Merge pull request #27 from rvandewater/main

Adding AUMCdb to MEDS-DEV

src/MEDS_DEV/datasets/AUMCdb/README.md

@@ -0,0 +1,34 @@
+# AUMCdb
+
+## Description
+
+The first freely accessible intensive care database from within the European Union containing de-identified health data related to tens of thousands of European intensive care unit admissions, including demographics, vital signs, laboratory tests and medications.
+
+## Access Requirements
+
+Taken from [official website](https://amsterdammedicaldatascience.nl/amsterdamumcdb/#requesting-access):
+
+- **Access Policy**: Fill out and sign the combined Access and End User License form.
+- **License (for files)**: Specify the license under which the dataset files are distributed.
+- **Data Use Agreement**: Agreement found [here](https://amsterdammedicaldatascience.nl/content/uploads/sites/2/2022/12/arfeula_v1.6.pdf).
+- **Required training**: Valid training courses include the Data or Specimens Only Research (DSOR) course from CITI, the Basic Course for Clinical Investigators (BROK) from NFU or an equivalent course. The DSOR course may be taken free of charge and is also needed to gain access to the MIMIC and eICU intensive care databases from the USA.
+
+## Supported Tasks
+
+- `tasks/mortality/in_icu/first_24h.yaml`
+
+## MEDS-transformation
+
+[MEDS_transforms](https://github.com/mmcdermott/MEDS_transforms) now includes the AUMCdb example. Please refer to the [guide](https://github.com/mmcdermott/MEDS_transforms/tree/main/AUMC_Example).
+
+## Sources
+
+1. [AUMCdb dataset](https://amsterdammedicaldatascience.nl/amsterdamumcdb/)
+2. [AUMCdb Research Paper](https://journals.lww.com/ccmjournal/fulltext/2021/06000/sharing_icu_patient_data_responsibly_under_the.16.aspx)
+3. [AUMCdb Data Use Agreement](https://amsterdammedicaldatascience.nl/content/uploads/sites/2/2022/12/arfeula_v1.6.pdf)
+4. [Data Repository](https://easy.dans.knaw.nl/ui/home)
+5. [Code Repository](https://github.com/AmsterdamUMC/AmsterdamUMCdb)
+
+## Disclaimer
+
+Please refer to the data owners and the most up-to-date information when using this dataset in your research.

src/MEDS_DEV/datasets/AUMCdb/predicates.yaml

@@ -0,0 +1,8 @@
+predicates:
+  icu_admission:
+    code: { regex: "^ICU_ADMISSION//.*" }
+  icu_discharge:
+    code: { regex: "^ICU_DISCHARGE//.*" }
+
+  death:
+    code: MEDS_DEATH

src/MEDS_DEV/datasets/README.md

@@ -6,8 +6,9 @@ To contribute a new dataset:
 
 1. Fork this repository
 2. Add your dataset predicates file in its respective folder (see `MIMIC-IV/predicates.yaml` for an example of predicate structure)
-3. Test locally to ensure your dataset works correctly
-4. Create a pull request with your changes
+3. Test locally to ensure your dataset works correctly. Ideally specify the used packages and versions in the dataset information.
+4. Specify the dataset information (including supported and custom tasks) in the template README.md file in the dataset's folder.
+5. Create a pull request with your changes
 
 ## Notes
 

src/MEDS_DEV/helpers/extract_task.sh

@@ -1,4 +1,27 @@
 #!/bin/bash
+print_help() {
+  echo "Usage: $(basename "$0") <MEDS_ROOT_DIR> <MEDS_DATASET_NAME> <MEDS_TASK_NAME> [additional parameters]"
+  echo
+  echo "Arguments:"
+  echo "  MEDS_ROOT_DIR       The root directory of the MEDS dataset to be used."
+  echo "  MEDS_DATASET_NAME   The name of the dataset to be used."
+  echo "  MEDS_TASK_NAME      The name of the task to be executed."
+  echo
+  echo "Additional parameters can be passed to the aces-cli command."
+  echo
+  echo "Example:"
+  echo "  $(basename "$0") /path/to/meds/root dataset_name task_name --some-parameter=value"
+}
+
+if [[ "$1" == "--help" || "$1" == "-h" ]]; then
+  print_help
+  exit 0
+fi
+
+if [[ $# -lt 3 ]]; then
+  echo "Error: Missing required arguments. See --help for usage."
+  exit 1
+fi
 
 export MEDS_ROOT_DIR=$1
 export MEDS_DATASET_NAME=$2
@@ -11,5 +34,7 @@ export MEDS_DEV_REPO_DIR
 
 SHARDS=$(expand_shards "$MEDS_ROOT_DIR"/data)
 
+echo "Running task $MEDS_TASK_NAME on dataset $MEDS_DATASET_NAME with MEDS_ROOT_DIR=$MEDS_ROOT_DIR and SHARDS=$SHARDS"
+
 aces-cli --config-path="$MEDS_DEV_REPO_DIR"/configs --config-name="_ACES_MD" \
   "hydra.searchpath=[pkg://aces.configs]" "data.shard=$SHARDS" -m "$@"

src/MEDS_DEV/tasks/criteria/mortality/in_icu/first_24h.yaml

@@ -22,10 +22,9 @@ description: >-
 predicates:
   icu_admission: ???
   icu_discharge: ???
-  hospital_discharge: ???
   death: ???
   discharge_or_death:
-    expr: or(icu_discharge, death, hospital_discharge)
+    expr: or(icu_discharge, death)
 
 trigger: icu_admission
 

src/MEDS_DEV/templates/dataset.md

@@ -0,0 +1,30 @@
+# New Dataset Template
+
+This is a template for creating a new dataset in MEDS-DEV. The dataset should be stored in a directory named after the dataset in the `src/MEDS-DEV/datasets` directory.
+
+## Description
+
+Describe the dataset in a few sentences. A link to the dataset's homepage and/or repository or a research paper is recommended.
+
+## Access Requirements
+
+Describe any access requirements for the dataset (e.g, human species research). If the dataset is publicly available, state that here. If the dataset is not publicly available, describe the process for obtaining access. We recommend the following topics be covered:
+
+- **Access Policy**: Describe the access policy for the dataset, including any restrictions or permissions required.
+- **License (for files)**: Specify the license under which the dataset files are distributed.
+- **Data Use Agreement**: Specify any data use agreement that must be signed to access the dataset.
+- **Required training**: Specify any training or certification required to access the dataset.
+
+## Supported Tasks
+
+Describe the existing tasks already present in MEDS-DEV that are covered. If there are new tasks that can be added, describe them here. Also note the `predicates.yaml` file that specifies the dataset's predicates.
+
+## MEDS-transformation
+
+Shortly specify the process of transforming this dataset to the MEDS format. If the dataset is already in the MEDS format when downloaded, specify that here.
+
+## Sources
+
+Summarize the sources of the dataset. If the dataset is a combination of multiple sources, list them here.
+
+1. https://link-to-dataset.org