Not able to run the application

[Yipinggggg]

Thanks for the awesome work!

I uploaded my datasets and followed all the instructions, but after clicking the visualize data, it seems it keeps running forever. Could you help please?

[GeorgiiKos]

Thank you for pointing this out. After some testing, I noticed inconsistencies in the data loading code that might have caused issues when loading datasets other than the ones we tested. I have fixed these inconsistencies and also updated the README file to provide clearer instructions. The changes have been pushed to the repository and are available in the web version as well. I apologize for the inconvenience and delayed response. Please let me know if the issue persists.

Generally, the data loading process is unfortunately inherently fragile because the application has to process multiple files. This requires all files to be consistent otherwise the application cannot process them. Here is a quick summary of the requirements that the data should fulfill:

• The application expects phase and instrument annotation files in CSV format (comma, tab, or semicolon-separated).
• Each surgery should have exactly one phase and one instrument annotation file.
• Each filename should contain an ID that uniquely identifies a surgery (e.g., surgery ID is 10 in video10_phase.csv).
• Each filename should contain a substring that allows the application to distinguish between phase and instrument annotation files (e.g., phase and inst in video10_phase.csv and video10_inst.csv).
• Phase annotation files should contain two columns: Frame,Phase.
• Phases should be encoded as consecutive numbers starting from 0.
• Instrument annotation files should contain a column Frame and further columns for the instruments (e.g., Frame,Grasper,Bipolar,Hook,Scissors,Clipper,Irrigator,SpecimenBag for the Cholec80 dataset).
• Instrument annotations should be encoded as binary values, where 1 means that the instrument is used and 0 means the instrument is not used.

As an example, in order to load Cholec80 in our application, we transformed the dataset the following way:

1. Copy phase and instrument annotation files to a single directory.
2. Change delimiter from tabs to commas.
3. Map textual phase labels to numbers (0-6).

I also provide screenshots of the Cholec80 dataset format, as well as the configuration form that work with our application, below.

This repository also contains a sample dataset that can serve as an example for the data format that the application expects. Furthermore, the application should log errors in the browser developer console (can be opened by pressing F12 on most browsers) that may provide insights into what is causing the problem.

I hope this helps, please let me know if the issue still persists. Also, please let me know if you are working with a publicly available dataset. In this case, I can offer to test the application with the dataset and provide a tailored guide or fix any bugs that may arise.

[Yipinggggg]

I assume the problem I had before arose from discontinuous surgery number. I just got it running!

Thanks a lot for the reply and the nice work!

[GeorgiiKos]

Glad to hear that! And thank you for pointing out a possible cause of the problem, I will look into it. In case you have suggestions for new features or encounter any further bugs, we will be happy to assist!