The document summarizes the development and validation of a hyperspectral imaging microscopy system called Xanoscope for pathology applications. Key points: 1) Xanoscope allows for fast, automated acquisition of hyperspectral images over contiguous wavelength bands with high spatial resolution, reducing scan times significantly compared to previous methods. 2) Validation experiments showed Xanoscope produces quantitative measurements and improves signal-to-noise ratio through optimized camera settings and image accumulation. 3) Xanoscope was used to scan 10 multiplexed fluorophores in cell lines, and linear unmixing was able to measure the individual contribution of each fluorophore at pixel-level for advanced pathological analysis.

1. M.S Thesis Defense 24 th November 2008 Dipen Rana Graduate Student Translational Research, Garner Lab UT Southwestern Medical Center at Dallas Integration of Hyperspectral Imaging Microscopy for Pathology Applications

2. Outline

4. Fluorescence: The Basic Principle 1= absorbance of light (excitation) 2= vibrational relaxation (stokes shift ) 3= emission of light http://www.olympusmicro.com/primer/techniques/fluorescence/introduction.html; invitrogen Interactions of molecules with light

6. Fluorophores – Cell Labeling

7. Fluorophores – Cell Labeling

13. Hyperspectral Imaging Microscopy (HMI) System Workstation Overview

14. Outline

16. Outline

17. Hyperspectral Imaging Microscopy System

19. Hyperspectral data cube provides 3 dimensional data Collecting a stack of Y- λ images by moving the stage incrementally in X direction creates a 3 dimensional “hyperspectral data cube”

20. Efficiency of Camera and spectrograph for generating high resolution hyperspectral data Slit Width

22. Outline

24. Architecture of Xanoscope

25. Initializing Xanoscope with default or last saved parameters

26. Adjusting the parameters of individual components for dynamic control

28. Controlling resolution to improve SNR and frame readout speed

29. Controlling resolution to improve SNR and frame readout speed 1X1 binning 1536X1024 pixels 2X2 binning 768X512 pixels 2X2 SNR = 4* Signal/ sqrt(4*Noise) = 2 [Signal/sqrt(Noise)] 1X1 SNR = Signal/ sqrt(Noise)

30. Custom settings for scan parameters

31. Sample scan preview along with scan information & Image Intensity values for quick reference

32. Sample scan preview along with scan information & Image Intensity values for quick reference

33. Saving the raw data files at user defined location for further analysis Raw data files are in binary format

34. Composite hyperspectral image cube build using raw data files along with quick analysis from composite image

35. Point spectrum and row spectrum for quick review of composite images

36. Outline

37. Determination of Spatial Resolution of Xanoscope images

39. Validating CCD linearity for quantitative measurements made using Xanoscope

40. Validating the gain settings for Xanoscope

41. Validating accumulations used in Xanoscope for improving Signal to noise ratio in low light imaging

42. Validating accumulations used in Xanoscope for improving Signal to noise ratio in low light imaging mimicking smoothness First image with low SNR ratio Accumulated images with smoothness effect

43. High precision control over stage to avoid overlapping of images 1 2 3 4 5 Image is captured at 80µm slit width and 40X objective lens Minimum stage step resolution= 0.2µm Total number steps moved for each image= (80/40)/0.2= 10 steps Total number steps moved for 5 images = 50 steps = 50 * 0.2 =10 μ m each division = 10µm

44. Performance Evaluation of Xanoscope by scanning 10 multiplexed fluorophores in different cell lines Samples provided by Uhr lab at Cancer immunobiology center, UTSW MCF7 – breast cancer Daudi- Human Burkitt's lymphoma cell line TP – patient breast cancer cells Normal breast cells from breast cancer reduction

45. Linear unmixing of multiplexed fluorophores to measure its individual contribution …………………………………………………… .2

46. Analysis of multiplexed fluorophores by linear unmixing of its overlapping spectra Standard emission spectrum of 10 fluorophores Contributions of each fluorophore at a pixel in an image

47. Visualization of individual contribution of 10 fluorophore-antibody conjugates in cell lines & touch preps

48. Statistical analysis of individual contribution of 10 fluorophore-antibody conjugates in cell lines & touch preps