Threshold-driven optimization for reference-based auto-planning

Troy Long, Mingli Chen, Steve Jiang, Weiguo Lu

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

We study threshold-driven optimization methodology for automatically generating a treatment plan that is motivated by a reference DVH for IMRT treatment planning. We present a framework for threshold-driven optimization for reference-based auto-planning (TORA). Commonly used voxel-based quadratic penalties have two components for penalizing under- and over-dosing of voxels: a reference dose threshold and associated penalty weight. Conventional manual- and auto-planning using such a function involves iteratively updating the preference weights while keeping the thresholds constant, an unintuitive and often inconsistent method for planning toward some reference DVH. However, driving a dose distribution by threshold values instead of preference weights can achieve similar plans with less computational effort. The proposed methodology spatially assigns reference DVH information to threshold values, and iteratively improves the quality of that assignment. The methodology effectively handles both sub-optimal and infeasible DVHs. TORA was applied to a prostate case and a liver case as a proof-of-concept. Reference DVHs were generated using a conventional voxel-based objective, then altered to be either infeasible or easy-to-achieve. TORA was able to closely recreate reference DVHs in 5-15 iterations of solving a simple convex sub-problem. TORA has the potential to be effective for auto-planning based on reference DVHs. As dose prediction and knowledge-based planning becomes more prevalent in the clinical setting, incorporating such data into the treatment planning model in a clear, efficient way will be crucial for automated planning. A threshold-focused objective tuning should be explored over conventional methods of updating preference weights for DVH-guided treatment planning.

Original languageEnglish (US)
Article number04NT01
JournalPhysics in Medicine and Biology
Volume63
Issue number4
DOIs
StatePublished - Feb 8 2018

Fingerprint

Weights and Measures
Prostate
Liver

Keywords

  • automated planning
  • knowledge-based planning
  • treatment plan optimization

ASJC Scopus subject areas

  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging

Cite this

Threshold-driven optimization for reference-based auto-planning. / Long, Troy; Chen, Mingli; Jiang, Steve; Lu, Weiguo.

In: Physics in Medicine and Biology, Vol. 63, No. 4, 04NT01, 08.02.2018.

Research output: Contribution to journalArticle

@article{d94c6eae21254d9f93ed2bf0e69537c1,
title = "Threshold-driven optimization for reference-based auto-planning",
abstract = "We study threshold-driven optimization methodology for automatically generating a treatment plan that is motivated by a reference DVH for IMRT treatment planning. We present a framework for threshold-driven optimization for reference-based auto-planning (TORA). Commonly used voxel-based quadratic penalties have two components for penalizing under- and over-dosing of voxels: a reference dose threshold and associated penalty weight. Conventional manual- and auto-planning using such a function involves iteratively updating the preference weights while keeping the thresholds constant, an unintuitive and often inconsistent method for planning toward some reference DVH. However, driving a dose distribution by threshold values instead of preference weights can achieve similar plans with less computational effort. The proposed methodology spatially assigns reference DVH information to threshold values, and iteratively improves the quality of that assignment. The methodology effectively handles both sub-optimal and infeasible DVHs. TORA was applied to a prostate case and a liver case as a proof-of-concept. Reference DVHs were generated using a conventional voxel-based objective, then altered to be either infeasible or easy-to-achieve. TORA was able to closely recreate reference DVHs in 5-15 iterations of solving a simple convex sub-problem. TORA has the potential to be effective for auto-planning based on reference DVHs. As dose prediction and knowledge-based planning becomes more prevalent in the clinical setting, incorporating such data into the treatment planning model in a clear, efficient way will be crucial for automated planning. A threshold-focused objective tuning should be explored over conventional methods of updating preference weights for DVH-guided treatment planning.",
keywords = "automated planning, knowledge-based planning, treatment plan optimization",
author = "Troy Long and Mingli Chen and Steve Jiang and Weiguo Lu",
year = "2018",
month = "2",
day = "8",
doi = "10.1088/1361-6560/aaa731",
language = "English (US)",
volume = "63",
journal = "Physics in Medicine and Biology",
issn = "0031-9155",
publisher = "IOP Publishing Ltd.",
number = "4",

