# READ ME File For 'Dataset for "Far-field unlabeled super-resolution imaging with superoscillatory illumination"'

Dataset DOI: 10.5258/SOTON/D0598

ReadMe Author: *Edward Rogers, University of Southampton*, etfr@soton.ac.uk

This dataset supports the publication:
*Far-field unlabeled super-resolution imaging with superoscillatory illumination*
Edward T. F. Rogers, Shmma Quraishe, Katrine S. Rogers, Tracey A. Newman, Peter J. S. Smith, and Nikolay I. Zheludev
APL Photonics **5**, 066107 (2020); https://doi.org/10.1063/1.5144918


## Contents


This dataset contains the data underlying: figures 2, 4, 5, 6 and 7 in the main manuscript; suplementary figure 2S; and suplementary videos 1-4.

Data from simulations is provided as Matlab data (.mat) files. Imaging data captured exeprimentally is provided as tiff files, with raw captured data and processed data as presented in the manuscript provided.

For Matlab files, the relevant variables for each figure are given in the descriptions below.
For polarisation-contrast images, each folder contains multiple tiff images, with filenames in the form:
`Chan[AB]_0001_0001_[time coord]_[z coord]`. Each x and t pair has 8 files: files suffixed `0001-0004`, which are the raw captured data for each of the four polarisation settings; and four files suffixed `..._Azimuth`, `..._Diattenuation`, `..._MeanTrans` and `..._Diattenuation_CMProcessed`, which contain the processed data. The processed files are formated as follows:
- `..._Azimuth`: 16bit integer econding the axis along which most light is refelected, in the range 0-pi.
- `..._Diattenuation`: 16bit integer encoding the degree of direflection at each point between 0 (equal refelction of all polarisations) and 1 (complete supression of reflection at a specific angle)
- `..._MeanTrans`: the mean reflection over poalrisation direction from each point on the sample 
- `..._Diattenuation_CMProcessed`: a false colour image containing azimuth encoded in hue and direfelction in brightness, as described in the manuscript.

In the case of images in the paper, a saturated image (files prefixed `Snapshot_...`) are also included, with the data saturated as described in the manuscript. 

## Manuscript figures:

#### Figure 1 - Principles of superoscillatory microscopy 
Schematic. No underlying data

#### Figure 2 - Superoscillatory microscope creates superoscillatory images with sub-diffraction features

Matlab file `fig_2.mat`. All panels use the same variable, `x`, on the x axis. Panel a plots variable `A` (blue solid) and `P` (orange dashed) against `x`. Panel b plots `PSF` against `x`. Panel c plots `object` (black) and `convImgI` (orange dashed) against `x`. Panel d plots `object` (black) and `imgI` (blue) against `x`, with yellow bars at those points where `abs(k_local)./k_0 > 1`. 
	
#### Figure 3 - Microscopy with superoscillatory illumination vs STED microscopy
Schematic. No underlying data.	

#### Figure 4 - Engineering of superoscillatory hotspots

The data underlying the simulation results are provided in the Matlab file fig_4.mat. The target widths and fields of view (both in units of wavelength) are stored in the variables widths and isos (=1) respectively. The images are rotationally symmetric and so are provided as a function of radius (r) with intenisty stored in the variable I. I is 10001x4, where the 10001 dimensions is the r directions, 5 is the number of widths. Each image is built up by rotataing the the 1D data around the central axis.

#### Figure 5 - Imaging the Siemens star.
- For panel a, all the data are in the Matlab file `fig 5 a siemens_star_pswf_tgt_width=0.4_iso=1_separ=10000_spher=0_astig=0.mat`
	
    - `x_img` and `y_img` contain the coordinates (in metres)of the sample, illumination and brightfield images.
    - The sample (i) is stored in `sample` 
    - The brightfield intensity (ii) is stored in `conventional_image`.
    - The illumination spot intensity (not shown) is stored in `I_spot`
		
    - The superoscillatory image (iii) is stored in `so_image`, on a different coordinate grid, stored in `so_x` and `so_y`.
		
- For panel b (i), the data is in `fig 5 b i SEM.jpg`
- For panel b (ii), the data is in `fig 5 b ii siemens_star 1.tif`
- For panel b (iii), the data is in the folder `fig 5 b iii ss 5deg cpswf 001a` in the standard format
	
- For panel c fig 5 c the data are in `siemens_star_comparison_data_20191025_133539.mat`. The experimental data for the superoscillatory and brightfield lines are stored in the Matlab structs `bf` and `so`, with the simulated data in `bf_sim` and `so_sim` respectively. For each line, the lower x-axis (spatial_frequency) is given is the field `pitch_axis` (accessed as e.g. `bf.pitch_axis`), and the (un-normalised) y data is given in  the field `SBR`. The upper x-axis is given by `1/(pitch_axis * 1e-3)`. The threshold level (black dotted line) is stored in `SBR_threshold` and the vertical dashed and dotted lines in the struct fields `res_SBR`.

#### Figure 6 - Superoscillatory imaging of biological structures
- Panel a: Captured widefield image and the cropped and processed version shown in the manuscript.
- Panel b: Captured files in the standard format. 
- Panel c: Captured files in the standard format. 
- Panel d: Matlab file res_test_data_20171106_123748.mat. x and y coordinates of the data are contained in the cell arrays `d_out` and `I_out`, respctively. Each line has separate x, `d_out(i)`, and y coordinates, `I_out(i)`, for `i = 1, 2, 3`.
	
	
#### Figure 7 - Different modalities of superoscillatory imaging of living cells
For panels a and b the respecive folders contain captured files in the standard format. For panel c, the image is not polarisation-contrast, so we provide only a data set (tiffs) captured at a single polaristion varying over time. For panel, we provide files in the standard format with the addition of data for Channel B (flourescent data). The flourescent data is not captured with polarisation sensitivity, so processed data is limited to _MeanTrans, the mean of the 4 captured images. 

## Supplementary figures

#### Figure 1S - Schematic of the superoscillatory microscope 
No underlying data
	
#### Figure 2S - Resolution improvement in a superoscillatory microscope
All data is stored in the same format as in the file used in figure 5, panel a of the main manuscript
- Panel 1 (i), (ii) and (iv) use the same simulated data as in the main manuscript figure 5, panel a, and can be taken from that file.
- The data in Panel 1 (iii) is stored in file: `fig 2S 1 (iii) - siemens_star_objective_tgt_width=0.4_iso=1_separ=10000_spher=0_astig=0.mat`. Noting that (counter-intuatively) the **convetional confocal image is stored in the variable `so_image`**.
- The data in panels 2-5 is stored in the equivalently named files, again that (counter-intuatively) the **convetional confocal image is stored in the variable `so_image`** in the appropriate file.


#### Supplementary videos 1 - 4. 
These videos have the same underlying data as figure 7 in the main manuscript and avi files are included in those folders. Figure 7 contains single frames of the provided videos


Geographic location of data collection: University of Southampton, U.K.

Related projects:
XXXXXXXXXXXXXXXXXXXX


Dataset available under a CC BY 4.0 licence


Publisher: University of Southampton, U.K.


Date: May 2020