Is it possible to link a ItemTable to a DynamicMap consists of overlays of VSpans?

I would like to achieve something like using boundsx tool to select multiple ranges on a plot and use multiple VSpans to visualize the underlying selections. My requirement is to first coarsely select the range by boundsx tool, and then fine tune the bounds in an editable ItemTable. I was able to use a external list to keep the selections but I have no idea how can I use the ItemTable to edit the underlying list (span_list).

I have seen a couple of examples such as curveedit. This example worked because the data of the curve can be linked to the table. In my case the data should be extracted from a DynamicMap and its wrapped elements. I wonder whether this is possible? Is there a better way to achieve this?

Thank you!

Here are the minimal example that could display the boundaries of VSpans but not able to edit them in the ItemTable:

import holoviews as hv
from holoviews import opts
import numpy as np

x = np.arange(10)
y = np.random.rand(10)

# dummy curve
curve = hv.Curve((x, y))

span_lists = []
bx = hv.streams.BoundsX(source=curve, boundsx=(0, 0))
def dmap_func(boundsx):
    if boundsx[0]:
    return hv.Overlay([hv.VSpan(x1=x[0], x2=x[1]) for x in span_lists])
dmap = hv.DynamicMap(dmap_func, streams=[bx])

def dtable_func(boundsx):
    left = [x[0] for x in span_lists]
    right = [x[1] for x in span_lists]
    return hv.ItemTable([('Left', left), ('Right', right)])
dmap_tb = hv.DynamicMap(dtable_func, streams=[bx])
(curve * dmap + dmap_tb).opts(
    opts.Curve(responsive=True, height=400),