Predicting COVID-19 mortality with electronic medical records.

This study aims to predict death after COVID-19 using only the past medical information routinely collected in electronic health records (EHRs) and to understand the differences in risk factors across age groups. Combining computational methods and clinical expertise, we curated clusters that represent 46 clinical conditions as potential risk factors for death after a COVID-19 infection. We trained age-stratified generalized linear models (GLMs) with component-wise gradient boosting to predict the probability of death based on what we know from the patients before they contracted the virus. Despite only relying on previously documented demographics and comorbidities, our models demonstrated similar performance to other prognostic models that require an assortment of symptoms, laboratory values, and images at the time of diagnosis or during the course of the illness. In general, we found age as the most important predictor of mortality in COVID-19 patients. A history of pneumonia, which is rarely asked in typical epidemiology studies, was one of the most important risk factors for predicting COVID-19 mortality. A history of diabetes with complications and cancer (breast and prostate) were notable risk factors for patients between the ages of 45 and 65 years. In patients aged 65-85 years, diseases that affect the pulmonary system, including interstitial lung disease, chronic obstructive pulmonary disease, lung cancer, and a smoking history, were important for predicting mortality. The ability to compute precise individual-level risk scores exclusively based on the EHR is crucial for effectively allocating and distributing resources, such as prioritizing vaccination among the general population.

Optimisation of exposure conditions for in vitro radiobiology experiments.

Despite the long history of using cell cultures in vitro for radiobiological studies, there is to date no study specifically addressing the dosimetric implications of flask selection and exposure environment in clonogenic assays. The consequent variability in dosimetry between laboratories impedes the comparison of results. In this study we compare the dose to cells adherent to the base of three types of commonly used culture flasks or plates. The cells are exposed to a 6MV clinical photon beam using either an open or a half blocked field. The depth of medium in each flask is varied with the medium surrounding the flask either water or air. The results show that the dose to the cells is more affected by the scattering conditions surrounding the flasks than by the level of filling within the flask. It is recommended that water or a water equivalent phantom material is used to surround the flasks or plates to approximate full scatter conditions at the cell layer. However for modulated fields, surrounding the 24 well plates with water-equivalent material is inadequate because of the large volume of air surrounding individual wells. Our results stress the importance of measuring the dose for new experimental configurations.

Mechanisms for covalent immobilization of horseradish peroxidase on ion-beam-treated polyethylene.

The surface of polyethylene was modified by plasma immersion ion implantation. Structure changes including carbonization and oxidation were observed. High surface energy of the modified polyethylene was attributed to the presence of free radicals on the surface. The surface energy decay with storage time after treatment was explained by a decay of the free radical concentration while the concentration of oxygen-containing groups increased with storage time. Horseradish peroxidase was covalently attached onto the modified surface by the reaction with free radicals. Appropriate blocking agents can block this reaction. All aminoacid residues can take part in the covalent attachment process, providing a universal mechanism of attachment for all proteins. The native conformation of attached protein is retained due to hydrophilic interactions in the interface region. The enzymatic activity of covalently attached protein remained high. The long-term activity of the modified layer to attach protein is explained by stabilisation of unpaired electrons in sp(2) carbon structures. A high concentration of free radicals can give multiple covalent bonds to the protein molecule and destroy the native conformation and with it the catalytic activity. The universal mechanism of protein attachment to free radicals could be extended to various methods of radiation damage of polymers.

Scintillation dosimeter arrays using air core light guides: simulation and experiment.

The performance of a scintillation dosimeter that uses a silvered air core light guide is examined by Monte Carlo (MC) simulations and by experiment to determine its suitability for array dosimetry in external beam radiotherapy. The air core light guide avoids the generation of the Cerenkov background that is produced in a conventional optical fibre. MC simulations using a 6 MV photon beam showed that silver thicknesses of less than 1 microm compensated for the effects of the other material components, to give the dosimeter water equivalence within 0.5%. A second dosimeter located adjacent to the primary dosimeter in any direction affected the dose measurement by less than 1.5%, when the centre-to-centre spacing was 1.3 mm or greater. When the dosimeter array is located perpendicular to the beam central axis, with a spacing of 2.5 mm, the calculated deviation from the dose deposited in water was less than 2%. When the dosimeter array is located parallel to the beam central axis with a spacing of 10 mm, the calculated dose readings deviated from water by less than 2.5%. The simulation results were confirmed with experiment for two neighbouring dosimeters and a small densely packed array. No proximity effects were measured within the experimental error of +/-1.5%. These results confirm the dosimetric accuracy of the air core dosimeter design without the need for correction factors. The dosimeter has excellent potential for use in arrays.

Light-scattering-induced artifacts in a complex polymer gel dosimetry phantom.

Certain polymer gels become turbid on exposure to ionizing radiation, a property exploited in medical dosimetry to produce three-dimensional dose maps for radiotherapy. These maps can be read using optical computed tomography (CT). A test phantom of complex shape ("layered tube") was developed to investigate the optical properties of polymer gel dosimeters when read using optical CT. Extinction coefficient profiles from tomographically reconstructed slices of the phantom exhibited several artifacts. A simple model invoking scattered light in the gel was able to account for all artifacts, which in a real dosimeter may have been mistaken for other phenomena, resulting in incorrect readings of dose.

Initial investigation of a novel light-scattering gel phantom for evaluation of optical CT scanners for radiotherapy gel dosimetry.

There is a need for stable gel materials for phantoms used to validate optical computerized tomography (CT) scanners used in conjunction with radiation-induced polymerizing gel dosimeters. Phantoms based on addition of light-absorbing dyes to gelatine to simulate gel dosimeters have been employed. However, to more accurately simulate polymerizing gels one requires phantoms that employ light-scattering colloidal suspensions added to the gel. In this paper, we present the initial results of using an optical CT scanner to evaluate a novel phantom in which radiation-exposed polymer gels are simulated by the addition of colloidal suspensions of varying turbidity. The phantom may be useful as a calibration transfer standard for polymer gel dosimeters. The tests reveal some phenomena peculiar to light-scattering gels that need to be taken into account when calibrating polymer gel dosimeters.