Urinary Tract Cancer Research


BLADDER CANCER DETECTION


Detection of bladder mass: example of mass in NC region. (a) Bladder segmentation. (b) Wall profile after thresholding to remove the urine within the entire bladder. The arrows indicate bladder lesion candidates. (c) Detected bladder mass candidates mapped back to CTU image.

Detection of bladder mass: example of mass in C region. (a) Bladder segmentation. (b) Wall profile after thresholding to remove the urine within the entire bladder. The arrows indicate bladder lesion candidates. The Orange arrow points to false positives, while the blue arrow points to a bladder lesion. (c) Detected bladder mass candidates mapped back to CTU image.

FROC curves for automatic computer detection after using LDA classifier.

Examples of detected bladder lesions (true positives). Lesions within both the contrast-enhanced and non-contrast regions of the bladder were correctly identified by the CAD system. (a) Lesion located in the non-contrast region. (b) Detected lesion within the contrast-enhanced bladder.