WEBVTT
Kind: captions
Language: en

00:00:13.849 --> 00:00:18.440
Next, I will introduce the

00:00:18.440 --> 00:00:20.260
Part 2 topic

00:00:20.320 --> 00:00:29.520
is the Magnetic Resonance Image Analysis Method and Method for Evaluating the Risks of Radiotherapy

00:00:29.700 --> 00:00:37.660
Cerebral AVM (arterio-venous malformation)  is a congenital cerebral vascular disease

00:00:37.660 --> 00:00:41.800
with an incidence rate.

00:00:41.800 --> 00:00:49.240
An AVM comprises an abnormal nidus of blood vessels

00:00:49.240 --> 00:00:53.920
that shunt blood directly from an artery
to a vein

00:00:53.920 --> 00:01:00.580
and thereby bypass an intervening capillary bed.

00:01:00.580 --> 00:01:08.420
it can take 2–3 years after GKRS Radiosurgery

00:01:08.420 --> 00:01:14.000
before the AVM is obliterated,

00:01:14.000 --> 00:01:20.920
and adverse radiation effects (ARE)

00:01:20.920 --> 00:01:27.700
during the latency period can lead to brain edema.

00:01:27.700 --> 00:01:35.640
There are many complication after GKRS Radiosurgery

00:01:35.640 --> 00:01:42.900
such as the Cerebral infracts and others.

00:01:42.900 --> 00:01:51.240
This is the Strategy of Radiosurgery for AVMs in Taipei hospital.

00:01:51.240 --> 00:02:03.760
Including the Imaging procedures, AVM nidus delineation, dose planning, dose prescription,

00:02:03.760 --> 00:02:08.440
staged radiosurgery and the follow-up.

00:02:08.760 --> 00:02:14.820
This figure shows the Histographic Characterization of Cerebral AVM.

00:02:14.820 --> 00:02:19.500
The intensity histogram of the T2w image

00:02:19.500 --> 00:02:24.820
was obtained within the radiation exposure region,

00:02:24.820 --> 00:02:30.140
with the nidus (dark blue), brain tissue (green), and CSF (red) demarcated manually.

00:02:30.140 --> 00:02:35.380
(A) The red curve indicates the region exposed to radiation;

00:02:35.380 --> 00:02:40.000
(B) The blue, green, and red colors

00:02:40.000 --> 00:02:44.940
indicates respectively indicate the nidus, brain tissue, and CSF

00:02:44.940 --> 00:02:49.180
delineated via manual segmentation.

00:02:49.180 --> 00:02:58.560
The figure (C) Intensity histogram of nidus, brain tissue, and CSF

00:02:58.560 --> 00:03:07.080
obtained using manual segmentation within the region exposed to radiation.

00:03:07.080 --> 00:03:14.120
Manual clustering was performed by an experienced neurosurgeon

00:03:14.120 --> 00:03:21.300
according to the radiation exposure region and T2w intensity.

00:03:21.300 --> 00:03:29.840
Preprocessing was performed to obtain the boundary of the prescription isodose region

00:03:29.860 --> 00:03:34.500
within the parenchymal brain in T2w images.

00:03:34.500 --> 00:03:41.500
The areas of the nidus, brain tissue, and CSF within the boundary

00:03:41.500 --> 00:03:46.100
were identified via manual demarcation.

00:03:46.100 --> 00:03:53.920
In fuzzy c-means (FCM) clustering, each data point may belong to more than one.

00:03:53.920 --> 00:04:02.560
This image analysis technique is well established in tumor detection.

00:04:02.560 --> 00:04:11.620
Clustering and cluster analysis involves assigning
data points to clusters to ensure

00:04:11.640 --> 00:04:18.680
that items in a given cluster are as similar as possible

00:04:18.680 --> 00:04:27.360
and that items belonging to different clusters are as
dissimilar as possible.

00:04:27.360 --> 00:04:36.760
Clusters are identified using similarity measures, such as distance, connectivity, and intensity.

00:04:36.760 --> 00:04:46.340
The choice about which measures to use is based on available data or the intended application.

00:04:46.340 --> 00:04:54.560
Our objective in this study was to use the FCM technique

00:04:54.560 --> 00:05:02.980
to classify the prescription isodose radiation exposure range in T2 weighted (T2W)

00:05:02.980 --> 00:05:10.700
images for GKRS plans with the aim of calculating

00:05:10.700 --> 00:05:19.680
precise percentages and volumes of the various forms of tissue.