Uncertainty-aware spot rejection rate as quality metric for proton therapy using a digital tracking calorimeter

被引：4

作者：

Schilling, Alexander ^{[1
,2
]}

Aehle, Max ^{[2
]}

Alme, Johan ^{[3
]}

Barnafoeldi, Gergely Gabor ^{[4
]}

Bodova, Tea ^{[3
]}

Borshchov, Vyacheslav ^{[5
]}

van den Brink, Anthony ^{[6
]}

Eikeland, Viljar ^{[3
]}

Feofilov, Gregory ^{[7
]}

Garth, Christoph ^{[8
]}

Gauger, Nicolas R. ^{[2
]}

Grottvik, Ola ^{[3
]}

Helstrup, Havard ^{[9
]}

Igolkin, Sergey ^{[7
]}

Keidel, Ralf ^{[1
,2
]}

Kobdaj, Chinorat ^{[10
]}

Kortus, Tobias ^{[1
]}

Leonhardt, Viktor ^{[8
]}

Mehendale, Shruti ^{[3
]}

Mulawade, Raju Ningappa ^{[1
]}

Odland, Odd Harald ^{[3
,11
]}

O'Neill, George ^{[3
]}

Papp, Gabor ^{[12
]}

Peitzmann, Thomas ^{[6
]}

Pettersen, Helge Egil Seime ^{[11
]}

Piersimoni, Pierluigi ^{[3
,13
]}

Protsenko, Maksym ^{[5
]}

Rauch, Max ^{[3
]}

Rehman, Attiq Ur ^{[3
]}

Richter, Matthias ^{[3
]}

Roehrich, Dieter ^{[3
]}

Santana, Joshua ^{[1
]}

Seco, Joao ^{[14
,15
]}

Songmoolnak, Arnon ^{[3
,10
]}

Sudar, Akos ^{[4
,16
]}

Tambave, Ganesh ^{[17
]}

Tymchuk, Ihor ^{[5
]}

Ullaland, Kjetil ^{[3
]}

Varga-Kofarago, Monika ^{[4
]}

Volz, Lennart ^{[18
,19
]}

Wagner, Boris ^{[3
]}

Wendzel, Steffen ^{[1
]}

Wiebel, Alexander ^{[1
]}

Xiao, Renzheng ^{[3
,20
]}

Yang, Shiming ^{[3
]}

Zillien, Sebastian ^{[1
]}

机构：

[1] Univ Appl Sci Worms, Ctr Technol & Transfer ZTT, D-67549 Worms, Germany

[2] Univ Kaiserslautern Landau, Chair Sci Comp, D-67663 Kaiserslautern, Germany

[3] Univ Bergen, Dept Phys & Technol, NO-5007 Bergen, Norway

[4] Wigner Res Ctr Phys, Budapest, Hungary

[5] Res & Prod Enterprise LTU RPELTU, Kharkiv, Ukraine

[6] Univ Utrecht, Inst Subatom Phys, Nikhef, Utrecht, Netherlands

[7] St Petersburg Univ, St Petersburg, Russia

[8] Univ Kaiserslautern Landau, Sci Visualizat Lab, D-67663 Kaiserslautern, Germany

[9] Western Norway Univ Appl Sci, Dept Comp Sci Elect Engn & Math Sci, NO-5020 Bergen, Norway

[10] Suranaree Univ Technol, Inst Sci, Nakhon Ratchasima, Thailand

[11] Haukeland Hosp, Dept Oncol & Med Phys, NO-5021 Bergen, Norway

[12] Eotvos Lorand Univ, Inst Phys, 1-A Pazmany P Setany, H-1117 Budapest, Hungary

[13] UniCamillus St Camillus Int Univ Hlth Sci, Rome, Italy

[14] DKFZ German Canc Res Ctr, Dept Biomed Phys Radiat Oncol, Heidelberg, Germany

[15] Heidelberg Univ, Dept Phys & Astron, Heidelberg, Germany

[16] Budapest Univ Technol & Econ, Budapest, Hungary

[17] Natl Inst Sci Educ & Res NISER, Ctr Med & Radiat Phys CMRP, Bhubaneswar, India

[18] GSI Helmholtz Ctr Heavy Ion Res GmbH, Biophys, Darmstadt, Germany

[19] UCL, Dept Med Phys & Biomed Engn, London, England

[20] China Three Gorges Univ, Coll Mech & Power Engn, People's Republ China, Yichang, Peoples R China

来源：

PHYSICS IN MEDICINE AND BIOLOGY | 2023年 / 68卷 / 19期

关键词：

machine learning; particle therapy; range verification; uncertainty; PROMPT GAMMA; RANGE UNCERTAINTIES; OPTIMIZATION; PET; VERIFICATION; TOOLKIT; PHANTOM;

D O I：

10.1088/1361-6560/acf5c2

中图分类号：

R318 [生物医学工程];

学科分类号：

0831 ;

摘要：

Objective. Proton therapy is highly sensitive to range uncertainties due to the nature of the dose deposition of charged particles. To ensure treatment quality, range verification methods can be used to verify that the individual spots in a pencil beam scanning treatment fraction match the treatment plan. This study introduces a novel metric for proton therapy quality control based on uncertainties in range verification of individual spots. Approach. We employ uncertainty-aware deep neural networks to predict the Bragg peak depth in an anthropomorphic phantom based on secondary charged particle detection in a silicon pixel telescope designed for proton computed tomography. The subsequently predicted Bragg peak positions, along with their uncertainties, are compared to the treatment plan, rejecting spots which are predicted to be outside the 95% confidence interval. The such-produced spot rejection rate presents a metric for the quality of the treatment fraction. Main results. The introduced spot rejection rate metric is shown to be well-defined for range predictors with well-calibrated uncertainties. Using this method, treatment errors in the form of lateral shifts can be detected down to 1 mm after around 1400 treated spots with spot intensities of 1 x 107 protons. The range verification model used in this metric predicts the Bragg peak depth to a mean absolute error of 1.107 & PLUSMN; 0.015 mm. Significance. Uncertainty-aware machine learning has potential applications in proton therapy quality control. This work presents the foundation for future developments in this area.

引用

页数：17

共 46 条

[21]

King DB, 2015, ACS SYM SER, V1214, P1, DOI 10.1021/bk-2015-1214.ch001

[22] In vivo proton range verification: a review [J].