Analyze in vivo raw data

How to launch the experiments

• Preparations:
• You should have the raw data then save them in the directory /dataset/\/SDS[1 or 2]/\/
• You should have phantom_simulated in the directory /dataset/phantom_simulated
• keep your pre-trained surrogate model in the directory /surrogate_model/\/
• keep your pre-trained prediction model in the directory /model_save/\/\/
• Make sure your OPs_used is synchronized as previous experiment (MCX simulation, training surrogate model, training prediction model)

• Make sure your ANN_models.py is synchronized as previous experiment (training surrogate model, training prediction model)

• Screenshot of filepath should look like:

• Launch:

  • S1_preprocess_short.ipynb: preprocess raw data of short channel.
  • S1_preprocess_long.ipynb: preprocess raw data of long channel.
  • S2_predict_measured_data.ipynb: use prediction model take processed data as input, get predicted blood oxygen change of IJV.

Diagram of preprocessing of raw data

How to process raw data

S1_preprocess_long.ipynb S1_preprocess_short.ipynb

• After finishing these two process, you would get 4 csv files.
  • Processed short in_vivo_results_exp.csv
  • Processed long in_vivo_results_exp.csv
  • Processed short calibration.csv : mapping to simulation intensity
  • Processed long calibration.csv : mapping to simulation intensity