Click here to

Session: Multi-Disciplinary: Imaging for Motion Management [Return to Session]

Verification of a System for Real-Time EPID Based Breath Hold Monitoring During DIBH Tangential Breast Cancer Radiotherapy

E Vasina1, P Greer2, D Thwaites3, J Lehmann4*, (1) University of Newcastle, AU, (2) Calvary Mater Newcastle, Newcastle, AU, (3) University of Sydney, Camperdown, Sydney, AU, (4) Newcastle Mater Hospital, Newcastle, AU


TH-IePD-TRACK 3-4 (Thursday, 7/29/2021) 12:30 PM - 1:00 PM [Eastern Time (GMT-4)]

Purpose: In breast cancer radiotherapy, the deep inspiration breath hold (DIBH) technique helps sparing heart, lungs, and liver from unintended radiation. The quality of DIBH is typically assessed via surrogates with varying degree of correlation with the patient’s anatomy. This work reports on the accuracy validation of an anatomy based system for assessment of the quality of DIBH using live MV images collected with the electronic portal imaging device (EPID).

Methods: The system consists of: (i) an image grabber application which acquires and saves MV EPID frames as TIFF files and (ii) image-processing software which analyses the files to assess the quality of the breath hold in real time. Measures used in the assessment are the lung depth (LD) and the distance from the breast surface to the posterior field edge (skin distance, SD) at user-defined locations. The system has been developed using 1000+ patient images.To assess the accuracy of the system, images of radiotherapy phantoms placed on and moved by a precision motion platform were analysed and the observed LD and SD were compared to those expected from the applied motion. A clinical surface imaging system was employed to confirm the phantom motions.

Results: The accuracy of LD and SD measurements for a PMMA CTDI phantom were 0.31(1.09) mm (mean (standard deviation)) and -0.10(0.14) mm, respectively. The accuracy of LD and SD measurements on an E2E SBRT phantom were 0.01(0.18) mm and 0.05(0.08) mm, respectively. The latency of the system is <350 ms: the sum of the time delays of the image grabber (100ms) and the image processing (<250ms) applications.

Conclusion: The system is able to assess the quality of DIBH breath hold via LD and SD with sub-millimetre accuracy. Processing times are short enough for real time clinical use. Clinical measurements with patients have begun.

Funding Support, Disclosures, and Conflict of Interest: This work is supported by National Health and Medical Research Council (NHMRC) grant 1147533 of the Australian Government.



    Not Applicable / None Entered.


    Not Applicable / None Entered.

    Contact Email