Limits...
Value of T1/T2-weighted magnetic resonance imaging registration to reduce the postbiopsy hemorrhage effect for prostate cancer localization.

You JY, Lee HJ, Hwang SI, Bae YJ, Kim H, Hong H, Choe G - Prostate Int (2015)

Bottom Line: For the trainees (Reader 3 and Reader 4), the AUC values were significantly higher (P < 0.05) for T1/T2RI (0.60 and 0.62, respectively) than for T2WI (0.54 and 0.56, respectively) in tumor detection, whereas no significant difference was observed for faculty members.There was no significant difference in AUC values between T1/T2RI and T2WI + DWI for all readers except for Reader 1.There was no additional diagnostic benefit for adding DWI with T1/T2RI for all readers.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiology, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, South Korea.

ABSTRACT

Background: The aim of this study was to evaluate the value of T1/T2-weighted imaging (T1/T2WI) registration to reduce the postbiopsy hemorrhage effect for prostate cancer localization on prostate magnetic resonance imaging (MRI).

Methods: Twenty-one men with pathology-proven prostate cancer who underwent preoperative MRI in a single institution were selected. The zonal anatomy was divided into 16 sections. T2WI, T1/T2-weighted registered imaging (T1/T2RI), T2WI combined with diffusion-weighted imaging (T2WI + DWI), and T1/T2RI combined with DWI (T1/T2RI + DWI) were scored for the likelihood of cancer by two radiology faculty members and two trainees, and were compared with histology results. Areas under the receiver operating characteristics curve (AUCs) were used to assess diagnostic accuracy.

Results: For the trainees (Reader 3 and Reader 4), the AUC values were significantly higher (P < 0.05) for T1/T2RI (0.60 and 0.62, respectively) than for T2WI (0.54 and 0.56, respectively) in tumor detection, whereas no significant difference was observed for faculty members. There was no significant difference in AUC values between T1/T2RI and T2WI + DWI for all readers except for Reader 1. There was no additional diagnostic benefit for adding DWI with T1/T2RI for all readers.

Conclusions: T1/T2WI registration is a feasible technique. For less experienced readers, T1/T2RI is better than T2WI in localization of prostate cancer.

No MeSH data available.


Related in: MedlinePlus

Good conspicuity of prostate cancer on T1/T2-weighted registered imaging in a 67-year-old man. (A) T2-weighted imaging shows suspicious focal low signal intensity in the right posterior peripheral zone at the apex of the prostate (arrow). The interobserver agreement for this region was poor because it was unclear. (B) T1-weighted imaging shows focal high signal intensity in that region (arrow), suggestive of partial hemorrhage. (C) After data processing, this lesion is shown as a more conspicuous low signal intensity region with the relatively clear margin of the tumor (arrow) on T1/T2-weighted registered imaging. (D) A photomicrograph of the pathologic specimen (hematoxylin–eosin stain; original magnification, 1×) shows a tumor with Gleason score of 7 (outlined area with arrow) in the corresponding region.
© Copyright Policy - CC BY-NC-ND
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4588389&req=5

fig5: Good conspicuity of prostate cancer on T1/T2-weighted registered imaging in a 67-year-old man. (A) T2-weighted imaging shows suspicious focal low signal intensity in the right posterior peripheral zone at the apex of the prostate (arrow). The interobserver agreement for this region was poor because it was unclear. (B) T1-weighted imaging shows focal high signal intensity in that region (arrow), suggestive of partial hemorrhage. (C) After data processing, this lesion is shown as a more conspicuous low signal intensity region with the relatively clear margin of the tumor (arrow) on T1/T2-weighted registered imaging. (D) A photomicrograph of the pathologic specimen (hematoxylin–eosin stain; original magnification, 1×) shows a tumor with Gleason score of 7 (outlined area with arrow) in the corresponding region.

Mentions: Our results showed that the T1/T2WI registration technique was feasible for reducing the effect of postbiopsy hemorrhage on T2WI. In cases with obvious postbiopsy hemorrhage, hemorrhage was often misdiagnosed as prostate cancer on T2WI. This is more common with less experienced readers, although it was not statistically significant. In those cases, hypointensity of hemorrhage on T2WI was completely corrected using the image registration technique, and therefore, it showed almost the same signal intensity as the surrounding normal prostate tissue on T1/T2RI (Fig. 4). In cases with a small amount of postbiopsy hemorrhage, tumor extent is sometimes overestimated because of partially overlapped hemorrhage. In those cases, hypointensity of the prostate cancer became more conspicuous with clear margin on T1/T2RI than that on T2WI, because the partially overlapped hemorrhage was suppressed using the image registration technique (Fig. 5). Furthermore, the quality of T1/T2RI was acceptable for interpretation without severe misregistration artifact. Although there was the interface artifact between the hemorrhage and surrounding tissue in some cases, it was minor and negligible. The quality of T1/T2RI seems to be decided by the qualities of the routine T1WI and T2WI.


