EP-1474: Feasibility of dose delivery error detection by a transmission detector for patient-specific QA

Abstract Background:The DVH-based and Gamma Index patient-specific QA measures commonly applied in radiotherapy planning are unable to simultaneously deliver detailed locations and magnitudes of discrepancy between isodoses of planned and delivered dose distributions. By exploiting the receiver operating characteristic (ROC) statistical classification tool, compliance between a planned and delivered isodose may be locally evaluated, both for OAR and PTV, at any given isodose level. Thus, a patient-specific QA tool may be developed to supplement those presently available in clinical radiotherapy.Materials and Methods:A method to locally establish and report dose delivery errors in 3-D isodoses of planned (reference) and delivered (evaluated) dose distributions as a function of both the dose level and spatial location was developed. At any given isodose level, the total volume of dose delivery containing the reference and the evaluated isodoses is locally decomposed into four subregions: true positive – subregions within both reference and evaluated isodoses, true negative – outside of both of these isodoses, false positive – inside the evaluated isodose occurs but not the reference one, and false negatives – inside the reference isodose but not the evaluated one, as established over the total volume of dose delivery. From this decomposition a confusion matrix is derived and various indices calculated to quantify the discrepancies between the given planned and delivered isodose distributions.Results:Examples of clinical photon radiotherapy plans underwent analysis using the method developed. At some isodose levels, at anatomically significant locations, dose delivery errors were found which would not be highlighted either by dose volume histogram (DVH)-based QA metrics or by gamma analysis.Conclusions:The proposed method which generalizes the DVH-based QA method approach and is able to spatially locate delivery errors at selected isodose levels may supplement the presently applied gamma analysis and DVH-based QA measures in patient-specific radiotherapy planning.

Download Full-text

Detecting MLC modeling errors using radiomics‐based machine learning in patient‐specific QA with an EPID for intensity‐modulated radiation therapy

Medical Physics ◽

10.1002/mp.14699 ◽

2020 ◽

Author(s):

Madoka Sakai ◽

Hisashi Nakano ◽

Daisuke Kawahara ◽

Satoshi Tanabe ◽

Takeshi Takizawa ◽

...

Keyword(s):

Machine Learning ◽

Radiation Therapy ◽

Intensity Modulated Radiation Therapy ◽

Patient Specific ◽

Intensity Modulated ◽

Modeling Errors ◽

Patient Specific Qa

Download Full-text

PO-1399: Evaluation of SRS MapCHECK™ for SABR patient-specific QA

Radiotherapy and Oncology ◽

10.1016/s0167-8140(21)01417-1 ◽

2020 ◽

Vol 152 ◽

pp. S743

Author(s):

Y. Miao ◽

G. Kidane ◽

A. Ifthaker ◽

L. Crees ◽

E. Almond

Keyword(s):

Patient Specific ◽

Patient Specific Qa

Download Full-text

A multi-institutional survey evaluating patient related QA – phase II

Current Directions in Biomedical Engineering ◽

10.1515/cdbme-2017-0134 ◽

2017 ◽

Vol 3 (2) ◽

pp. 639-642

Author(s):

Tobias Teichmann ◽

Henning Salz ◽

Michael Schwedas ◽

Simon Howitz ◽

Tilo Wiezorek

Keyword(s):

Phase I ◽

Phase Ii ◽

Reference Conditions ◽

Patient Specific ◽

Beam Model ◽

Medical Systems ◽

Output Factor ◽

Treatment Techniques ◽

Patient Specific Qa ◽

Output Factors

AbstractIn phase I of the survey a planning intercomparison of patient-related QA was performed at 12 institutions. The participating clinics created phantom based IMRT and VMAT plans which were measured utilizing the ArcCheck diode array. Mobius3D (M3D) was used in phase II. It acts as a secondary dose verification tool for patient-specific QA based on average linac beam data collected by Mobius Medical Systems. All Quasimodo linac plans will be analyzed for the continuation of the intercomparison. We aim to determine if Mobius3D is suited for use with diverse treatment techniques, if beam model customization is needed. Initially we computed first Mobius3D results by transferring all plans from phase I to our Mobius3D server. Because of some larger PTV mean dose differences we checked if output factor customization would be beneficial. We performed measurements and output factor correction to account for discrepancies in reference conditions. Compared to Mobius3D's preconfigured average beam data values, these corrected output factors differed by ±1.5% for field sizes between 7x7cm2 and 30x30cm2 and to −3.9% for 3x3cm2. Our method of correcting the output factors turns out good congruence to M3D's reference values for these medium field sizes.

Download Full-text