Source code for intensity_normalization.normalize.kde

"""Kernel density estimation-based tissue mode normalization
Author: Jacob Reinhold <jcreinhold@gmail.com>
Created on: 01 Jun 2021
"""

from __future__ import annotations

__all__ = ["KDENormalize"]

import typing

import intensity_normalization.normalize.base as intnormb
import intensity_normalization.typing as intnormt
import intensity_normalization.util.histogram_tools as intnormhisttool


class KDENormalize(intnormb.LocationScaleCLIMixin, intnormb.SingleImageNormalizeCLI):
    def __init__(self, norm_value: float = 1.0, **kwargs: typing.Any):
        """
        Use kernel density estimation to fit a smoothed histogram of intensities
        of a (skull-stripped) brain MR image, then find the peak of the white
        matter (by default) in the smoothed histogram. Finally, normalize the
        white matter mode of the image to norm_value (default = 1.)
        """
        super().__init__(norm_value=norm_value, **kwargs)

    def calculate_location(
        self,
        image: intnormt.ImageLike,
        /,
        mask: intnormt.ImageLike | None = None,
        *,
        modality: intnormt.Modality = intnormt.Modality.T1,
    ) -> float:
        return 0.0

    def calculate_scale(
        self,
        image: intnormt.ImageLike,
        /,
        mask: intnormt.ImageLike | None = None,
        *,
        modality: intnormt.Modality = intnormt.Modality.T1,
    ) -> float:
        voi = self._get_voi(image, mask, modality=modality)
        tissue_mode = intnormhisttool.get_tissue_mode(voi, modality=modality)
        return tissue_mode

    @staticmethod
    def name() -> str:
        return "kde"

    @staticmethod
    def fullname() -> str:
        return "Kernel Density Estimation"

    @staticmethod
    def description() -> str:
        desc = "Use kernel density estimation to find the WM mode from "
        desc += "a smoothed histogram and normalize an MR image."
        return desc