Forums General AlignRT setup results Reply To: AlignRT setup results

Marko Laaksomaa


Introduction: During this week I did some different additional calculations to presented FB mastectomy (M) and resection (n0) setup results, mostly regarding tangential image accuracy without IGRT. I added isocenter couch correction values (based on the online match and which include observer variation) in SI direction to offline matched tangential image MID_PTV ribs residual error values in SI. In that way we get isocenter errors in tangential images in SI direction without IGRT. All the fractions where only tangential image was acquired were also taken into account. Also residual errors without IGRT, to the most commonly used matching compromise point in between of MID_PTV ribs (lateral border) in AP image and MID_PTV sternum in LAT image in orthogonal images was calculated similarly in AP, SI and LAT directions (=online couch shift + error to the structure) in M group.

Results: Tangential image MID_PTV ribs (N, R) SI, M, L 0.16, 0.33, 0.65.M, A 0.1, 0.19, 0.38. n0, L 0.22, 0.35, 0.79. n0, A 0.11, 0.20, 0.41. and residual errors larger than 4.5 mm M, L 20%, A 4.5%, n0, L 24%, A 4%. Compromise MID_PTV ribs  (AP- image) + sternum (LAT- image) (N, R)  AP, SI, LAT, M, L 0.14, 0.16, 0.18, 0.37, 0.29, 0.26, 0.60, 0.63, 0.69. M, A 0.13, 0.15, 0.15, 0.19, 0.22, 0.23, 0.45, 0.53, 0.54, residual errors larger than 4.5 mm M, L 27, 20, 13%, A 7, 10, 9%. AP- image MID_PTV ribs- th6 (roll) (S) M, LAT L 0.11, 0.15. A 0.11, 0.10.

Conclusion: With AlignRT setup in group M, residual errors to the most commonly used compromised matching location in orthogonal images (MID_ribs + sternum) were in 90% of the fractions inside 4.5 mm accuracy after systematic isocenter error was corrected with 4 mm AL in the first three fractions. The possibility that more than 4.5mm errors exists without IGRT with mastectomy patients is approximately two times more possible 1) in total in group L than in group A in MID_ribs + sternum location and 2) in SI direction in location th1 (demonstrating lymph node area) than in location MID_ribs+sternum (demonstrating general matching point for tangential images ribs) in group A . Realized tangential image accuracy in MID_ribs without IGRT in SI direction was noticed adequate with the margins of 4mm in group A in both n0 and mastectomy groups.

Without IGRT (N)= systematic isocenter error was corrected with 4 mm AL during the first three fractions based on IGRT

R= residual error

S= error from structure to structure

systematic, random, margin with the margin formula 2.5Σ+0.7σ (cm)

I think this clinical study that I wanted to share at this point is finished now. It is free to comment and discuss. Do you feel any similarities with your findings?

Have a good weekend.