Literature DB >> 32743017

Machine learning for the prediction of pseudorealistic pediatric abdominal phantoms for radiation dose reconstruction.

Marco Virgolin1, Ziyuan Wang2, Tanja Alderliesten2,3, Peter A N Bosman1,4.   

Abstract

Purpose: Current phantoms used for the dose reconstruction of long-term childhood cancer survivors lack individualization. We design a method to predict highly individualized abdominal three-dimensional (3-D) phantoms automatically. Approach: We train machine learning (ML) models to map (2-D) patient features to 3-D organ-at-risk (OAR) metrics upon a database of 60 pediatric abdominal computed tomographies with liver and spleen segmentations. Next, we use the models in an automatic pipeline that outputs a personalized phantom given the patient's features, by assembling 3-D imaging from the database. A step to improve phantom realism (i.e., avoid OAR overlap) is included. We compare five ML algorithms, in terms of predicting OAR left-right (LR), anterior-posterior (AP), inferior-superior (IS) positions, and surface Dice-Sørensen coefficient (sDSC). Furthermore, two existing human-designed phantom construction criteria and two additional control methods are investigated for comparison.
Results: Different ML algorithms result in similar test mean absolute errors: ∼ 8    mm for liver LR, IS, and spleen AP, IS; ∼ 5    mm for liver AP and spleen LR; ∼ 80 % for abdomen sDSC; and ∼ 60 % to 65% for liver and spleen sDSC. One ML algorithm (GP-GOMEA) significantly performs the best for 6/9 metrics. The control methods and the human-designed criteria in particular perform generally worse, sometimes substantially ( + 5 - mm error for spleen IS, - 10 % sDSC for liver). The automatic step to improve realism generally results in limited metric accuracy loss, but fails in one case (out of 60).
Conclusion: Our ML-based pipeline leads to phantoms that are significantly and substantially more individualized than currently used human-designed criteria.
© 2020 Society of Photo-Optical Instrumentation Engineers (SPIE).

Entities:  

Keywords:  dose reconstruction; machine learning; pediatric cancer; phantom; radiation treatment

Year:  2020        PMID: 32743017      PMCID: PMC7390892          DOI: 10.1117/1.JMI.7.4.046501

Source DB:  PubMed          Journal:  J Med Imaging (Bellingham)        ISSN: 2329-4302


  42 in total

1.  The relationship between CT scout landmarks and lung boundaries on chest CT: guidelines for minimizing excess z-axis scan length.

Authors:  Stuart L Cohen; Thomas J Ward; Matthew D Cham
Journal:  Eur Radiol       Date:  2019-08-13       Impact factor: 5.315

Review 2.  A review of uncertainties in radiotherapy dose reconstruction and their impacts on dose-response relationships.

Authors:  Jérémi Vũ Bezin; Rodrigue S Allodji; Jean-Pierre Mège; Guillaume Beldjoudi; Fleur Saunier; Jean Chavaudra; Eric Deutsch; Florent de Vathaire; Valérie Bernier; Christian Carrie; Dimitri Lefkopoulos; Ibrahima Diallo
Journal:  J Radiol Prot       Date:  2017-01-24       Impact factor: 1.394

3.  Automated size-specific CT dose monitoring program: assessing variability in CT dose.

Authors:  Olav Christianson; Xiang Li; Donald Frush; Ehsan Samei
Journal:  Med Phys       Date:  2012-11       Impact factor: 4.071

4.  On the feasibility of automatically selecting similar patients in highly individualized radiotherapy dose reconstruction for historic data of pediatric cancer survivors.

Authors:  Marco Virgolin; Irma W E M van Dijk; Jan Wiersma; Cécile M Ronckers; Cees Witteveen; Arjan Bel; Tanja Alderliesten; Peter A N Bosman
Journal:  Med Phys       Date:  2018-03-07       Impact factor: 4.071

5.  Magnitude and variability of respiratory-induced diaphragm motion in children during image-guided radiotherapy.

Authors:  Sophie C Huijskens; Irma W E M van Dijk; Jorrit Visser; Coen R N Rasch; Tanja Alderliesten; Arjan Bel
Journal:  Radiother Oncol       Date:  2017-03-28       Impact factor: 6.280

6.  Reconstruction of organ dose for external radiotherapy patients in retrospective epidemiologic studies.

Authors:  Choonik Lee; Jae Won Jung; Christopher Pelletier; Anil Pyakuryal; Stephanie Lamart; Jong Oh Kim; Choonsik Lee
Journal:  Phys Med Biol       Date:  2015-02-26       Impact factor: 3.609

7.  Regularization Paths for Generalized Linear Models via Coordinate Descent.

Authors:  Jerome Friedman; Trevor Hastie; Rob Tibshirani
Journal:  J Stat Softw       Date:  2010       Impact factor: 6.440

8.  Evaluation of late adverse events in long-term wilms' tumor survivors.

Authors:  Irma W E M van Dijk; Foppe Oldenburger; Mathilde C Cardous-Ubbink; Maud M Geenen; Richard C Heinen; Jan de Kraker; Flora E van Leeuwen; Helena J H van der Pal; Huib N Caron; Caro C E Koning; Leontien C M Kremer
Journal:  Int J Radiat Oncol Biol Phys       Date:  2010-02-04       Impact factor: 7.038

9.  Basic anatomical and physiological data for use in radiological protection: reference values. A report of age- and gender-related differences in the anatomical and physiological characteristics of reference individuals. ICRP Publication 89.

Authors: 
Journal:  Ann ICRP       Date:  2002

10.  Navigator channel adaptation to reconstruct three dimensional heart volumes from two dimensional radiotherapy planning data.

Authors:  Angela Ng; Thao-Nguyen Nguyen; Joanne L Moseley; David C Hodgson; Michael B Sharpe; Kristy K Brock
Journal:  BMC Med Phys       Date:  2012-01-18
View more
  1 in total

1.  Validation and Comparison of Radiograph-Based Organ Dose Reconstruction Approaches for Wilms Tumor Radiation Treatment Plans.

Authors:  Ziyuan Wang; Marco Virgolin; Brian V Balgobind; Irma W E M van Dijk; Susan A Smith; Rebecca M Howell; Matthew M Mille; Choonsik Lee; Choonik Lee; Cécile M Ronckers; Peter A N Bosman; Arjan Bel; Tanja Alderliesten
Journal:  Adv Radiat Oncol       Date:  2022-07-04
  1 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.