Base images in RGB format from .E2E files

Hello, when I want to save base image in RGB format, I encounter this problem.
How to read RGB images instead of grayscale images from E2E files?

I have used your codes to read E2E files and tried to modify the code to get 24-bit-depth images.
However, it did not work, and it still output 8-bit-depth grayscale images (base image).
Do I need to modify the codes in src/widgets/cvimagewidget.cpp in OCT-Marker or any files (maybe codes) in LibE2E?

[kaygdev]

Hello edenwu0625,

if you want save the grayscaled images as RGB images or read RGB data from an E2E file?

1. When you want to save RGB base images from the context menu in the OCT-Marker, than you can modify the save function. You must simply add imageTmp.convertTo(QImage::Format_RGB32); in void CVImageWidget::saveBaseImage() after the line cvImage2qtImage(outputImage, imageTmp);.
   If you want to modify the xoct export from LibOCT, you can add cv::cvtColor(image, image, cv::COLOR_GRAY2RGB); in XOctWritter::writeImage before cv::imencode(imageExtention, image, buffer);.

2. When you want to read RGB data from E2E file, you must modify the Image class from LibE2E and maybe copyBScan in he_e2eread.cpp. But I have only seen the 16 bit grayscaled images in E2E files.

Best Kay

Hello Kay,

Thank you for the information, I successfully saved RGB base images from the context menu in the OCT-Marker by adding imageTmp.convertTo(QImage::Format_RGB32); after the line cvImage2qtImage(outputImage, imageTmp);.

1. When I use the original OCT-Marker (without any changes) to show and save images, I got only CV_8UC1 images.
   I understand I can add imageTmp.convertTo(QImage::Format_Grayscale16); , but I am not sure if I save 8 bit or 16 bit grayscaled content in Grayscale16 base images.
   How can I show 16 bit grayscaled images and save lossless 16 bit grayscaled content in Grayscale16 base images?

2. Have you ever seen any patient's information or data like Average Thickness data in thickness map or RNFLT / MRW data in E2E files?

[kaygdev]

Hello edenwu0625,

I am pleased to have helped you. To your questions:

1. when you use this conversation on this image, than you save the 8 bit image in a 16 bit image. The 16 bit data can you save from the context menu with “Save raw image”. For activate this point, you must set the load option “hold OCT raw data” from Options -> Load options and than load the file again. But it is possible that the image is not this what you expect. I transform the Data with a fix lookup table in the view image. The uocte wiki say, it is a transformed float value (32 bit to 16 bit). When you process the 16 bit values, you should consider to use this transformation also.
   The LUT transformation is applied by LibOctData, you can find this in https://github.com/neurodial/LibOctData/blob/master/octdata/import/he_e2e/he_gray_transform.cpp
   The used type of transformation can you choose in OCT-Marker by Options -> Load Options -> E2E Gray.

2. Patient data like name, birthdate, and diagnose is in the file and should show in OCT-Marker. Data like Refraction, preview images, reference SLO image and many more is also in the file and read by LibE2E but not shown in OCT-Marker. I have never seen the Average Thickness data but I have never search this. In the E2E file there are many data fragments that I could not interpret. Some of them seem to be sections with initialized/garbage values.
   The segmentation data is mostly in the file and read by OCT-Marker. It can calculate a thickness map in the “Layer Segmentation” marker mode. For other data like thickness grid, we have used the samirix program in the lab ( https://github.com/neurodial/am_SAMIRIX ).

Best Kay