Data from: Non-invasive estimation of absorbed ionizing radiation dose in mice using Near-Infrared Spectroscopy (NIRS) and aquaphotomics
Data files
Apr 19, 2024 version files 3.57 MB
-
Radiation_data.csv
3.56 MB
-
README.md
9.32 KB
Abstract
Accurate measurement of ionizing radiation exposure, whether therapeutic or accidental, is of utmost importance in various scenarios. This paper presents a study that addresses this critical need by utilizing near-infrared (NIR) spectroscopy and aquaphotomics to estimate radiation dose exposure in mouse models subjected to X-ray irradiation. The analysis of NIR spectra acquired from the mouse abdomen enabled non-invasive estimation of radiation doses ranging from 0.5 to 6.5 Gy, immediately following the irradiation exposure. The findings were consistent with the impact of total body irradiation in mice, as evidenced by measures such as animal survival rate, alterations in body weight observed over a 30-day post-exposure period, and changes in hematocrit levels. The spectroscopic measurements were based on detecting changes in the molecular structure of body water after radiation exposure, utilizing the water spectral pattern as a multidimensional biomarker. While further validation in nonhuman primates is necessary, the findings demonstrate a simple, non-destructive, and rapid method that holds promise for the estimation of radiation exposure across a range of doses, applicable to both clinical applications and catastrophic radiation events. These advancements in radiation dose quantification have significant implications for the timely and precise assessment of radiation exposure in humans.
README: Non-Invasive Estimation of Absorbed Ionizing Radiation Dose in Mice Using Near-Infrared Spectroscopy (NIRS) and Aquaphotomics
https://doi.org/10.5061/dryad.vq83bk41s
The dataset contains the near-infrared spectra of mouse abdomen acquired with the purpose of assessing and monitor the effects of exposure to ionizing irradiation. The experiments were conducted with female mice categorized into six groups, each exposed to specific levels of irradiation, ranging from 0 Gy (control) to 6.5 Gy (lethal dose). The study involved a testing window with longitudinal follow-ups over a one-month period, covering immediate post-exposure assessments (0h to 72h) and subsequent testing every 5 days until the one-month follow-up period concluded. After a week of acclimatization, 8-week-old mice were randomly divided into 6 groups with 8 mice per group (48 mice in total). Each mice group was then subjected to varying total body X-rays irradiation in doses of 0 (control group), 0.5, 2, 4, 6, and 6.5 Gy (160 kVp, 3mA, 1.0 mm aluminium filter) at a dose rate of 0.622 Gy/min using an X-ray generator MX-160Labo (MediXtec, Tokyo, Japan) with a distance of 300 mm between the focus and target. NIR spectra acquisition occurred at specific time points: prior to irradiation, after one week of acclimatization, immediately post-irradiation (0 h), and at 6, 12, 24, 48, and 72 hours, followed by subsequent measurements every 5 days until 30 days post-irradiation. At each measurement time, the spectra were collected from mouse abdomen using a portable, hand-held microNIR spectrometer (Viavi Solutions, Santa Rosa, CA, USA) in the spectral range 908 – 1670 nm, with approximately 7 nm resolution step. Five consecutive spectra were acquired for each mouse, at each time when the measurements were performed. The acquired dataset was pre-processed using extended multiplicative scatter correction and analyzed using multivariate data analysis as described in the research article.
Description of the data and file structure
The dataset is provided in Excel file Radiation_data, all the data used in the analysis is provided in Sheet1. The dataset starts with 14 columns where names, class and dependent variables are provided, followed by near infrared spectra. The first row in the dataset provides names of the variables (columns 1 to 14), followed by the wavelengths at which the absorbance was measured (starting from 908.1, 914.294 etc.). The wavelengths are expressed in nm, and there is in total 125 wavelengths at which the absorbance was measured in the range from 908.1 nm to 1676.2 nm.
The first 14 columns contain class (descriptive, categorical variables, starting with C_) and dependent variables (measured, continuous variables, starting with Y_) contain the following information:
C_name: the name assigned to the spectrum during acquisition of spectra using the MicroNIR 1700 ES. The name was given using VIAVI MicroNIR Pro spectral acquistion software. The name of the spectrum contains a string of characters and numbers that have a meaning. For example the name of the spectra C1M1_4_20230117094827 means the following:
C1: the mice group 1 (each group contained 8 mice in total) out of 6 groups of mice that will undergo irradiation exposure to one of the following levels: C1 - O Gy (control group), C2 - 0.5 Gy, C3 - 2 Gy, C4 - 4 Gy, C5 - 6 Gy, and C6 - 6.5 Gy. The group number corresponds to the level of irradiation.
M1: the mice number; in each of the groups from C1 to C6 there were 8 mice in total, and the number next to M corresponds to a one of eight mice, it is unique for each mice in the group.
So, the meaning of the C1M1_4_20230117094827 would be: mice group 1 (control group, irradiation level 0 Gy), mouse number 1 from this group, 4th consecutively measured spectrum from the mouse abdomen, acquired on 17.01.2023 (the number after the date 20230117 represents exact time stamp at which the spectrum was acquired)
C_MOUSE: class variable that can have values from 1 to 8. There were six groups of mice, each containing 8 mice in total.
