This Knighton_readme_20210511.txt file was generated on 20210511 by Brian Cottle ------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset Spectral data for: Towards Intraoperative Quantification of Atrial Fibrosis Using Light-Scattering Spectroscopy and Convolutional Neural Networks 2. Author Information Principal Investigator Contact Information Name: Frank Sachse Institution: University of Utah Address: The Nora Eccles Harrison CVRTI 95 S 2000 E, Room 204A, Salt Lake City, UT 84112-5000 Email: frank.sachse@utah.edu Associate or Co-investigator Contact Information Name: Robert Hitchcock Institution: University of Utah Address: 36 S Wasatch Drive, SMBB 4509, Salt Lake City, UT 84112 Email: r.hitchcock@utah.edu Alternate Contact Information Name: Nathan Knighton University of Utah Address: 201 Presidents' Cir, Salt Lake City, UT 84112 Email: nate.knighton@utah.edu Authors: Nathan J. Knighton Brian K Cottle Email: brian.cottle@utah.edu Institution: University of Utah Address: The Nora Eccles Harrison CVRTI 95 S 2000 E, Room 204, Salt Lake City, UT 84112-5000 Bailey E.B. Kelson Robert W. Hitchcock Frank B. Sachse 3. Date of data collection (single date, range, approximate date) 20190101 to 20190208, 20200721 to 20200807 4. Geographic location of data collection (where was data collected?): Salt Lake City, UT 5. Information about funding sources that supported the collection of the data: National Institutes of Health National Institutes of Health (NIH)9000 Rockville Pike Bethesda, Maryland 20892; telephone 301-496-4000 Grant Number:R56 HL128813 Grant Number:R01 HL135077 Nora Eccles Treadwell Foundation 136 EAST SOUTH TEMPLE STE 2100 SALT LAKE CITY, UT 84111 -------------------------- SHARING/ACCESS INFORMATION -------------------------- 1. Licenses/restrictions placed on the data: CC BY NC - Allows others to use and share your data non-commercially and with attribution. 2. Links to publications that cite or use the data: Knighton, N. J., Cottle B. K., Kelson B. E. B., Hitchcock R. W., Sachse F. B., 2021 Towards Intraoperative Quantification of Atrial Fibrosis Using Light-Scattering Spectroscopy and Convolutional Neural Networks. Sensors (21)18, 6033; https://doi.org/10.3390/s21186033. Note:This article belongs to the Special Issue Optical Imaging and Biophotonic Sensors (OIBS)) 3. Links to other publicly accessible locations of the data: 4. Links/relationships to ancillary data sets: 5. Was data derived from another source? No If yes, list source(s): 6. Recommended citation for the data: Knighton, N. J., Cottle B. K., Kelson B. E. B., Hitchcock R. W., Sachse F. B., 2020. Spectral data for: Towards Intraoperative Quantification of Atrial Fibrosis Using Light-Scattering Spectroscopy and Convolutional Neural Networks. The Hive: University of Utah Research Data Repository.https://doi.org/10.7278/S50D-3Q4J-SC4Y --------------------- DATA & FILE OVERVIEW --------------------- 1. File List (spectral_data_heart.zip) The spectral_data_heart.zip file contains the following folders: A. Folder/Filename: Depth Short description: Raw spectral data for depth analysis, including one .csv file containing wavelength information. B. Folder/Filename: VolumeFraction Short description: Raw spectral data for volume fraction analysis, including one .csv file containing wavelength information. C. Folder/Filename: Permutations Short description: Raw spectral data for Permutations analysis, including one .csv file containing wavelength information. 2. Relationship between files: All raw data files are comma delimited text files (.csv). Each folder has .csv files that are constructed and named in a similar manner, describing the fiber combination used as well as whether the data were used for testing or training purposes. 3. Additional related data collected that was not included in the current data package: None 4. Are there multiple versions of the dataset? No If yes, list versions: Name of file that was updated: i. Why was the file updated? ii. When was the file updated? Name of file that was updated: i. Why was the file updated? ii. When was the file updated? -------------------------- METHODOLOGICAL INFORMATION -------------------------- 1. Description of methods used for collection/generation of data: We adapted the experimental approach for use with fixed canine tissue constructs created by placing thin tissue sections of myocardium or aortic tissue on top of each other. Tissue layout within constructs was dependent on the classification objective. Constructs were described numerically by their composition of myocardium and aortic tissue, and read left-to-right with the leftmost value being the top layer of the tissue constructs. “0” and “1” identified myocardial and aortic sections, respectively. Thus, ‘00000000’ is a construct of only myocardium, ‘11110000’ a construct of aortic tissue in the top four layers with myocardium in the bottom four layers, and ‘11111111’ a construct of only aortic tissue. Class names were based on these descriptors. Spectra were measured for 30 ms. Every measurement gathered a single spectra from each of the five collection fibers. The probe was placed at various locations and orientations on the construct between each measurement. Ten measurements were obtained from a construct whereupon the construct was disassembled and a new construct of the same layout was built using new tissue sections. The measurement process was repeated on the new constructs. Extreme outliers were removed. 