PyFLI (pyfli-lib): A Unified Open-Source Python Library for FLIM Analysis and Fluorescence Lifetime Imaging

PyFLI is an open-source Python library for Fluorescence Lifetime Imaging (FLI/FLIM) analysis. It loads TCSPC, SPAD and ICCD data from different manufacturers through one interface, for both microscopy (FLIM) and macroscopic imaging (MFLI), and provides:

• Traditional analysis: NLSF and MLE lifetime fitting, RLD, Laguerre deconvolution and phasor plot analysis, on CPU or GPU.
• Advanced analysis: FLIM Deep-learning inference, and compressed-sensing reconstruction for single-pixel hyperspectral FLI.
• Advanced statistical analysis:
  • Precision limits: CRLB-derived bounds for fitted parameters, computed from the Poisson Fisher information matrix (Cramér-Rao lower bound, CRLB).
  • Photon economy: coefficient of variation of lifetime estimates as a function of photon count.
  • Error analysis: per-pixel error against simulated ground truth, and goodness-of-fit with Poisson deviance.
  • Method comparison: results from every fitting method binned and compared across factors such as photon count, plus mono- vs. bi-exponential model classification.
  • Distribution tests: classical and multivariate tests that compare simulated and experimental data.
  • Cross-software benchmarking: import results processed in other software and compare them with PyFLI.
• A standardized platform for FLIM deep learning: a complete pipeline for training, running and benchmarking FLIM deep-learning models. It includes a detector-aware simulator that generates labeled training data with known ground truth, and a common benchmark that compares networks against analytical methods.

Supported Data Acquisition Methods

The platform provides native support for several high-end imaging systems:

1. ICCD: Intensified Charge-Coupled Device cameras for fast-gated, wide-field imaging.
2. SPAD: High-speed SPAD (Single-Photon Avalanche Diode) architectures for high-resolution photon counting.
3. TCSPC: Standardized processing for Time-Correlated Single Photon Counting microscopy data.

Data Processing & Analysis

PyFLI (pyfli-lib) implements standard analytical methods to extract lifetime information:

• Non-linear Least Squares Fitting (NLSF): Robust mathematical approach for exponential decay modeling.
• Phasor Plot Analysis: Graphical, model-free transformation of fluorescence decay into a 2D polar plot for easy species separation.
• Maximum Likelihood Estimation (MLE): Statistical estimator optimized for low-photon regimes.
• Rapid Lifetime Determination (RLD): Computationally efficient method for real-time applications and high-frame-rate data.
• Laguerre Method: Model-free IRF deconvolution followed by multi-exponential lifetime extraction on a per-pixel basis.
• Bayesian Inference: Probabilistic parameter estimation with uncertainty quantification for lifetime fitting.

Installation

Install the stable version directly from PyPI:

pip install pyfli-lib

For users requiring GPU-based processing, install the optional tensor/AI dependencies:

pip install "pyfli-lib[gpu]"

Quick Start

Even though the package is installed as pyfli-lib, you import it as pyfli in your scripts:

from pyfli import DataOperations

loader = DataOperations(
    data_path="experimental_data.sdt",
    irf_path="instrument_data.txt",
    bg_path="background_data.tif",
    mask_path="background_data.png",
)
decay_data = loader.load_data()
irf_data = loader.load_irf()

Documentation

Full documentation, examples and API reference are at pyfli.org. Guides:

• How to perform phasor plot analysis in Python with PyFLI
• Training deep-learning FLIM models on simulated data
• Example notebooks: NLSF/MLE fitting, phasor analysis, FLIM simulation and SPAD data I/O

Citation

If you use PyFLI in your research, please cite this package:

Pandey V., Erbas I., Barroso M., Radev S., Intes X. PyFLI: A Python Library for Simulation, Parameter Estimation, and Benchmarking in Fluorescence Lifetime Imaging. https://arxiv.org/abs/2609.11994

@misc{pandey2026pyflipythonlibrarysimulation,
      title={PyFLI: A Python Library for Simulation, Parameter Estimation, and Benchmarking in Fluorescence Lifetime Imaging},
      author={Vikas Pandey and Ismail Erbas and Margarida Barroso and Stefan Radev and Xavier Intes},
      year={2026},
      eprint={2609.11994},
      archivePrefix={arXiv},
      primaryClass={q-bio.QM},
      url={https://arxiv.org/abs/2609.11994},
}

Repository & Issues

The source code is hosted on GitHub. Please report any bugs or feature requests via the issues tracker.

• Website: PyFLI, the open-source FLIM analysis framework (pyfli.org)
• GitHub: https://github.com/vkp217/pyfli-lib
• Contact: For any queries, reach out at pyfli4lifetime@gmail.com or support@pyfli.org