Non-contrast CT at comparable dose to an abdominal radiograph in patients with acute renal colic; impact of iterative reconstruction on image quality and diagnostic performance

Abstract

Objectives: The aim was to assess the performance of low-dose non-contrast CT of the urinary tract (LD-CT) acquired at radiation exposures close to that of abdominal radiography using adaptive statistical iterative reconstruction (ASiR).

Methods: Thirty-three patients with clinically suspected renal colic were prospectively included. Conventional dose (CD-CT) and LD-CT data sets were contemporaneously acquired. LD-CT images were reconstructed with 40 %, 70 % and 90 % ASiR. Image quality was subjectively and objectively measured. Images were also clinically interpreted.

Results: Mean ED was 0.48 ± 0.07 mSv for LD-CT compared with 4.43 ± 3.14 mSv for CD-CT. Increasing the percentage ASiR resulted in a step-wise reduction in mean objective noise (p < 0.001 for all comparisons). Seventy % ASiR LD-CT images had higher diagnostic acceptability and spatial resolution than 90 % ASiR LD-CT images (p < 0.001). Twenty-seven calculi (diameter = 5.5 ± 1.7 mm), including all ureteric stones, were correctly identified using 70 % ASiR LD-CT with two false positives and 16 false negatives (diameter = 2.3 ± 0.7 mm) equating to a sensitivity and specificity of 72 % and 94 %. Seventy % ASiR LD-CT had a sensitivity and specificity of 87 % and 100 % for detection of calculi >3 mm.

Conclusion: Reconstruction of LD-CT images with 70 % ASiR resulted in superior image quality than FBP, 40 % ASIR and 90 % ASIR. LD-CT with ASIR demonstrates high sensitivity and specificity for detection of calculi >3 mm.

Teaching points: • Low-dose CT studies for urinary calculus detection were performed with a mean dose of 0.48 ± 0.07 mSv • Low-dose CT with 70 % ASiR detected calculi >3 mm with a sensitivity and specificity of 87 % and 100 % • Reconstruction with 70 % ASiR was superior to filtered back projection, 40 % ASiR and 90 % ASiR images.

Fig. 1

Changes in mean objective noise (mean standard deviation of HU measured in 10 anatomical levels) encountered when 5-mm LD-CT images were reconstructed with FBP, and ASiR-40, ASiR-70, ASiR-90 and CD-CT images were reconstructed with FBP and ASiR-40 algorithms. (Star symbol denotes statistically significant difference, p < 0.05)

Fig. 2

Changes in subjective image quality parameters (median score as measured at 5 anatomical levels) encountered when 5-mm LD-CT images were reconstructed with FBP, ASiR-40, ASiR-70 and ASiR-90. (All differences were statistically significant except comparison of contrast resolution scores between ASiR-70 and ASiR-90 images)

Fig. 3

Changes in streak artefact score in the abdomen (light grey) and pelvic (dark grey) stations in patients with increasing BMI. 0, no streak artefact; 1, streak artefact present but not interfering with image interpretation; 2 streak artefact present and interfering with image interpretation

Fig. 4

A 48-year-old female patient (BMI = 25.6) with acute right ureteric colic. a and b Axial conventional dose (5.1 mSv) non-contrast CT image reconstructed with 40 % ASiR (a) and 100 % FBP (b) showing an 8-mm right ureteric calculus (arrowhead) and an additional 2-mm renal calculus in the right lower pole (arrow) that was not prospectively detected on the low-dose images. c-f Axial low-dose (0.56 mSv) non-contrast CT image reconstructed with 100 % FBP (c), 40 % ASiR (d), 70 % ASiR (e) and 90 % ASiR (f) showing the obstructing 8-mm calculus in the proximal right ureter (arrowhead). The location of the smaller 2.1-mm calculus is also demonstrated (arrow). Note the ‘over smoothening’ of the soft tissues on the 90 % ASiR image compared to the other low-dose reconstructions

Fig. 5

A 53-year-old female patient (BMI = 26.3) with acute left ureteric colic. a Axial low-dose (0.43 mSv) 2-mm non-contrast CT image reconstructed with 70 % ASiR showing a 5-mm calculus in the lower pole of the right kidney (arrowhead). b Axial conventional dose (5 mSv) 2-mm non-contrast CT image reconstructed with 40 % ASiR showing an additional 1.7-mm renal calculus in the right lower pole (arrow)

Fig. 6

A 40-year-old female patient (BMI = 29.8) with suspected renal calculi. a Axial low-dose (0.45-mSv) 2-mm non-contrast CT image reconstructed with 70 % ASiR and b axial conventional dose (7 mSv) 2-mm non-contrast CT image reconstructed with 40 % ASiR showing a 3.2-cm low-density left adrenal mass consistent with a benign adenoma (arrow). This left adrenal mass was missed when viewed on the low-dose non-contrast CT images

Fig. 7

A 29-year-old female patient (BMI = 20.8) with haematuria. a and b Axial conventional dose (3.6-mSv) 5-mm non-contrast CT images reconstructed with 40 % ASiR (a) and 100 % FBP (b) showing a 5.4-cm low-density left adnexal mass consistent with a dermoid cyst (arrow). This left adnexal mass was missed when viewed on the low-dose non-contrast CT images. c-f Axial low-dose (0.48-mSv) 5-mm non-contrast CT images reconstructed with 100 % FBP (c), 40 % ASiR (d), 70 % ASiR (e) and 90 % ASiR (f) showing the inferior detail of the left adnexal lesion relative to the conventional dose images (arrow). Again note the ‘over smoothening’ of the soft tissues on the 90 % ASiR image compared to the other low-dose reconstructions