Value of T1/T2-weighted magnetic resonance imaging registration to reduce the postbiopsy hemorrhage effect for prostate cancer localization.

You JY, Lee HJ, Hwang SI, Bae YJ, Kim H, Hong H, Choe G - Prostate Int (2015)

Good conspicuity of prostate cancer on T1/T2-weighted registered imaging in a 67-year-old man. (A) T2-weighted imaging shows suspicious focal low signal intensity in the right posterior peripheral zone at the apex of the prostate (arrow). The interobserver agreement for this region was poor because it was unclear. (B) T1-weighted imaging shows focal high signal intensity in that region (arrow), suggestive of partial hemorrhage. (C) After data processing, this lesion is shown as a more conspicuous low signal intensity region with the relatively clear margin of the tumor (arrow) on T1/T2-weighted registered imaging. (D) A photomicrograph of the pathologic specimen (hematoxylin–eosin stain; original magnification, 1×) shows a tumor with Gleason score of 7 (outlined area with arrow) in the corresponding region.
© Copyright Policy - CC BY-NC-ND
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4588389&req=5

fig5: Good conspicuity of prostate cancer on T1/T2-weighted registered imaging in a 67-year-old man. (A) T2-weighted imaging shows suspicious focal low signal intensity in the right posterior peripheral zone at the apex of the prostate (arrow). The interobserver agreement for this region was poor because it was unclear. (B) T1-weighted imaging shows focal high signal intensity in that region (arrow), suggestive of partial hemorrhage. (C) After data processing, this lesion is shown as a more conspicuous low signal intensity region with the relatively clear margin of the tumor (arrow) on T1/T2-weighted registered imaging. (D) A photomicrograph of the pathologic specimen (hematoxylin–eosin stain; original magnification, 1×) shows a tumor with Gleason score of 7 (outlined area with arrow) in the corresponding region.
Mentions: Our results showed that the T1/T2WI registration technique was feasible for reducing the effect of postbiopsy hemorrhage on T2WI. In cases with obvious postbiopsy hemorrhage, hemorrhage was often misdiagnosed as prostate cancer on T2WI. This is more common with less experienced readers, although it was not statistically significant. In those cases, hypointensity of hemorrhage on T2WI was completely corrected using the image registration technique, and therefore, it showed almost the same signal intensity as the surrounding normal prostate tissue on T1/T2RI (Fig. 4). In cases with a small amount of postbiopsy hemorrhage, tumor extent is sometimes overestimated because of partially overlapped hemorrhage. In those cases, hypointensity of the prostate cancer became more conspicuous with clear margin on T1/T2RI than that on T2WI, because the partially overlapped hemorrhage was suppressed using the image registration technique (Fig. 5). Furthermore, the quality of T1/T2RI was acceptable for interpretation without severe misregistration artifact. Although there was the interface artifact between the hemorrhage and surrounding tissue in some cases, it was minor and negligible. The quality of T1/T2RI seems to be decided by the qualities of the routine T1WI and T2WI.

Bottom Line: For the trainees (Reader 3 and Reader 4), the AUC values were significantly higher (P < 0.05) for T1/T2RI (0.60 and 0.62, respectively) than for T2WI (0.54 and 0.56, respectively) in tumor detection, whereas no significant difference was observed for faculty members.There was no significant difference in AUC values between T1/T2RI and T2WI + DWI for all readers except for Reader 1.There was no additional diagnostic benefit for adding DWI with T1/T2RI for all readers.

View Article: PubMed Central - PubMed

Affiliation: Department of Radiology, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, South Korea.

ABSTRACT

Background: The aim of this study was to evaluate the value of T1/T2-weighted imaging (T1/T2WI) registration to reduce the postbiopsy hemorrhage effect for prostate cancer localization on prostate magnetic resonance imaging (MRI).

Methods: Twenty-one men with pathology-proven prostate cancer who underwent preoperative MRI in a single institution were selected. The zonal anatomy was divided into 16 sections. T2WI, T1/T2-weighted registered imaging (T1/T2RI), T2WI combined with diffusion-weighted imaging (T2WI + DWI), and T1/T2RI combined with DWI (T1/T2RI + DWI) were scored for the likelihood of cancer by two radiology faculty members and two trainees, and were compared with histology results. Areas under the receiver operating characteristics curve (AUCs) were used to assess diagnostic accuracy.

Results: For the trainees (Reader 3 and Reader 4), the AUC values were significantly higher (P < 0.05) for T1/T2RI (0.60 and 0.62, respectively) than for T2WI (0.54 and 0.56, respectively) in tumor detection, whereas no significant difference was observed for faculty members. There was no significant difference in AUC values between T1/T2RI and T2WI + DWI for all readers except for Reader 1. There was no additional diagnostic benefit for adding DWI with T1/T2RI for all readers.

Conclusions: T1/T2WI registration is a feasible technique. For less experienced readers, T1/T2RI is better than T2WI in localization of prostate cancer.

No MeSH data available.


Related in: MedlinePlus