/Knee-Merchant-Landmark-Detection

Measurement of Merchant Angles is a fully automated method to measure Patellar Congruence Angle and Tilt Angle on Merchant Knee radiographs, employing CNN landmark localizers

Primary LanguageJupyter Notebook

Automatation-Drawing-Merchant-Lines-and-Angles-of-Knee

As soon as possible , the paper will be published

The Dataset

For training algorithm, the training data was prepared from LERA - Lower Extremity Radiographs . It has consists of Merchant view of the knee and their landmark heatmaps.

Dataset Description

In this retrospective, HIPAA-compliant, IRB-approved study, we collected data from 182 patients who underwent a radiographic examination at the Stanford University Medical Center between 2003 and 2014. The dataset consists of images of the foot, knee, ankle, or hip associated with each patient.

On Hugging Face App , Check Out !

Train Your Own Model

Firstly , you need to load resized data as a numpy array .

#For An Example
import numpy as np

images=np.load("To_Path/stanford_merchantdata256x256.npz")
#X Merchant Views
x=images.f.x
#Y Heatmaps 
y=images.f.y

Then , you can train your own model with one click. For the code running smoothly , you should care of paths. For example , " np.load" , "model.save" etc... Before you run the code , you should check the code and its comments.

MerchantApp

It has easy usage. After open app , you should click file , then click predict. When the file dialog is open , you can choose your dicom file.

Download App

merchantapp

On Command

For usage on command . In command, you can predict many dicoms file.

-- DicomsPath = Your Dicoms File of Path. The file must consist of dicoms . It can have one more than. It only supports dicom extension.

-- OutputPath = Your Results File of Path. The predictions will be saved into this file. It will consist of csv and the predicted images.

-- ModelPath = Your Trained Model of Path .

#Example For Usage 

cd Automatation-Drawing-Merchant-Lines-Angles-of-Knee

pip install -r requirement.txt

python main.py --DicomsPath "To_DicomsPath" --OutputPath "To_YourOutputFolderPath" --ModelPath "To_ModelPath"

Methods

Trained Model will be shared as soon as !

Model

model_diagram

Adaptive Wing Loss

awl