from 2019/05/21 Summer Research at Hopkins under Professor Emad Boctor
The proeject starts with Matlab code, completing simulation based on different phantoms (using K-wave software). With data retrieved from the simulation, use python to process and analyze with neural network Try to decrease the # of elements used in the transducer to increase the flexibility of current probes => "LEGO" like ultrasound probe
Process:
- write bone generator for lumbar puncture based the position of the needle
- process the phantom with K-wave simulation, get the RF data and write beamforming code to process the data and transform it into the data that reflects the intensity at different intersection of the grid
- process the intensity value with supervised learning to train the neural network and the test data would be beamformed data from real ultrasound devices in the lab