2. Methods for processing the data: Raw data were sorted by class for each of the three experiments (Depth, Permutations, and VolumeFraction), and were also sorted by collection fiber. These dataset were then saved into 30 different .csv files for each experiment. 15 of these .csv files contain training data for the neural networks (one for each of the potential fiber combinations), and 15 contain testing data for the neural networks (also one for each of the potential fiber combinations). Each of the three folders, one for each experiment, also contain one .csv file that includes the wavelength data for all the spectra in that experiment. Each spectra was also normalized to the mean value 3. Instrument- or software-specific information needed to interpret the data: Software to load, view, or manipulate .csv files. MATLAB 2019 or newer or a Python environment 3.8 or newer recommended. 4. Standards and calibration information, if appropriate: None 5. Environmental/experimental conditions: None 6. Describe any quality-assurance procedures performed on the data: None 7. People involved with sample collection, processing, analysis and/or submission: N. Knighton B. Cottle B. Kelson --------------------------------------------------- DEPTH: Depth_Train_*.csv and Depth_Test_*.csv files --------------------------------------------------- 1. Number of variables: 6 2. Number of cases/rows: All numbers below are described using zero-based indexing. Rows are individual samples. ————————————————— For all “.csv” files: Column 0: Class to which the spectrum are associated. Binary descriptions are as outlined in the paper. 0 = 00000000 1 = 00000011 2 = 00001100 3 = 00110000 4 = 11000000 5 = 11111111 ————————————————— For filenames “R1.csv”, “R2.csv”,”R3.csv”,”R4.csv”,”R5.csv”: Columns 1-3600: Spectral measurements from the spectrometers ————————————————— For filenames involving combinations of “R1.csv”, “R2.csv”,”R3.csv”,”R4.csv”,”R5.csv”: Columns 1 through 3599: Spectral measurements from the first of the two listed combinations Columns 3600 through 7199: Spectral measurements from the second of the two listed combinations 3. Variable List A. Name: [00000000] Description: [All myocardium stack] B. Name: [11111111] Description: [All Aortic Stack] “0” and “1” identified myocardial and aortic sections, respectively. Thus, [00000000] is a construct of only myocardium, [11110000] a construct of aortic tissue in the top four layers with myocardium in the bottom four layers, and [11111111] a construct of only aortic tissue, and so forth. 4. Missing data codes: N/A 5. Specialized formats of other abbreviations used N/A ------------------------------------------------------------------------------- VOLUME FRACTION: VolumeFraction_Train_*.csv and VolumeFraction_Test_*.csv files ------------------------------------------------------------------------------- 1. Number of variables: 9 2. Number of cases/rows: All numbers below are described using zero-based indexing. Rows are individual samples. ————————————————— For all “.csv” files: Column 0: Class to which the spectrum are associated. Binary descriptions are as outlined in the paper. 0 = 00000000 1 = 01000000 2 = 01100000 3 = 01110000 4 = 01111000 5 = 01111100 6 = 01111110 7 = 01111111 8 = 11111111 ————————————————— For filenames “R1.csv”, “R2.csv”,”R3.csv”,”R4.csv”,”R5.csv”: Columns 1-3600: Spectral measurements from the spectrometers ————————————————— For filenames involving combinations of “R1.csv”, “R2.csv”,”R3.csv”,”R4.csv”,”R5.csv”: Columns 1 through 3599: Spectral measurements from the first of the two listed combinations Columns 3600 through 7199: Spectral measurements from the second of the two listed combinations 3. Variable List A. Name: [00000000] Description: [All myocardium stack] B. Name: [11111111] Description: [All Aortic Stack] “0” and “1” identified myocardial and aortic sections, respectively. Thus, [00000000] is a construct of only myocardium, [11110000] a construct of aortic tissue in the top four layers with myocardium in the bottom four layers, and [11111111] a construct of only aortic tissue, and so forth. 4. Missing data codes: N/A 5. Specialized formats of other abbreviations used N/A ------------------------------------------------------------------------ PERMUTATIONS: Perumtations_Train_*.csv and Perumtations_Test_*.csv files ------------------------------------------------------------------------ 1. Number of variables: 6 2. Number of cases/rows: All numbers below are described using zero-based indexing. Rows are individual samples. ————————————————— For all “.csv” files: Column 0: Class to which the spectrum are associated. Binary descriptions are as outlined in the paper. 0 = 00000000 1 = 00001111 2 = 00110011 3 = 11001100 4 = 11110000 5 = 11111111 ————————————————— For filenames “R1.csv”, “R2.csv”,”R3.csv”,”R4.csv”,”R5.csv”: Columns 1-3600: Spectral measurements from the spectrometers ————————————————— For filenames involving combinations of “R1.csv”, “R2.csv”,”R3.csv”,”R4.csv”,”R5.csv”: Columns 1 through 3599: Spectral measurements from the first of the two listed combinations Columns 3600 through 7199: Spectral measurements from the second of the two listed combinations 3. Variable List A. Name: [00000000] Description: [All myocardium stack] B. Name: [11111111] Description: [All Aortic Stack] “0” and “1” identified myocardial and aortic sections, respectively. Thus, [00000000] is a construct of only myocardium, [11110000] a construct of aortic tissue in the top four layers with myocardium in the bottom four layers, and [11111111] a construct of only aortic tissue, and so forth. 4. Missing data codes: N/A 5. Specialized formats of other abbreviations used N/A