TY - JOUR
T1 - Comparison of two-phase 201T1 SPECT with chest CT to differentiate thoracic malignancies from benign lesions
AU - Yu, Y. H.
AU - Hsu, W. H.
AU - Hsu, N. Y.
AU - Chang, Y. C.
AU - Tsai, P. P.
AU - Tseng, G. C.
AU - Sun, S. S.
AU - Chiui, Y. F.
PY - 2008/3
Y1 - 2008/3
N2 - Aim. This study was conducted to compare the performance of 201Tl single photon emission computed tomography (201Tl SPECT) with chest computed tomography (CT) in differentiating thoracic malignancies from benign lesions. Methods. One hundred and seventy patients with confirmed diagnostic thoracic lesions found in chest radiographs were prospectively examined by 201Tl SPECT. The performance of 201Tl SPECT in differentiating thoracic malignancies from benign lesions was evaluated in 161 patients with a measurable retention index (RI), using the region-of-interest method. Chest CT scans were retrospectively collected from 165 patients and were interpreted by two independent observers. Results. The areas under the receiver operating characteristics curves were 0.85 using the RI value to differentiate thoracic malignancies from benign lesions. The sensitivity, specificity, and accuracy were 71.9%,83.1%, and 76.4%, respectively, with a cutoff level for the RI set at 20%. Similarly, the sensitivity, specificity and accuracy of chest CT scans to differentiate malignancies from benign lesions were 78.2%, 69.7% and 74.9%, respectively. Focusing on patients with concordant results in both 201Tl SPECT and chest CT scam, we can differentiate thoracic malignancies from benign lesions with a sensitivity of 89. 1%, a specificity of 90%, and an accuracy of 89.4%. Conclusion. Both 201Tl SPECT and chest CT scans are useful imaging tools in differentiating thoracic malignancies from benign lesions, with an accuracy of around 75%. By combining these two image modalities, the accuracy improves to 89.4%, which may circumvent the need for invasive procedures for certain equivocal cases, using either single image alone.
AB - Aim. This study was conducted to compare the performance of 201Tl single photon emission computed tomography (201Tl SPECT) with chest computed tomography (CT) in differentiating thoracic malignancies from benign lesions. Methods. One hundred and seventy patients with confirmed diagnostic thoracic lesions found in chest radiographs were prospectively examined by 201Tl SPECT. The performance of 201Tl SPECT in differentiating thoracic malignancies from benign lesions was evaluated in 161 patients with a measurable retention index (RI), using the region-of-interest method. Chest CT scans were retrospectively collected from 165 patients and were interpreted by two independent observers. Results. The areas under the receiver operating characteristics curves were 0.85 using the RI value to differentiate thoracic malignancies from benign lesions. The sensitivity, specificity, and accuracy were 71.9%,83.1%, and 76.4%, respectively, with a cutoff level for the RI set at 20%. Similarly, the sensitivity, specificity and accuracy of chest CT scans to differentiate malignancies from benign lesions were 78.2%, 69.7% and 74.9%, respectively. Focusing on patients with concordant results in both 201Tl SPECT and chest CT scam, we can differentiate thoracic malignancies from benign lesions with a sensitivity of 89. 1%, a specificity of 90%, and an accuracy of 89.4%. Conclusion. Both 201Tl SPECT and chest CT scans are useful imaging tools in differentiating thoracic malignancies from benign lesions, with an accuracy of around 75%. By combining these two image modalities, the accuracy improves to 89.4%, which may circumvent the need for invasive procedures for certain equivocal cases, using either single image alone.
KW - Computed tomography
KW - Lung neoplasms
KW - Single photon
KW - Thallium radioisotopes
KW - Tomography, emission computed
UR - http://www.scopus.com/inward/record.url?scp=40449112963&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=40449112963&partnerID=8YFLogxK
M3 - Article
C2 - 18235422
AN - SCOPUS:40449112963
SN - 1824-4785
VL - 52
SP - 66
EP - 73
JO - Quarterly Journal of Nuclear Medicine and Molecular Imaging
JF - Quarterly Journal of Nuclear Medicine and Molecular Imaging
IS - 1
ER -