A Coordinate System for Tumor Identification in Positron Emission Tomography (PET) Imaging

Abstract

Purpose: PET can be useful in determining the progression of malignant disease over time as well as the response to therapy. To achieve this, the physician must be able to unambiguously identify and characterize individual tumors among several different scans.

Methods: We have developed a coordinate system for identifying individual tumor sites on PET scans, selecting the carina on the transmission scan as a point of origin. Using this system, each tumor is given a set of spherical coordinates that identifies its position: a rho (ρ, displacement from carina), a theta (θ, angle between the A–P axis and the tumor), and a phi (φ, angle between the polar axis and the tumor). We tested this method on a patient with metastatic thyroid cancer, who underwent ¹⁸FDG and ¹²⁴I-Iodide PET scans in the same week. This sytem was also used on a patient with metastatic prostate cancer, who had two FDG scans done 7 weeks apart. The patient underwent chemotherapy treatment during this period, and the scans were performed to assess therapy response.

Results: The patient with thyroid cancer had a total of 90 tumors, 82 of them identified in the ¹⁸FDG scan and 35 in the ¹²⁴I-Iodide scan, with 27 tumors identified in both. For ρ, θ, and φ among the 27 matching pairs of tumors, the mean differences were 6.80 ± 5 mm, 6.22 ± 4.54°, and 5.51 ± 5.81°, respectively. The disparity in coordinate values between corresponding tumors can be explained by the distinctive uptake patterns of the radiopharmaceuticals. The patient with prostate cancer had 9 tumors identifiable in both the pre- and post-therapy scans. The mean differences for ρ, θ, and φ among the 9 pairs of tumors were 1.93 ± 1.65 mm, 6.67 ± 5.53°, and 2.04 ± 2.02°, respectively. After thorough analysis, we have determined that corresponding tumors with ρ < 15 mm, θ and φ < 15° difference usually indicate a match.

Conclusion: This coordinate system facilitates the identification and characterization of individual tumors among multiple scans, thus aiding in both the assessment of diagnostic capabilities of different tracers, and the tracking of tumors following therapy.