This repository will be a summary and outlook on all our open, medical, AI advancements.

Last update: Dec 30, 2022

Related tags

Overview

medical by LAION

This repository will be a summary and outlook on all our open, medical, AI advancements. See the medical-general channel in the medical category on our discord server for more information!

Current projects

fMRI to image

Project ideas

Fine-tune clip for pathology images [1]
Replicate PLOS paper with current image generation techniques

Datasets

[1] image-text pairs of pathology images (~10k) Go to dataset

Methods

Fine-tuning existing CLIP models

Comments

Add conda env setup
Adding two files to help setting up a conda environment setup that will run all of the fMRI notebooks: environment.yaml and setup.sh.

Example usage (added to README):

conda create -n medical-v1 python=3.10 conda activate medical-v1 conda env update -f environment.yaml # this one must be run individually since it uses git pip install git+https://github.com/openai/CLIP.git

The setup.sh script list the conda and pip commands to create this environment.
opened by jimgoo 1

feat: add better seed functionality

This PR aims to add a better seed function for the notebook to ensure reproducibility. Instead of just setting the seed for numpy and torch we also set the seed for random, os and set the deterministic flag to True. viz.

def seed_everything(seed=GLOBAL_SEED):
    """"
    Seed everything.
    """   
    random.seed(seed)
    os.environ['PYTHONHASHSEED'] = str(seed)
    np.random.seed(seed)
    torch.manual_seed(seed)
    torch.cuda.manual_seed(seed)
    torch.cuda.manual_seed_all(seed)
    torch.backends.cudnn.deterministic = True

opened by SauravMaheshkar 1

style: Update `.gitignore`, delete `.DS_Store` and add README to `fMRI/` dir
Changes are summarised as follows:

Update .gitignore with global paths for ipynb checkpoints and .DS_Store.

Remove the previously committed .DS_Store binary.

Add a README to the fMRI/ dir with direct links to open the notebooks in colab.
opened by SauravMaheshkar 1
added link to TCGA pathology reports.

The matching images still need to be pulled together, but they are available to the public from different sources, (e.g. by cohort from https://www.cancerimagingarchive.net/collections/), or to people with TCGA access from GDC.

opened by dlivitz 1
Remove subj_01 from h5 keys

Removing the subj_01 part from the keys for compatibility with https://huggingface.co/datasets/pscotti/naturalscenesdataset/blob/main/subj01_nsdgeneral.hdf5.

opened by jimgoo 0

This repository will be a summary and outlook on all our open, medical, AI advancements.

Related tags

Overview

medical by LAION

Current projects

Project ideas

Datasets

Methods

Comments

Add conda env setup

feat: add better seed functionality

style: Update `.gitignore`, delete `.DS_Store` and add README to `fMRI/` dir

added link to TCGA pathology reports.

Remove subj_01 from h5 keys

Owner

LAION AI

Build a medical knowledge graph based on Unified Language Medical System (UMLS)

We evaluate our method on different datasets (including ShapeNet, CUB-200-2011, and Pascal3D+) and achieve state-of-the-art results, outperforming all the other supervised and unsupervised methods and 3D representations, all in terms of performance, accuracy, and training time.

An unofficial styleguide and best practices summary for PyTorch

Cross-platform CLI tool to generate your Github profile's stats and summary.

Codebase for the Summary Loop paper at ACL2020

Summary Explorer is a tool to visually explore the state-of-the-art in text summarization.

This repository is an unoffical PyTorch implementation of Medical segmentation in 3D and 2D.

PyTorch implementation of our Adam-NSCL algorithm from our CVPR2021 (oral) paper "Training Networks in Null Space for Continual Learning"

Convolutional neural network web app trained to track our infant’s sleep schedule using our Google Nest camera.

The official pytorch implementation of our paper "Is Space-Time Attention All You Need for Video Understanding?"

Medical-Image-Triage-and-Classification-System-Based-on-COVID-19-CT-and-X-ray-Scan-Dataset

The official repository for our paper "The Devil is in the Detail: Simple Tricks Improve Systematic Generalization of Transformers". We significantly improve the systematic generalization of transformer models on a variety of datasets using simple tricks and careful considerations.

This repository holds code and data for our PETS'22 article 'From "Onion Not Found" to Guard Discovery'.

This repository contains the entire code for our work "Two-Timescale End-to-End Learning for Channel Acquisition and Hybrid Precoding"

Code for the prototype tool in our paper "CoProtector: Protect Open-Source Code against Unauthorized Training Usage with Data Poisoning".

CoTr: Efficiently Bridging CNN and Transformer for 3D Medical Image Segmentation

Medical Image Segmentation using Squeeze-and-Expansion Transformers

nnDetection is a self-configuring framework for 3D (volumetric) medical object detection which can be applied to new data sets without manual intervention. It includes guides for 12 data sets that were used to develop and evaluate the performance of the proposed method.

TorchIO is a Medical image preprocessing and augmentation toolkit for deep learning. Part of the PyTorch Ecosystem.