ultrasound_metrics.metrics.contrast

ultrasound_metrics.metrics.contrast#

Contrast metric.

References

Functions#

contrast(→ float)

Calculate the contrast between two regions.

Module Contents#

ultrasound_metrics.metrics.contrast.contrast(values_inside: jaxtyping.Num[ultrasound_metrics._utils.array_api.ArrayAPIObj, *inside_dims], values_outside: jaxtyping.Num[ultrasound_metrics._utils.array_api.ArrayAPIObj, *outside_dims], *, db: bool = False) float[source]#

Calculate the contrast between two regions.

Equation and argument names follow the convention from [1]. Alternative formulations for frequency-independent analysis are discussed in [2].

  • inside: the region of interest (e.g. lesion)

  • outside: the background region

  • contrast: power-contrast ratio, not dB or amplitude-ratio

The contrast between two regions is defined as:

\[C = \frac{\mu_i}{\mu_o}\]

where:

\[\mu_i = \mathbb{E}\left\{|s_i|^2\right\}\]
\[\mu_o = \mathbb{E}\left\{|s_o|^2\right\}\]

are, respectively, the mean signal power inside and outside the lesion, where \(s\) denotes the signal. Contrast can take any positive real value, and \(C \to \infty\) as \(\mu_o \to 0\).

Contrast is often expressed in decibels as:

\[C[\text{dB}] = 10\log_{10} C\]
Parameters:
  • values_inside – Pixel values from inside the lesion (I), i.e. the region of interest. For hypoechoic lesions like cysts, these are typically lower intensity values.

  • values_outside – Pixel values from outside the lesion (O), i.e. the background region.

  • db – Whether to return the contrast in dB.

Returns:

The power-contrast between the inside and outside regions.

Return type:

float