Dataset for Fluorescent Nuclear Track Detectors for out-of-field neutron dosimetry in proton therapy

This dataset was generated in:
Schmidt, S., Muñoz, I. D., Yukihara, E. G., Vedelago, J. (2025). Fluorescent Nuclear Track Detectors for out-of-field neutron dosimetry in proton therapy. Submitted to Medical Physics.

Dataset overview:
This dataset contains fluorescence (“fl”) and reflection (“re”) microscopy images of alumina-based Fluorescent Nuclear Track Detectors (FNTDs). These detectors were used to measure out-of-field neutron dose values in a clinical proton beam.

Experimental setup:
Phantom: RW3 and PMMA phantom
- Dimensions: 30 cm × 30 cm × 30 cm
- Target volume: 10 cm × 10 cm × 10 cm, located at 10 cm to 20 cm depth
Detector placement
- For RW3-I and PMMA-I, all three FNTDs were positioned 8.1 cm downstream of the phantom isocenter
- For RW3-II and PMMA-II, all three FNTDs were positioned 14.1 cm downstream of the phantom isocenter
- No lateral (x-direction) shift
- The FNTDs were placed 1 cm to 2 cm above the isocenter level
Irradiation context
- Out-of-field neutron dosimetry in primary proton beams

Image acquisition:
Detector readout was conducted with a dedicated confocal laser scanning microscopy (FXR700RG).
Each detector was scanned at multiple x–y positions, resulting in 175 x–y image locations per FNTD. At each x–y location, a z-stack consisting of 11 images was recorded. For every image, two channels were acquired:
- fl – fluorescence image (ionizing particle tracks)
- re – reflection image (surface and structural information)

The structure of each zip file is summarized below:
Component Value
------------------------- -----------
Number of detectors 3
x–y images per detector 175
z-stack slices per image 11
Channels per image 2
Configuration file 1
--------------------------------------
A total of 11,550 images per file (~ 5.5 GB unzipped)

File naming convention:
Each image name follows the structure
FNTD_aa_ybb_xcc_zddee_f.png
where
aa = detector number (01, 02, ...)
bb = y index
cc = x index
dd = z slice index
ee = channel ("fl" or "re")
f = detector index without leading zero (0, 1, ...)

Note 1: To properly visualize 16-bit images, it may be necessary to apply intensity scaling (e.g. using ImageJ/FIJI). If not, the images might appear black.
Note 2: For images with x indices of 1 or 2, the dominant signal is due to tracks from a Cm-244 source, which is used for inter-detector sensitivity correction. Therefore, out-of-field analysis is only conducted for images with different x indices.