1. Ultrasound File Format Avatar Ultrasound File Format
  2. UFF
  3. UFF

Wiki

Clone wiki

UFF / v0.3 / uff.channel_data

View History

uff.channel_data < uff

UFF class that contains all the information needed to store and later process channel data.

Parameters

Parameter Datatype Description
authors string (Optional) string with the authors of the data
description string (Optional) string describing the data
local_time string (Optional) string defining the time the dataset was acquired following ISO 8601
country_code string (Optional) string defining the country, following ISO 3166-1
system string (Optional) string defining the system used to acquired the dataset
repetition_rate double or single (Optional) Inverse of the time delay between consecutive repetitions of the whole sequence, often known as framerate
data double or single dataset of dimensions [frames x events x channels x samples] in HDF5
probes uff.probe List of the probes used to transmit/recive the sequence
unique_waves uff.wave List of the unique waves (or beams) used in the sequence
unique_events uff.event List of the unique transmit/receive events used in the sequence
unique_excitations uff.excitation List of the unique excitations used in the sequence
sequence uff.timed_event List of the times_events that describe the sequence
sound_speed double or single Reference sound speed for Tx and Rx events [m/s]

diverging_wave_channel_data.png

Notes

The parameter authors identifies the authors of the data; description describes the acquisition scheme, motivation and application; local_time and country_code identify the time and place the data were acquired; system describes the hardware used in the acquisition; sound_speed contains the reference speed of sound that was used in the system to produce the transmitted waves; and repetition_rate is the sequence repetition rate, also referred to as framerate in some scenarios.

The object uff.channel_data contains all the probes used in the acquisition, a list of the unique_waves that have been transmitted, and a list of the unique_events that form the sequence. The sequence is specified as an array of uff.timed_events, each member containing a reference to an event, and the time_offset since the beginning of the current repetition, also known as frame.

The HDF5 dataset data contains the channel data, organized as a matrix of dimensions (HDF5 notation), :

[frames x events x channels x samples]

where samples is the number of temporal samples acquired by the system, channels is the number of active channels, unique_events is the number of events in the sequence (not unique events), and repetitions is the number of times the described sequence was repeated. Notice that "HDF5 uses C storage conventions, assuming that the last listed dimension is the fastest-changing dimension and the first-listed dimension is the slowest changing." (https://support.hdfgroup.org/HDF5/doc1.6/UG/12_Dataspaces.html). This means that by accessing the data with MATLAB's HDF5 API or Python's h5py the dimension order will be:

[samples x channels x events x frames]

This proposal has the limitation of requiring that all event acquisitions have the same number of time samples and active channels

File examples

Updated