}

TY - JOUR

T1 - Threshold-driven optimization for reference-based auto-planning

AU - Long, Troy

AU - Chen, Mingli

AU - Jiang, Steve

AU - Lu, Weiguo

PY - 2018/2/8

Y1 - 2018/2/8

N2 - We study threshold-driven optimization methodology for automatically generating a treatment plan that is motivated by a reference DVH for IMRT treatment planning. We present a framework for threshold-driven optimization for reference-based auto-planning (TORA). Commonly used voxel-based quadratic penalties have two components for penalizing under- and over-dosing of voxels: a reference dose threshold and associated penalty weight. Conventional manual- and auto-planning using such a function involves iteratively updating the preference weights while keeping the thresholds constant, an unintuitive and often inconsistent method for planning toward some reference DVH. However, driving a dose distribution by threshold values instead of preference weights can achieve similar plans with less computational effort. The proposed methodology spatially assigns reference DVH information to threshold values, and iteratively improves the quality of that assignment. The methodology effectively handles both sub-optimal and infeasible DVHs. TORA was applied to a prostate case and a liver case as a proof-of-concept. Reference DVHs were generated using a conventional voxel-based objective, then altered to be either infeasible or easy-to-achieve. TORA was able to closely recreate reference DVHs in 5-15 iterations of solving a simple convex sub-problem. TORA has the potential to be effective for auto-planning based on reference DVHs. As dose prediction and knowledge-based planning becomes more prevalent in the clinical setting, incorporating such data into the treatment planning model in a clear, efficient way will be crucial for automated planning. A threshold-focused objective tuning should be explored over conventional methods of updating preference weights for DVH-guided treatment planning.

AB - We study threshold-driven optimization methodology for automatically generating a treatment plan that is motivated by a reference DVH for IMRT treatment planning. We present a framework for threshold-driven optimization for reference-based auto-planning (TORA). Commonly used voxel-based quadratic penalties have two components for penalizing under- and over-dosing of voxels: a reference dose threshold and associated penalty weight. Conventional manual- and auto-planning using such a function involves iteratively updating the preference weights while keeping the thresholds constant, an unintuitive and often inconsistent method for planning toward some reference DVH. However, driving a dose distribution by threshold values instead of preference weights can achieve similar plans with less computational effort. The proposed methodology spatially assigns reference DVH information to threshold values, and iteratively improves the quality of that assignment. The methodology effectively handles both sub-optimal and infeasible DVHs. TORA was applied to a prostate case and a liver case as a proof-of-concept. Reference DVHs were generated using a conventional voxel-based objective, then altered to be either infeasible or easy-to-achieve. TORA was able to closely recreate reference DVHs in 5-15 iterations of solving a simple convex sub-problem. TORA has the potential to be effective for auto-planning based on reference DVHs. As dose prediction and knowledge-based planning becomes more prevalent in the clinical setting, incorporating such data into the treatment planning model in a clear, efficient way will be crucial for automated planning. A threshold-focused objective tuning should be explored over conventional methods of updating preference weights for DVH-guided treatment planning.

KW - automated planning

KW - knowledge-based planning

KW - treatment plan optimization

UR - http://www.scopus.com/inward/record.url?scp=85042293681&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85042293681&partnerID=8YFLogxK

U2 - 10.1088/1361-6560/aaa731

DO - 10.1088/1361-6560/aaa731

M3 - Article

C2 - 29328050

AN - SCOPUS:85042293681

VL - 63

JO - Physics in Medicine and Biology

JF - Physics in Medicine and Biology

SN - 0031-9155

IS - 4

M1 - 04NT01

ER -