RSNA 2005 

Abstract Archives of the RSNA, 2005


LPB13-04

Diagnostic Accuracy of PET, CT, and Fused Image Interpretations in Patients with Malignant Lymphoma

Scientific Posters

Presented on November 27, 2005
Presented as part of LPB13: Nuclear Medicine (Lung Cancer, Lymphoma, and Staging with PET, SPECT, and Hybrid Techniques)

Participants

Munenobu Nogami MD, Presenter: Nothing to Disclose
Yuji Nakamoto MD,PhD, Abstract Co-Author: Nothing to Disclose
Setsu Sakamoto MD, Abstract Co-Author: Nothing to Disclose
Kazuto Fukushima MD, Abstract Co-Author: Nothing to Disclose
Tomohisa Okada MD, Abstract Co-Author: Nothing to Disclose
Tsuneo Saga MD, PhD, Abstract Co-Author: Nothing to Disclose
Tatsuya Higashi MD, PhD, Abstract Co-Author: Nothing to Disclose
Michio Senda MD, Abstract Co-Author: Nothing to Disclose
Kazuro Sugimura MD, PhD, Abstract Co-Author: Nothing to Disclose
et al, Abstract Co-Author: Nothing to Disclose

PURPOSE

Recently, the significance of image fusion has been recognized in oncology. Although our institute has no inline PET/CT system, whole-body FDG-PET and CT images are acquired separately using a fixation device for positioning, and combined manually on a workstation, yielding higher diagnostic confidence. The purpose of this study was to assess the diagnostic accuracy of interpretation of PET, CT and fused images, in patients with malignant lymphoma.

METHOD AND MATERIALS

Fifty patients (28 males and 22 females; mean age, 56.1 yr; range 20-76 yr) with histologically-confirmed malignant lymphoma underwent a FDG-PET scan, followed by a CT scan. The CT examinations were performed with (n=45) or without (n=5) contrast media. PET, CT and fused images were interpreted separately without any information and visually using a 5-point grading scale (0=definitely negative, 1=probably negative, 2=equivocal, 3=probably positive, 4=definitely positive) for the following 8 regions; cervical, supraclavicular, axillary, mediastinal, para-aortic to iliac, mesenteric, inguinal, and extra-nodal. Diagnostic accuracy was compared, based on the final diagnoses determined by histological confirmation and/or clinical course. Also, the receiver operating characteristic (ROC) analysis was applied.

RESULTS

There were 86 involved regions in 19 patients. Interpretation of PET images only (sensitivity = 83.7%, specificity = 98.7%, accuracy = 96.4%) and fused images (96.5%, 98.9%, 98.6%) yielded higher diagnostic performances than that of CT images (61.6%, 95.1%, 89.9%) (p<0.001). However, interpretation of PET and fused images did not differ significantly. According to the ROC analysis, area under the curve of fused and PET images was 0.999 and 0.986, respectively, while that of CT images was 0.831. Among the regions, the extra-nodal region was more accurately diagnosed in PET or fused images, as compared with CT interpretation (p<0.001).

CONCLUSION

Our preliminary data indicate that the most accurate diagnosis was obtained by interpreting fused images, followed by interpreting PET alone. Although fused images are clinically valuable, the area-based diagnostic accuracy of PET alone was also reasonably high in patients with lymphoma.

PURPOSE

Recently, the significance of image fusion has been recognized in oncology. Although our institute has no inline PET/CT system, whole-body FDG-PET and CT images are acquired separately using a fixation device for positioning, and combined manually on a workstation, yielding higher diagnostic confidence. The purpose of this study was to assess the diagnostic accuracy of interpretation of PET, CT and fused images, in patients with malignant lymphoma.

METHOD AND MATERIALS

Fifty patients (28 males and 22 females; mean age, 56.1 yr; range 20-76 yr) with histologically-confirmed malignant lymphoma underwent a FDG-PET scan, followed by a CT scan. The CT examinations were performed with (n=45) or without (n=5) contrast media. PET, CT and fused images were interpreted separately without any information and visually using a 5-point grading scale (0=definitely negative, 1=probably negative, 2=equivocal, 3=probably positive, 4=definitely positive) for the following 8 regions; cervical, supraclavicular, axillary, mediastinal, para-aortic to iliac, mesenteric, inguinal, and extra-nodal. Diagnostic accuracy was compared, based on the final diagnoses determined by histological confirmation and/or clinical course. Also, the receiver operating characteristic (ROC) analysis was applied.

RESULTS

There were 86 involved regions in 19 patients. Interpretation of PET images only (sensitivity = 83.7%, specificity = 98.7%, accuracy = 96.4%) and fused images (96.5%, 98.9%, 98.6%) yielded higher diagnostic performances than that of CT images (61.6%, 95.1%, 89.9%) (p<0.001). However, interpretation of PET and fused images did not differ significantly. According to the ROC analysis, area under the curve of fused and PET images was 0.999 and 0.986, respectively, while that of CT images was 0.831. Among the regions, the extra-nodal region was more accurately diagnosed in PET or fused images, as compared with CT interpretation (p<0.001).

CONCLUSION

Our preliminary data indicate that the most accurate diagnosis was obtained by interpreting fused images, followed by interpreting PET alone. Although fused images are clinically valuable, the area-based diagnostic accuracy of PET alone was also reasonably high in patients with lymphoma.

Cite This Abstract

Nogami, M, Nakamoto, Y, Sakamoto, S, Fukushima, K, Okada, T, Saga, T, Higashi, T, Senda, M, Sugimura, K, et al, , Diagnostic Accuracy of PET, CT, and Fused Image Interpretations in Patients with Malignant Lymphoma.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4419631.html