Welcome to our shared tools and resources repository. In the spirit of open science and reproducibility, we curate the software, pipelines, and tools developed by our lab and our incredible network of current and past collaborators that support our published research. Explore the sections below to view the contributions from each participating lab, and feel free to adapt these tools to accelerate your own discoveries. As we are always looking to improve, we welcome your feedback, questions, or ideas for collaboration.

Chuah-Dugger Lab, UC Davis

• BrainSec2.0 [github]
  BrainSec2.0 is a transformer-based segmentation toolkit for large-scale Whole-Slide Images (WSI) of the human brain. It extends SegFormer with LoRA / QLoRA fine-tuning, quantization, and ONNX-based inference for GPU and CPU environments. The toolkit achieves research-grade accuracy while remaining lightweight enough for macOS laptops and low-resource desktops.
• Infarct detection [github]
  Segmentation and whole-slide–image infarct analysis.
• VLM PEFT [github]
  A unified benchmark suite for evaluating PEFT methods on pathology VLM tasks across multiple datasets. This repository brings together multiple parameter-efficient fine-tuning (PEFT) methods for vision-language models (VLMs), allowing users to reproduce all benchmark experiments with a single script.

Brain Digital Slide Archive

The Digital Slide Archive is a system for working with large microscopy images. It can be used to organize images from a variety of asset stores, such as local file systems and S3. Provide user access controls. Image annotation and review. Run algorithms on all or parts of images. It is leveraging the following Digital Slide Achieve code [github]. 

Gutman Lab, Emory University

• NFT Detection [github]
  ML workflow for detecting Neurofibrillary Tangles (NFTs) in whole slide images. 

Jerry Lou, UC Irvine 

• Morphometric assessment of brain arteriolosclerosis [github]
  Applying machine learning to assist in the morphometric assessment of brain arteriolosclerosis through automation. ArtSeg (arteriolosclerosis segmentation) and VasMorph (vascular morphometry) pipeline code, models, and dataset pointers.

Keiser Lab, UC San Francisco 

• Amyloid Detection using Yolo [github]
  Learning fast and fine-grained detection of amyloid neuropathologies from coarse-grained expert labels
• Automated detection of neurofibrillary tangles in digitized whole slide images using deep learning [github]: 
  A tangle-tracer pipeline to transform point annotations done on AT8-stained WSIs to segmentation ground truth masks. UNet model weights which can be fine-tuned to segment neurofibrillary tangles (NFTs) in novel whole slide pathology images (WSIs). YOLOv8 model weights which can be fine-tuned to detect NFTs in WSIs
• Classification of Alzheimer's disease pathologies [github]
  Interpretable classification of Alzheimer's disease pathologies with a convolutional neural network pipeline. Jupyter notebooks to reproduce preprocessing, CNN training, confidence maps, saliency, and CERAD-like whole-slide scoring.