Tracking tumor growth rates in patients with malignant gliomas: a test of two algorithms.

BACKGROUND AND PURPOSE

Two 3D image analysis algorithms, nearest-neighbor tissue segmentation and surface modeling, were applied separately to serial MR images in patients with glioblastoma multiforme (GBM). Rates of volumetric change were tracked for contrast-enhancing tumor tissue. Our purpose was to compare the two image analysis algorithms in their ability to track tumor volume relative to a manually defined standard of reference.

METHODS

Three-dimensional T2-weighted and contrast-enhanced T1-weighted spoiled gradient-echo MR volumes were acquired in 10 patients with GBM. One of two protocols was observed: 1) a nearest-neighbor algorithm, which used manually determined or propagated tags and automatically segmented tissues into specific classes to determine tissue volume; or 2) a surface modeling algorithm, which used operator-defined contrast-enhancing boundaries to convert traced points into a parametric mesh model. Volumes were automatically calculated from the mesh models. Volumes determined by each algorithm were compared with the standard of reference, generated by manual segmentation of contrast-enhancing tissue in each cross section of a scan.

RESULTS

Nearest-neighbor algorithm enhancement volumes were highly correlated with manually segmented volumes, as were growth rates, which were measured in terms of halving and doubling times. Enhancement volumes generated by the surface modeling algorithm were also highly correlated with the standard of reference, although growth rates were not.

CONCLUSION

The nearest-neighbor tissue segmentation algorithm provides significant power in quantifying tumor volume and in tracking growth rates of contrast-enhancing tissue in patients with GBM. The surface modeling algorithm is able to quantify tumor volume reliably as well.