Submit Format:¶

you should submit a docker container including your algorithm following the instruction: https://github.com/PerceptionComputingLab/TDSC-ABUS2023/blob/main/how to build a docker image for your algorithm.pdf

Please upload the docker container to a cloud storage service of your choice, such as Google Drive or Baidu Netdisk. Once the upload is complete, generate a shareable link and send it to our official email address: tdscabus@gmail.com.

Classification Result Format:¶

1. The classification results should be saved as a .csv file under classification folder.

2. The prediction .csv file should contain the columns case(an index) and prob(the predicted malignant prob of this case), and the results should be saved in the following format:

Segmentation Result Format:¶

1. The segmentation results should be saved as .nii.gz under segmentation folder.

2. The segmentation results should be named the same way as the corresponding image files, and the folder structure should be the same as the example below:

|— Segmentation

| |—— MASK_130.nrrd

| |—— MASK_131.nrrd

| |—— MASK_132.nrrd

| |——...

Detection Result Format:¶

1. The detection results should be saved as a .csv file under detection folder.

2. The prediction .csv file should have eight columns:

public_id(patient ID),coordX,coordY,coordZ(the center of the bounding box),x_length,y_length, z_length(length of the bounding boxes in three dimensions)and probability(detection confidence), and the results should be saved in the following format: