How to document?
Upload a public test case¶
An image is worth a thousand words. Hence we consider it best practice to upload at least one test case and make it public. To have a case with a result visible to the public, you should have a case with a Successful run of the last container available in your Results. To make a successful particular run visible to the public, firstly, navigate to the "View result details" by clicking on "i" icon. Secondly, click on the "Edit this Result" button. Thirdly, check the "Public" box and press "Save".
Write a documentation page¶
Good documentation will ensure proper usage of your Algorithm. So here is a small overview of the recommended documentation. We also created a documentation example further down the page.
For public Algorithms, we'd like you to add contact information in case users need to get in touch:
- Email address
- Link to publication (recommended)
For the user to make optimal use of the developed model, a description is desired. This can be done in approximately 5-10 sentences containing the following information:
- Task of the model
- The type of network the model is based on
- The input the model requires
- Target group
This section should give a precise description of the in- and output data. Optionally, a description of the model could be added.
Validation and performance¶
This section is for a description of method validation and performance. Include some metrics, and perhaps some visual results. Also, describe the ground truth/reference standard that you used to test and optimize your method.
This section contains important background information to keep in mind for the user when using this algorithm.
Common errors and solutions¶
If your algorithm is prone to certain errors that are easily solvable with a small change, please provide the error and its solution so users don't have to contact the developer for common errors.
- Name of the model: Segmentation of cysts in the pancreas
- Last edited: 2021-09-09
- ** Email address:** email@example.com
- ** Associated publications: Doe John, Doe Jane, et al.. Deep learning-based cyst detection and segmentation in MRI. Transactions in Medical Image Analysis. 2022;11:11111.
This model first detects cysts based on the probability of the presence of cysts. Above a certain threshold of probability, it segments the cysts. The model requires a DICOM MR image as input and creates a probability map and segmentation mask as output. The model is based on nnUnet.
- Datatype: Coronal MR image of the thorax
- File format: DICOM image of size (512, 512, 3)
- Target group: Patients older than 18 and younger than 95
- Results: Probability between 0 and 1 stating the possibility of a cyst being present in the pancreas
- Type of image output: Probability map with a heat map color-coding that indicates which regions in the image were influential to the prediction of the model: colder (blue-green) and warmer (yellow-red) colors respectively indicate low and high probability regions
Validation and performance¶
Ground truth: Manually segmentation by experienced radiologist at the Radboudumc, Nijmegen
|Coronal MR thorax images||90 images||University Medical Center Utrecht, Utrecht||Between 2010-2014||AUC||0.96|
|Coronal MR thorax images||100 images||Leiden University Medical Center, Leiden||Between 2009-2012||AUC||0.89|
This algorithm was developed for research purposes only. Please be aware that the uploaded images and results will be visible to the owner of the algorithm. Make sure to properly anonymize your data before you consider uploading.