characteristics, and (iv) interactive segmentation or analysis.
20. The method of claim 15 , the expanded presentation windows being sequentially displayed according to the computed timewise presentation sequence for the CAD detections such that one expanded presentation window is presented at a time.