Prompt gamma emission prediction using a long short-term memory network

被引：0

作者：

Xiao, Fan ^{[1
]}

Radonic, Domagoj ^{[1
,2
]}

Kriechbaum, Michael ^{[2
]}

Wahl, Niklas ^{[3
,4
]}

Neishabouri, Ahmad ^{[4
,5
]}

Delopoulos, Nikolaos ^{[1
]}

Parodi, Katia ^{[2
]}

Corradini, Stefanie ^{[1
]}

Belka, Claus ^{[1
,6
,7
,8
]}

Kurz, Christopher ^{[1
]}

Landry, Guillaume ^{[1
]}

Dedes, George ^{[2
]}

机构：

[1] Ludwig Maximilians Univ Munchen, LMU Univ Hosp, Dept Radiat Oncol, Munich, Germany

[2] Ludwig Maximilians Univ Munchen, Dept Med Phys, Munich, Germany

[3] German Canc Res Ctr, Dept Med Phys Radiat Oncol, Heidelberg, Germany

[4] Heidelberg Inst Radiat Oncol HIRO, Natl Ctr Radiat Oncol NCRO, Heidelberg, Germany

[5] German Canc Res Ctr, Clin Cooperat Unit Radiat Oncol, Heidelberg, Germany

[6] German Canc Consortium DKTK, Partner Site Munich, Partnership DKFZ, Munich, Germany

[7] LMU Univ Hosp, Munich, Germany

[8] Bavarian Canc Res Ctr BZKF, Munich, Germany

来源：

PHYSICS IN MEDICINE AND BIOLOGY | 2024年 / 69卷 / 23期

关键词：

deep learning; proton therapy; prompt gamma; range verification; LSTM; BEAM RANGE VERIFICATION; MONTE-CARLO; PROTON THERAPY; UNCERTAINTIES; SENSITIVITY;

D O I：

10.1088/1361-6560/ad8e2a

中图分类号：

R318 [生物医学工程];

学科分类号：

0831 ;

摘要：

Objective: To present a long short-term memory (LSTM)-based prompt gamma (PG) emissionprediction method for proton therapy.Approach: Computed tomography (CT) scans of 33 patientswith a prostate tumor were included in the dataset. A set of 107histories proton pencil beam (PB)swas generated for Monte Carlo (MC) dose and PG simulation. For training (20 patients) andvalidation (3 patients), over 6000 PBs at 150, 175 and 200MeV were simulated. 3D relativestopping power (RSP), PG and dose cuboids that included the PB were extracted. Three modelswere trained, validated and tested based on an LSTM-based network: (1) input RSP and outputPG, (2) input RSP with dose and output PG (single-energy), and (3) input RSP/dose and outputPG (multi-energy). 540 PBs at each of the four energy levels (150, 175, 200, and 125-210MeV)were simulated across 10 patients to test the three models. The gamma passing rate (2%/2mm)and PG range shift were evaluated and compared among the three models.Results: The model withinput RSP/dose and output PG (multi-energy) showed the best performance in terms of gammapassing rate and range shift metrics. Its mean gamma passing rate of testing PBs of 125-210MeVwas 98.5% and the worst case was 92.8%. Its mean absolute range shift between predicted and MCPGs was 0.15mm, where the maximum shift was 1.1mm. The prediction time of our models waswithin 130ms per PB.Significance: We developed a sub-second LSTM-based PG emissionprediction method. Its accuracy in prostate patients has been confirmed across an extensive rangeof proton energies

引用

页数：17

共 50 条

[21] Using long short-term memory networks for river flow prediction [J].

Xu, Wei ;

Jiang, Yanan ;

Zhang, Xiaoli ;

Li, Yi ;

Zhang, Run ;

Fu, Guangtao .

HYDROLOGY RESEARCH, 2020, 51 (06) :1358-1376

[22] Regional Logistics Demand Prediction: A Long Short-Term Memory Network Method [J].

Li, Ya ;

Wei, Zhanguo .

SUSTAINABILITY, 2022, 14 (20)

[23] Convolutional long short-term memory neural network for groundwater change prediction [J].

Patra, Sumriti Ranjan ;

Chu, Hone-Jay .

FRONTIERS IN WATER, 2024, 6

[24] Antarctic sea ice prediction with A convolutional long short-term memory network [J].

Dong, Xiaoran ;

Yang, Qinghua ;

Nie, Yafei ;

Zampieri, Lorenzo ;

Wang, Jiuke ;

Liu, Jiping ;

Chen, Dake .

OCEAN MODELLING, 2024, 190

[25] A Long Short-Term Memory Network Stock Price Prediction with Leading Indicators [J].

Wu, Jimmy Ming-Tai ;

Sun, Lingyun ;

Srivastava, Gautam ;

Lin, Jerry Chun-Wei .

BIG DATA, 2021, 9 (05) :343-357

[26] Multimodal Continuous Prediction of Emotions in Movies using Long Short-Term Memory Networks [J].

Sivaprasad, Sarath ;

Joshi, Tanmayee ;

Agrawal, Rishabh ;

Pedanekar, Niranjan .

ICMR '18: PROCEEDINGS OF THE 2018 ACM INTERNATIONAL CONFERENCE ON MULTIMEDIA RETRIEVAL, 2018, :413-419

[27] Recognising formula entailment using long short-term memory network [J].

Pathak, Amarnath ;

Pakray, Partha .

JOURNAL OF INFORMATION SCIENCE, 2023,

[28] Short-Term Prediction of Wind Power Based on Deep Long Short-Term Memory [J].

Qu Xiaoyun ;

Kang Xiaoning ;

Zhang Chao ;

Jiang Shuai ;

Ma Xiuda .

2016 IEEE PES ASIA-PACIFIC POWER AND ENERGY ENGINEERING CONFERENCE (APPEEC), 2016, :1148-1152

[29] Short-term Individual Electric Vehicle Charging Behavior Prediction Using Long Short-term Memory Networks [J].

Khwaja, Ahmed S. ;

Venkatesh, Bala ;

Anpalagan, Alagan .

2020 IEEE 25TH INTERNATIONAL WORKSHOP ON COMPUTER AIDED MODELING AND DESIGN OF COMMUNICATION LINKS AND NETWORKS (CAMAD), 2020,

[30] Traffic Flow Prediction using Long Short-Term Memory Network and Optimized Spatial Temporal Dependencies [J].

Ara, Zinat ;

Hashemi, Mahdi .

2021 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA), 2021, :1550-1557

← 1 2 3 4 5 →