We used public MIMIC-III and CAS-BP dataset and our collected private Mindray dataset in this study.

• MIMIC III waveform dataset
• MIndray dataset

• Python==2.7
• Pytorch==0.4.1
• CUDA==9.0
• Scikit-learn==0.23.2
• Numpy==1.16.5
• Imbalanced-learn==1.7.0
• Scipy==1.3.1
• Pandas==0.23.4
• Matplotlib==3.3.2

To add new dataset (e.g., NewData), it should be placed in a folder named: NewData in the datasets directory.

Since "NewData" has several domains, each domain should be split into train/test splits with naming style as "train_i_x.pt" and "test_i_x.pt" for each i-th fold.

The structure of data files should in dictionary form as follows: train.pt = {"samples": data, "labels: labels}, and similarly for test.pt.

Next, you have to add a class with the name NewData in the configs/data_model_configs.py file. You can find similar parameter settings for existing datasets as guidelines. Also, you have to specify the cross-domain scenarios in self.scenarios variable.

Last, you have to add another class with the name NewData in the configs/hparams.py file to specify the training parameters.

• SEM-ResNet
• GRU
• ResNet
• TCN
• Inception
• Xception
• mWDN
• GRU-FCN
• Transformer

• AutoML
• KNN
• SVR
• RF
• XGBoost
• Stacking
• MLP

• SMART-BP

• To add a new Regressor, place it in algorithms/algorithms.py file.
• To add a new Architecture, place it in models/arch.py file.

The experiments are organised in a hierarchical way such that:

• Several filters are collected under one directory assigned by --filters.
• Several segmentation methods are collected under one directory assigned by --segments.
• Several feature extraction methods are collected under one directory assigned by --feat_extrat.
• Several visualization methods are collected under one directory assigned by --visualization.

For BP interval classification:

python trainers/train.py  --experiment_description cls  \
                --run_description run_1 \
                --da_method SEM-ResNet \
                --dataset MIMIC \
                --backbone resnet \
                --num_runs 200 \
                --mode train

For BP values estimation:

python trainers/train.py  --experiment_description reg  \
                --run_description run_2 \
                --da_method AutoML \
                --dataset MIMIC \
                --backbone XGBoost \
                --mode train

python cascade.py

import os
from PIL import Image

def view_images_in_folder(folder_path):
    # Get a list of all files in the folder
    file_list = os.listdir(folder_path)

    # Iterate over the files
    for filename in file_list:
        # Check if the file is an image (supports more formats than just .jpg)
        if filename.endswith(".jpg") or filename.endswith(".png") or filename.endswith(".jpeg"):
            # Construct the full file path
            file_path = os.path.join(folder_path, filename)

            try:
                # Open and display the image using Pillow
                img = Image.open(file_path)
                img.show()
            except Exception as e:
                print(f"Error opening {file_path}: {e}")

# Provide the folder path to view the images
folder_path = "./misc"
view_images_in_folder(folder_path)

If you found this work useful for you, please consider citing it.

@article{SMART-BP,
  title   = {SMART-BP: Sem-Resnet and Auto-Regressor Based on a Two-Stage Framework for Noninvasive Blood Pressure Measurement},
  author  = {Chenbin Ma, Yangyang Sun, Peng Zhang, Fan Song, Youdan Feng, Yufang He, GuangLei Zhang},
  journal = {####},
  year    = {2023}
}

For any issues/questions regarding the paper or reproducing the results, please contact any of the following.

Chenbin Ma: [email protected]

Department of Biomedical Engineering, Beihang University, 37 Xueyuan Road, Beijing, 100853