OCT Converter

Python-based tools for reading OCT and fundus data.

Description

In ophthalmology, data acquired from the scanner is often exported in the manufacturer's proprietary file format. OCT-Converter provides python-based tools for extracting images (optical coherence tomography and fundus), as well as associated metadata, from these files.

Supported file formats

• .fds (Topcon)
• .fda (Topcon)
• .e2e (Heidelberg)
• .img (Zeiss)
• .oct (Bioptigen)
• .OCT (Optovue)
• .dcm (both the standard kind, see Dicom, and Zeiss' obfuscated version, see ZEISSDicom)

Installation

Requires python 3.7 or higher.

pip install oct-converter

Usage

A number of example usage scripts are included in examples/.

Here is an example of reading a .fds file:

from oct_converter.readers import FDS

# An example .fds file can be downloaded from the Biobank website:
# https://biobank.ndph.ox.ac.uk/showcase/refer.cgi?id=30
filepath = '/home/mark/Downloads/eg_oct_fds.fds'
fds = FDS(filepath)

oct_volume = fds.read_oct_volume()  # returns an OCT volume with additional metadata if available
oct_volume.peek() # plots a montage of the volume
oct_volume.save('fds_testing.avi')  # save volume as a movie
oct_volume.save('fds_testing.png')  # save volume as a set of sequential images, fds_testing_[1...N].png
oct_volume.save_projection('projection.png') # save 2D projection

fundus_image = fds.read_fundus_image()  # returns a  Fundus image with additional metadata if available
fundus_image.save('fds_testing_fundus.jpg')

Contributions

Are welcome! Here is a development roadmap, including some easy first issues. Please open a new issue to discuss any potential contributions.

Updates

18 November 2022

• Initial support for reading the Zeiss Dicom format.

7 August 2022

• Contours (layer segmentations) are now extracted from .e2e files.
• Acquisition date is now extracted from .e2e files.

16 June 2022

• Initial support for reading Optovue OCTs
• Laterality is now extracted separately for each OCT/fundus image for .e2e files.
• More patient info extracted from .e2e files (name, sex, birthdate, patient ID)

24 Aug 2021

• Reading the Bioptigen .OCT format is now supported

11 June 2021

• Can now specify whether Zeiss .img data needs to be de-interlaced during reading.

14 May 2021

• Can save 2D projections of OCT volumes.

30 October 2020

• Extract fundus and laterality data from .e2e
• Now attempts to extract additional volumetric data from .e2e files that was previously missed

22 August 2020

• Experimental support for reading OCT data from .fda files.

14 July 2020

• Can now read fundus data from .fda files.

Related projects

• uocte inspired and enabled this project
• LibE2E and LibOctData provided some additional descriptions of the .e2e file spec
• eyepy can read HEYEY XML and VOL formats, and eyelab is a tool for annotating this data

Clinical use

We can't guarantee images extracted with OCT-Converter will match those extracted or viewed with the manufacturer's software. Any use in clinical settings is at the user's own risk.