Function create_table_neighbour_code_to_contour_length

monai/metrics/utils.py:839–878

Returns an array mapping neighbourhood code to the contour length. Adapted from https://github.com/deepmind/surface-distance In 2D, each point has 4 neighbors. Thus, are 16 configurations. A configuration is encoded with '1' meaning "inside the object" and '0' "outside the obje

(spacing_mm, device=None)

def create_table_neighbour_code_to_contour_length(spacing_mm, device=None):
    """
    Returns an array mapping neighbourhood code to the contour length.
    Adapted from https://github.com/deepmind/surface-distance

    In 2D, each point has 4 neighbors. Thus, are 16 configurations. A
    configuration is encoded with '1' meaning "inside the object" and '0' "outside
    the object". For example,
    "0101" and "1010" both encode an edge along the first spatial axis with length spacing[0] mm;
    "0011" and "1100" both encode an edge along the second spatial axis with length spacing[1] mm.

    Args:
        spacing_mm: 2-element list-like structure. Pixel spacing along the 1st and 2nd spatial axes.
        device: device to put the table on.

    Returns:
        A 16-element array mapping neighbourhood code to the contour length.
        ENCODING_KERNEL[2] which is the kernel used to compute the neighbourhood code.
    """
    spacing_mm = ensure_tuple_rep(spacing_mm, 2)
    first, second = spacing_mm  # spacing along the first and second spatial dimension respectively
    diag = 0.5 * np.linalg.norm(spacing_mm)

    neighbour_code_to_contour_length = np.zeros([16], dtype=diag.dtype)
    neighbour_code_to_contour_length[int("0001", 2)] = diag
    neighbour_code_to_contour_length[int("0010", 2)] = diag
    neighbour_code_to_contour_length[int("0011", 2)] = second
    neighbour_code_to_contour_length[int("0100", 2)] = diag
    neighbour_code_to_contour_length[int("0101", 2)] = first
    neighbour_code_to_contour_length[int("0110", 2)] = 2 * diag
    neighbour_code_to_contour_length[int("0111", 2)] = diag
    neighbour_code_to_contour_length[int("1000", 2)] = diag
    neighbour_code_to_contour_length[int("1001", 2)] = 2 * diag
    neighbour_code_to_contour_length[int("1010", 2)] = first
    neighbour_code_to_contour_length[int("1011", 2)] = diag
    neighbour_code_to_contour_length[int("1100", 2)] = second
    neighbour_code_to_contour_length[int("1101", 2)] = diag
    neighbour_code_to_contour_length[int("1110", 2)] = diag
    neighbour_code_to_contour_length = convert_to_tensor(neighbour_code_to_contour_length, device=device)
    return neighbour_code_to_contour_length, torch.as_tensor([[ENCODING_KERNEL[2]]], device=device)


def get_code_to_measure_table(spacing, device=None):