C_GyGroup: class variable that can have values from 1 to 6. The values 1 to 6 correspond to the level of exposure to irradiation: 1 - 0 Gy (control group), 2 - 0.5 Gy, 3 - 2 Gy, 4 - 4 Gy, 5 - 6 Gy and 6 - 6.5 Gy.
C_Date: class variable that provides information about the day when the spectrum has been acquired. It follows the format year/month/day. For example 230117 would mean that the spectrum was acquired 2023/01/17 - 17th January 2023.
C_EXPERIMENT: class variable that can have values from 1 to 13, describing when with respect to irradiation the spectrum was acquired. The meaning of the values 1 to 13 is as follows. 1 - before the irradiation (mice arrived to the experimental location and were given one week acclimatization time), 2 - immediately after irradiation (0h), 3 - 6 hours after irradiation exposure, 4 - 12 hours after irradiation exposure, 5 - 24 hours after irradiation exposure, 6 - 48 hours after irradiation exposure, 7 - 72 hours after irradiation exposure, 8 - 5 days after irradiation exposure, 9 - 10 days after irradiation exposure, 10 - 15 days after irradiation exposure, 11 - 20 days after irradiation exposure, 12 - 25 days after irradiation exposure and 13 - 30 days after irradiation exposure.
Y_Hours: dependent (continuous, numerical) variable: describes the number of hours passed after the irradiation. For the spectra collected during C_EXPERIMENT = 1 and C_EXPERIMENT = 2, Y_Hours is the same and equal to zero, because it was before and immediately after the irradiation (0 h). Caution should be taken during time analysis, that only the data corresponding to C_EXPERIMENT = 2 to 13 correspond to actual number of hours passed since irradiation (0h, 6h, 12 h etc.)
Y_GyDose: dependent, continuous numerical variable providing the information about irradiation dose (expressed in Gy) to which the mice have been exposed to, and can have values 0 Gy, 0.5 Gy, 2 Gy, 4 Gy, 6 Gy and 6.5 Gy.
Y_conSNr: dependent variable describing the order of consecutive spectra acquired during one continuous use of the instrument. The instrument was placed on the mouse abdomen, and without any movement or removal of the sensor from the abdomen, 5 consecutive measurements were performed, providing 5 spectra. The Y_conSNr variable can therefore have values from 1 to 5.
Y_Temp: dependent variable, the internal temperature of the instrument (sensor) during spectral acquisition. These values are also provided together with the spectral absorbance, by the instrument itself. The temperature is expressed in degree of celsius (°C).
Y_Weight: dependent, continuous numerical variable, measured body weight of each mouse prior to spectral measurements on a particual day. The weight is expressed in grams (g).
Y_Hematocrit: dependent, continuous numerical variable, measured hematocrit levels in serum (%). This variable only has values for the last experiment C_EXPERIMENT = 13, because on the 30th day the surviving mice were anaesthetized and sacrificed. The missing values are marked by "*". There are no measurement of hematocrit that can be matched to the measured spectra.
Y_SampleNr: dependent variable providing the ID number of a particular measurement consisting of 5 consecutively acquired spectra, performed on a particular animal, from a particular group, in a specific point in time
The columns that follow first 14 columns contain independent variables i.e. spectral measurements. Each spectrum is given in one row of data starting from the column M to column EG, where each column corresponds to a particular wavelength of light, and each cell represents the value of absorbance measured at that particular wavelenght.
Sharing/Access information
Data is also available as the Supplementary Material provided with the Manuscript.
Code/Software
The analysis of the submitted Dataset was performed using R version 4.3.1 (2023-06-16 ucrt) -- "Beagle Scouts" Copyright (C) 2023 The R Foundation for Statistical Computing, Platform: x86_64-w64-mingw32/x64 (64-bit) and a package aquap2. For the code to be used, a user needs to download R software from this link https://cran.r-project.org/bin/windows/base/o and install aquap2 package that can be downloaded from GitHub https://github.com/bpollner/aquap2 or from the following link https://www.aquaphotomics.com/aquap2/. The last link contains detailed instructions for installation and running aquap2. The three scripts provided are called Routine, Metadata and Anproc. Routine script contains script with the code for the analysis. The user should first run two lines of this script which will generate folder structure with several folders. The scripts Metadata and Anproc should be copied to the Folder called Metadata which will be created when the first two lines of Routine script are ran. Each of the provided scripts contain comments explaining what each command line means. For more help regarding the use of aquap2, refer to the Documentation for package ‘aquap2’.
Methods
Near infrared spectra acquisition was performed during a period of 30 days, at specific time points: prior to ionizing irradiation (after one week of acclimatization to laboratory conditions), immediately post-irradiation (0 h), and at 6, 12, 24, 48, and 72 hours, followed by subsequent measurements every 5 days until 30 days post-irradiation. At each measurement time, the spectra were collected from mouse abdomen using a portable, hand-held microNIR spectrometer (Viavi Solutions, Santa Rosa, CA, USA) in the spectral range 908 – 1670 nm, with approximately 7 nm resolution step. Five consecutive spectra were acquired for each mouse, at each time when the measurements were performed.
The spectral data were analyzed using multivariate analysis methods and preprocessed using extended multiplicative scatter correction.