This event will also have a BBQ and tour of the M-35 facilities! See the email for details.
Speakers:
Benjamin Puzantian, Carleton University
"Application of supervised machine learning to biological dosimetry of astronauts"
Radiation in low earth orbit (LEO) consists of a mixture of sparsely and densely ionizing radiation which can induce cytogenetic damage in astronauts. Biodosimetry using fluorescence in situ hybridization (FISH) has been a method routinely used to assess radiation exposure of astronauts from both the National Aeronautics and Space Administration (NASA) and the Canadian and European Space Agencies (CSA and ESA, performed at Health Canada) after long-duration missions to the International Space Station (ISS). Personalized calibration curves (CCs) were generated by irradiating pre-flight whole blood samples to different X-ray or γ-ray doses up to 2 Gy and measuring cytogenetic damage in lymphocytes using FISH. Post-flight assessments were done within a couple weeks of landing and repeated 6-18 months later. Due to the limited number of astronauts who have completed missions of at least three months to the ISS, it has been challenging to perform in-depth statistical analysis of their biodosimetry assessments. To address this, the biodosimetry datasets of 36 NASA and 10 CSA and ESA astronauts were pooled and anazlyzed.
Statistical analysis of damage pre- and post-flight revealed significant inter-individual and inter-agency differences for several types of enumerated damage. Supervised machine learning (SML) algorithms were developed by pooling the measured damage endpoints from the pre-flight in vitro samples along with additional sample information. Damage measured in post-flight samples was used to estimate radiation exposure with fits from both the CCs and SML algorithms. These estimates were compared to physical dosimetry, which revealed variations in biological doses amongst astronauts. While there were several advantages to using the SML model, CC estimation was found to be the preferred method for assessing space radiation exposures. In cases where the individual CC was missing or in question, the SML model combined with the pooled dataset provided a strong alternative for estimating exposure.
Jacques Dubeau, Detec
“The measurement of neutron energy spectra inside and outside the shielding of medical accelerators using the Nested Neutron Spectrometer.”
Under some operating conditions, neutrons may be generated by medical accelerators. In the case of therapy systems (LINACs and proton accelerators) the production of neutrons is a by-product of the treatment modality, and this may lead to unwanted patient dose. On the other hand, in the case of medical isotope production, neutrons are essential in the production of the desired material. The measurement of the neutron energy spectra inside an accelerator bunker is complicated by the very large neutron fluence rates. However, outside a well-designed shielding room, neutron fluence rates are necessarily very low. In this talk, the application of the Nested Neutron Spectrometer, for the measurement of energy differential neutron fluence spectra in the range of a few 10’s to nearly 108 cm-2 s-1, at LINACs, proton therapy accelerators and cyclotrons will be presented.