READ ME File For 'Dataset for Doctoral Thesis "Investigating Biological Optical Transparency Windows in the Near and Shortwave Infrared for Diagnosis and Therapy"' Dataset DOI: 10.5258/SOTON/D3521 ReadMe Author: HIROKI COOK, University of Southampton, ORCID ID 0009-0003-8984-0021 This dataset supports the thesis entitled 'Investigating Biological Optical Transparency Windows in the Near and Shortwave Infrared for Diagnosis and Therapy' AWARDED BY: University of Southampton DATE OF AWARD: 2025 Date of data collection: 2020-2024 Information about geographic location of data collection: Southampton, UK Licence: CC BY Related projects/Funders: Mahajan - Transformative Healthcare 2050 -------------------- DATA & FILE OVERVIEW -------------------- This dataset contains: Dataset for Chapter 5/ ├── Data for Figure 5_3 Explained Variance.xlsx # Numerical data describing effect of increasing principal components on variance ├── Data for Figure 5_4 Cartilage Vibrational Spectra.xlsx # Averaged spectral data for cartilage samples ├── Data for Figure 5_5 Diffuse vs Transreflected Spectra.xlsx # Averaged spectral data for cartilage on different substrates ├── Data for Figure 5_6 Table 5_2 Table 5_3 PCA LDA.xlsx # Numerical data from Principal Component Analysis - Linear Discriminant Analysis ├── Data for Figure 5_7 Table 5_4 Table 5_5 SVM.xlsx # Numerical data from Support Vector Machine analysis └── Data for Figure 5_8 Figure 5_9 Spectromics Biomarkers.xlsx # Numerical data from statistical tests performed on spectral data Dataset for Chapter 6/ ├── Data for Figure 6_9 Bone Core Vibrational Spectra.xlsx # Spectral data from along bone core samples │ ├── Data for Figure 6_10 F77 OP Core Hyperspectral Scan/ │ ├── Archive F77 OP NIRSWIR 5283 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 5283 cm^-1 │ ├── Archive F77 OP NIRSWIR 7156 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 7156 cm^-1 │ ├── Archive F77 OP Raman 960 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 960 cm^-1 │ ├── Archive F77 OP Raman 1302 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 1302 cm^-1 │ ├── Archive F77 OP Raman 1445 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 1445 cm^-1 │ ├── Archive F77 OP Raman 1649 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 1649 cm^-1 │ ├── Python Code Output for F77 OP Core/ # Spectral data (graph images and numerical data) and microscopy images output from hyperspectral scan │ └── Bone Core F77 OP Photogrammetry Model.obj # 3D photogrammetry model │ ├── Data for Figure 6_11 F57 OA Core Hyperspectral Scan/ │ ├── Archive F57 OA NIRSWIR 5283 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 5283 cm^-1 │ ├── Archive F57 OA NIRSWIR 7156 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 7156 cm^-1 │ ├── Archive F57 OA Raman 960 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 960 cm^-1 │ ├── Archive F57 OA Raman 1302 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 1302 cm^-1 │ ├── Archive F57 OA Raman 1445 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 1445 cm^-1 │ ├── Archive F57 OA Raman 1649 cm-1 Processed/ # Spectral intensity data forming 2D slices (TIF) of 3D hyperspectral data set at 1649 cm^-1 │ ├── Python Code Output for F57 OA Core/ # Spectral data (graph images and numerical data) and microscopy images output from hyperspectral scan │ └── Bone Core F57 OA Photogrammetry Model.obj # 3D photogrammetry model │ ├── Data for Figure 6_12 F77 OP Cartilage vs Bone.xlsx # Spectral data from along bone core sample ├── Data for Figure 6_13 Sliced and Natural Cartilage.xlsx # Spectral data from cartilage samples │ ├── Data for Figure 6_14 F65 OA Femoral Head Hyperspectral Scan/ │ ├── Archive F65 OA NIRSWIR 5346 cm-1/ # Spectral intensity data forming 2D slices (PNG) of 3D hyperspectral data set at 5346 cm^-1 │ ├── Archive F65 OA NIRSWIR 6435 cm-1/ # Spectral intensity data forming 2D slices (PNG) of 3D hyperspectral data set at 6435 cm^-1 │ ├── Archive F65 OA NIRSWIR 7172 cm-1/ # Spectral intensity data forming 2D slices (PNG) of 3D hyperspectral data set at 7172 cm^-1 │ ├── Archive F65 OA NIRSWIR 8648 cm-1/ # Spectral intensity data forming 2D slices (PNG) of 3D hyperspectral data set at 8648 cm^-1 │ ├── Python Code Output for F65 OA Femoral Head/ # Spectral data (graph images and numerical data) and microscopy images output from hyperspectral scan │ └── Femoral Head OA Photogrammetry Model.glb # 3D photogrammetry model │ └── Data for Figure 6_15 Fiducial Markers.xlsx # Spectral data from femoral head cartilage and polystyrene markers Dataset for Chapter 7/ ├── Data for Figure 7_3 MOPA Characterisation.xlsx # Numerical data from spectral characterisation of master oscillator power amplifier (MOPA) ├── Data for Figure 7_7 Onion Cells.xlsx # Numerical data describing affected area (AA) and thermally altered area (TAA) with changing pulse parameters ├── Data for Figure 7_8 Onion Cells.xlsx # Numerical data describing AA and TAA, and subsequent ablation efficiency (AE) with changing pulse parameters ├── Data for Figure 7_9 Onion Cells.xlsx # Numerical data describing AA and TAA with changing pulse paramaters ├── Data for Figure 7_10 Fig 7_11 Fig 7_12 Fig 7_13 Neuroblastoma Cells.xlsx # Numerical data describing AA with changing pulse parameters ├── Data for Figure 7_14 Neuroblastoma Cells.xlsx # Numerical data describing AA with changing pulse parameters └── Data for Figure 7_15 Figure 7_16 Single Neuroblastoma Cell.xlsx # Numerical data describing AA under ideal pulse parameters -------------------------- METHODOLOGICAL INFORMATION -------------------------- All methodology described in detail in Thesis titled "Investigating Biological Optical Transparency Windows in the Near and Shortwave Infrared for Diagnosis and Therapy" Method for "Dataset for Chapter 5" can also be found in Cook, Hiroki, et al. ‘Holistic Vibrational Spectromics Assessment of Human Cartilage for Osteoarthritis Diagnosis’. Biomedical Optics Express, vol. 15, no. 7, July 2024, p. 4264. DOI: 10.1364/BOE.520171 Method for "Dataset for Chapter 7" can also be found in Gerard, Matthew D., Cook, Hiroki, et al. ‘Single-Cell High-Precision Ablation Using Nanosecond-Pulsed Thulium-Doped Fiber Laser’. Optical Engineering, vol. 63, no. 08, Aug. 2024. DOI: 10.1117/1.OE.63.8.086102. DATASET FOR CHAPTER 5: Raman spectral data was collected from cartilage samples with a Renishaw InVia microscope system with samples excited using a 785 nm laser focused through a Leica 50x (0.75 NA) short working distance (~200 µm) objective. Renishaw WiRE 4.1 software was used to collect data and set measuring parameters.The instrument was calibrated to the 520 cm^-1 peak of a silicon reference sample before each experiment, dark background signal subtracted, and cosmic rays removed after acquisition using WiRE. Near- and Shortwave Infrared (NIR-SWIR) spectral data was collected from cartilage samples with incident excitation light was provided by a broadband halogen lamp (HL-2000-FHSA-LL, Ocean Insight) and signal collected via an NIRQuest2.5+ (Ocean Insight) spectrometer. Both were coupled to a ferrule fibre optic reflectance probe (R400-7-VIS-NIR, Ocean Insight) with a profile of 6 annular fibres for excitation and 1 central fibre for collection. Two planoconvex uncoated lenses collimated and focused light onto sample, allowing for contact-less measurements. Spectral data underwent pre-processing transformations prior to classification via multivariate analysis, carried out in iRootLab (0.15.07.09-v) toolbox within MATLAB R2020a software (MathWorks). DATASET FOR CHAPTER 6: Spectral data was captured on a custom-built multimodal micro-spectroscopy system with motorised stage (XYZ) and rotational stage (theta), capable of 3D hyperspectral scans of tissue sample surfaces. Operating software was written in Python and using Micro-Manager (ver. 2.0.2 20230815)(see Thesis Supplementary Section). Raman laser excitation was delivered from a 785 nm continuous wave laser (Laser 2000), which was conveyed via a relay lens pair (both AC254-30-B-ML, Thorlabs) to the inverted microscope (Eclipse Ti-E, Nikon), then onto the reflective objective lens (50105-02, Newport) via a 785 nm dichroic beamsplitter (Semrock). Raman spectral data was collected via a crossed Czerny-Turner spectrograph and CCD science camera (SR-303i-A-SIL and DU420A-BR-DD, Andor Technology). Raman scattering signal intensity was recorded by the CCD in full vertical binning mode with an exposure time of 20s, calibrated with the 1004 cm-1 peak of a polystyrene