MLC positioning verification for small fields: a new investigation into automatic EPID-based verification methods

Joshua Hiatt, Godfrey Mukwada, Michael Barnes, Hans Lynggaard Riis, Du Huynh, Pejman Rowshanfarzad

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)


Multileaf-collimator (MLC) defined small fields in radiotherapy are used in high dose, ultra-conformal techniques such as stereotactic radiotherapy and stereotactic radiosurgery. Proximity to critical structures and irreversible damage arising from inaccurate delivery mean that correct positioning of the MLC system is of the utmost importance. Some of the existing techniques for MLC positioning quality assurance make use of electronic portal imaging device (EPID) images. However, conventional collimation verification algorithms based on the full width at half maximum (FWHM) fail when applied to small field images acquired by an EPID due to overlapping aperture penumbrae, lateral electron disequilibrium and radiation source occlusion. The objective of this study was to investigate sub-pixel edge detection and other techniques with the aim of developing an automatic and autonomous EPID-based method suitable for MLC positional verification of small static fields with arbitrary shapes. Methods investigated included derivative interpolation, Laplacian of Gaussian (LoG) and an algorithm based on the partial area effect hypothesis. None of these methods were found to be suitable for MLC positioning verification in small field conditions. A method is proposed which uses a manufacturer-specific empirically modified FWHM algorithm which shows improvement over the conventional techniques in the small field size range. With a measured mean absolute difference from planned position for Varian linacs of 0.01 ± 0.26 mm, compared with the erroneous FWHM value of 0.70 ± 0.51 mm. For Elekta linacs the proposed algorithm returned 0.26 ± 0.25 mm, in contrast to the FWHM result of 1.79 ± 1.07 mm.

Original languageEnglish
Pages (from-to)945-955
Number of pages11
JournalAustralasian Physical and Engineering Sciences in Medicine
Issue number4
Publication statusPublished - Dec 2018


Dive into the research topics of 'MLC positioning verification for small fields: a new investigation into automatic EPID-based verification methods'. Together they form a unique fingerprint.

Cite this