Polyp Detection with MDCT: A Phantom-Based Evaluation of the Impact of Dose and Spatial Resolution

HOME

HELP

FEEDBACK

SUBSCRIPTIONS

Polyp Detection with MDCT: A Phantom-Based Evaluation of the Impact of Dose and Spatial Resolution

Abdülvahit Özgün¹, Erik Rollvén², Lennart Blomqvist², Staffan Bremmer², Richard Odh¹ and Annette Fransson¹

¹ Department of Medical Physics, S-171 76 Karolinska University Hospital, Stockholm, Sweden.
² Department of Diagnostic Radiology, Karolinska University Hospital, Stockholm, Sweden.

View larger version (42K):

[in a new window]

Fig. 1. —Axial CT scans show polyps of different shapes. 1, pedunculated; 2, broad-based; 3, ulcerated or depressed with elevated margins; and 4, sessile or flat.

View larger version (100K):

[in a new window]

Fig. 2. —Endoluminal views from CT scans of colon polyps of different shapes. 1A and 1B, pedunculated; 2A and 2B, broad-based; 3A and 3B, ulcerated or depressed with elevated margins; and 4A and 4B, sessile or flat.

View larger version (124K):

[in a new window]

Fig. 3. —Photograph shows typical setup of two colon phantoms inside water tank used during CT.

View larger version (59K):

[in a new window]

Fig. 4. —Diagram shows rotational scheme of 24 phantoms used during data acquisition. R1, R2, and R3 refer to radiologists 1, 2, and 3, respectively. Colon phantoms are displayed as rectangular boxes. The reference number of phantom used to acquire data for a given sequence is indicated in the left subsection of each box. The numbers in the right subsection refer to the number of polyps in size groups X, Y, and Z, respectively. For each scan sequence, the number of polyps in each size group was kept constant and equal to 20. No radiologist examined the same phantom twice with the scheme, and all radiologists had examined all 24 phantoms after completion of the study.

View larger version (40K):

[in a new window]

Fig. 5. —Bar graph shows percentage of detected polyps as function of dose-length product (DLP) using the spatial resolution of the standard protocol. Black bars = 2D multiplanar reconstruction technique, striped bars = flythrough technique.

View larger version (15K):

[in a new window]

Fig. 6. —Graph shows percentage of detected polyps of different sizes as a function of dose-length product (DLP), which was obtained from data with the same spatial resolution as standard protocol. ${blacksquare}$ = polyps > 5 mm;

= 2- to 5-mm polyps;

= 0- to 2-mm polyps.

View larger version (40K):

[in a new window]

Fig. 7. —Graph shows percentage of detected polyps as a function of polyp size for different settings of slice width and pitch factor. Radiation dose (i.e., dose-length product) was kept constant and equal to standard sequence. Black bars = slice width of 1.25 mm and pitch of 0.75, gray bars = slice width of 1.25 mm and pitch of 1.5, striped bars = slice width of 2.5 mm and pitch of 1.5 (standard protocol), white bars = slice width of 3.75 mm and pitch of 0.75.

View larger version (42K):

[in a new window]

Fig. 8. —Graph shows detection rate of polyps in phantoms with and without water for three CT sequences. CT_1 used slice width of 1.25 mm and pitch of 0.75 (dose-length product [DLP], 100%). CT_7 used slice width of 2.5 mm and pitch of 1.5 (DLP, 80%); CT_10 used slice width of 3.75 mm and pitch of 0.75 (DLP, 60%). Black bars = 2D multiplanar reconstruction (MPR) technique, white bars = 2D MPR technique with water, gray bars = fly-through technique, striped bars = fly-through technique with water.

CiteULike

Complore

Connotea

Del.icio.us Add to Digg

Digg

Technorati What's this?