standard sample. NIR-SWIR spectral data was collected using a broadband excitation light source (HL-2000-FHSA-LL, Ocean Insight) and spectrometer (NIRQuest2.5+, software ver. 2.0.16, Ocean Insight) coupled to a reflective off-axis collimator (RC08SMA-P01, Thorlabs) via a bifurcated fibre optic cable (BFY1000LS02, Thorlabs), directed through the reflective objective lens. NIR-SWIR absorption signal was collected with 10 ms exposure time, averaged over 1000 acquisitions, bright reference (I_0) captured from a gold mirror slide mounted on stage. Spectral acquisition and processing, including wavenumber calibration and dark background correction, was carried out via custom operating software written in Python (see Thesis Supplementary Section). Photogrammetry data was collected via PolyCam mobile app software, which enabled automatic reconstruction, and exported a full colour model and surface mesh. 3D hyperspectral data reconstruction was carried out in Python and with Dragonfly (Comet Technologies) for scaling and manual alignment with the photogrammetry mesh. DATASET FOR CHAPTER 7: Master oscillator power amplifier (MOPA) carried out by Matthew D. Gerard: Power slope efficiency measured using an Ophir power meter. Signal spectrum using an optical spectrum analyser. Optical pulse shape using a photodiode attached to an oscilloscope (Tektronix). Beam quality was measured using an IR scan-head with 75 mm planoconvex lens, observing how beam diverged as the scan-head was moved perpendicularly, hyperbolic curve fitted using Mathematica software to determine M^2 value. Remaining data collected jointly by Hiroki Cook and Matthew D. Gerard. Brightfield images (transmission geometry) were captured using a scientific camera (Digital Sight DS-Fi1c, Nikon) mounted to the microscope side port, with proprietary software (NIS Elements F, Nikon). Video frames capturing the ablation event were analysed to accurately track cell fate under laser exposure. An 1800 nm short pass filter (DMSP1800R, Thorlabs) reflected the 1950 nm ablating beam towards the sample whilst allowing shorter visible light to a mounted camera to observe ablation in real time. Measurement of affected areas on the cell samples was carried out using ImageJ software (FIJI), on images scaled using microscopy standards. This included the ablation area (AA), thermally altered area (TAA), and ablation efficiency (AE). Environmental/experimental conditions: Air-conditioned optics laboratory at 293 K; air-conditioned positive-pressure wet laboratory. People involved with sample collection, processing, analysis and/or submission: Hiroki Cook for "Dataset for Chapter 5" and "Dataset for Chapter 6" Hiroki Cook and Matthew D. Gerard for "Dataset for Chapter 7" -------------------------- DATA-SPECIFIC INFORMATION -------------------------- DATASET FOR CHAPTER 5: Variable list, defining any abbreviations, units of measure, codes or symbols used: Principal Components (PC) - unitless Explained Variance - unitless Raman Spectral Intensity - arbitrary units Near- and Shortwave Infrared (NIR-SWIR) Spectral Absorbance - arbitrary units Concatenated Spectral Intensity - arbitrary units Wavenumber - reciprocal centimetre (cm^-1) Wavelength - nanometre (nm) Principal Component Analysis - Linear Discriminant Analysis (PCA-LDA) - Mahalonobian distance, unitless Support Vector Machine Accuracy - percentage Quality Parameters: Accuracy, Sensitivity, Speciality, F-Score - percentage DATASET FOR CHAPTER 6: Variable list, defining any abbreviations, units of measure, codes or symbols used: Raman Spectral Intensity - arbitrary units NIR-SWIR Spectral Absorbance - arbitrary units Wavenumber - reciprocal centimetre (cm^-1) Coordinate Position, XYZ - millimetre (mm) Angular Position, theta - degrees DATASET FOR CHAPTER 7: Variable list, defining any abbreviations, units of measure, codes or symbols used: Average Power (P_avg) - Watts (W) Pulse Duration (tau) - nanoseconds (ns) Repetition Rate (R) - Hertz (Hz) Exposure Time (t_exp) - seconds (s) Position - millimetre (mm) Power - decibel milliwatt (dBm) Ablation Area (AA) - micrometres squared (um^2) Thermally Ablated Area (TAA) - micrometres squared (um^2) Ablation Efficiency - percentage Date that the file was created: June, 2025