# ECHO2022 — Data

来源：https://www.kaggle.com/competitions/echo2022/data

采集时间：2026-10-18T13:41:13Z

# Files
*  **train_data/train_data/2CH** - 2CH echo loops, training data
*  **train_data/train_data/4CH** - 4CH echo loops, training data
*  **test_data/test_data/4CH** - 4CH echo loops, test data
*  **test_data/test_data/2CH** - 2CH echo loops, test data
*   **train_data.csv** - patient number and the corresponding ejection fraction
*   **sample_submission.csv** - a sample submission file in the correct format (the numbers in LV_ef are random numbers between 0 and 100)

## .npy files
When importing the .npy-files using `numpy.load` the resulting arrays are structured as follow:

```
(number of images in the loop, image height, image width)
```
## .csv files
The csv files have two colums

- `Patient_number` - the patient number (notice that the test patients have one more digit in the patient number than the train patients)
- `LV_ef` - a float number with one decimal containing a number between 0 and 100 (%)