DeepLIIF: Inferred Multiplex Immunofluorescence for IHC Quantification

Nadeem Lab released DeepLIIF in 2022 as a framework for inferring multiplex immunofluorescence from immunohistochemical (IHC) images. The data supports automated cell segmentation and protein quantification across whole-slide images as validated in Nature Machine Intelligence and CVPR. It provides a computational method to generate multiplexed data from standard, single-stained pathology slides.

Use Cases

• Cell segmentation using inferred multiplex immunofluorescence channels
• Super-resolution of IHC images to enhance digital pathology quantification
• Multitask learning for extracting cell-biology features from whole-slide imaging

Strengths

• Validated across multiple peer-reviewed publications including Nature Machine Intelligence and CVPR
• Provides inferred multiplex immunofluorescence channels for high-dimensional analysis
• Includes whole-slide imaging (WSI) compatibility for large-scale pathology tasks

Limitations

• Specific record counts and dataset volume are not detailed in the repository metadata
• High computational overhead for GAN-based inference on whole-slide images
• Requires specialized domain knowledge in digital pathology to interpret IHC/mIF results

Provenance

Source

Nadeem Lab, published in Nature Machine Intelligence (2022)

Collection Method

Deep learning inference from IHC images using GAN-based architectures

Time Range

2022-2024

Freshness

Last updated March 2026; includes updates from MICCAI 2024 publications.