pyrecist 0.1.0

A package for evaluating RECIST criteria from CSV files

pyrecist

A Python package for computing RECIST 1.1 assessments from lesion measurements stored in tabular format.

Instalation

To install the package, run the following command in your terminal:

pip install pyrecist

Features

• RECIST classification for each follow-up study based exclusively on target lesions.
• Evaluations are performed in chronological order based on study dates, assuming the first study is the baseline and all others are follow-ups.
• Once Progressive Disease (PD) is reached, subsequent classifications are marked as None.
• Supports input in CSV format.
• Not yet supported:
  • Appearance of new lesions.
  • Evaluation of non-target lesions.

Usage

The CLI script takes a .csv file containing RECIST measurements and generates an output .csv file with the corresponding classifications for each follow-up.

For example, to evaluate the measurements on example.csv and save the results to /home/example_user, run:

pyrecist example.csv -o /home/example_user

You can test the package using the synthetic measurements available in the tests/data directory.

Input format

The input should be a CSV file where each row represents a RECIST measurement taken from a study. The following five columns are required:

Column header	Type	Description	Example
patient_id	Integer	Pseudonymized identifier of the patient	1
study_date	String	Date of the study in the format YYYYMMDD	20190220
measurement	Float	RECIST measurement in mm	15.43
lesion_label_alias	String	Letter used as an alias to identify a lesion within a patient	A
lesion_category	String	Lesion category (target, non-target)	target

Run pyrecist -h to see how to customize the name of the expected column headers. For example, if your CSV file contain the measurements in a column called recist_measurements, then you can run:

pyrecist example.csv -o /home/example_user --measurement_header recist_measurements