No more glowing in the dark: how deep learning improves exposure date estimation in thermoluminescence dosimetry. (15th November 2021)

• Record Type: Journal Article
• Authors: Mentzel, F; Derugin, E; Jansen, H; Kröninger, K; Nackenhorst, O; Walbersloh, J; Weingarten, J
• Is Part Of: Journal of radiological protection. Volume 41:Number 4(2021)
• Page Start: S506
• Page End: S521
• Publication Date: 2021-11-15
• Subjects: dose monitoring -- glow curve analysis -- machine learning -- convolutional neural networks -- fading time estimation; Radiation -- Safety measures -- Periodicals; Health Physics -- Periodicals; Radiation Monitoring -- Periodicals; Radiation Protection -- Periodicals; Rayonnement -- Sécurité -- Mesures -- Périodiques; Electronic journals; Fulltext; Internet Resource; Periodical; 363.179905
• DOI: 10.1088/1361-6498/ac20ae
• Languages: English
• ISSNs: 0952-4746

Abstract

Abstract: The time- or temperature-resolved detector signal from a thermoluminescence dosimeter can reveal additional information about circumstances of an exposure to ionising irradiation. We present studies using deep neural networks to estimate the date of a single irradiation with 12 mSv within a monitoring interval of 42 days from glow curves of novel TL-DOS personal dosimeters developed by the Materialprüfungsamt NRW in cooperation with TU Dortmund University. Using a deep convolutional network, the irradiation date can be predicted from raw time-resolved glow curve data with an uncertainty of roughly 1–2 days on a 68% confidence level without the need for a prior transformation into temperature space and a subsequent glow curve deconvolution (GCD). This corresponds to a significant improvement in prediction accuracy compared to a prior publication, which yielded a prediction uncertainty of 2–4 days using features obtained from a GCD as input to a neural network.