Datasets:
BearSubj13
commited on
Commit
•
3a4e5ff
1
Parent(s):
1c1bffb
Upload README.md
Browse files
README.md
ADDED
@@ -0,0 +1,105 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
CoronaryDominance dataset
|
2 |
+
|
3 |
+
DATA DESCRIPTION
|
4 |
+
The data we present here includes three parts - the main dataset, the real distribution dataset, and the domain shift dataset.
|
5 |
+
P.S. IMPORTANT!!! The angiograms of the real distribution dataset are from the CoronarySYNTAX dataset.
|
6 |
+
|
7 |
+
The data we present here includes three parts - the main dataset, the real distribution dataset, and the domain shift dataset. The main dataset with
|
8 |
+
1,025 studies (319 of whom have left dominance, and 706 of whom have right dominance) from the Philips Allura Clarity cardiovascular imaging system is
|
9 |
+
for model training and hyper-parameters fine-tuning. For the main dataset, each study falls into one of five categories: bad quality, artifact, small
|
10 |
+
diameter, occlusion, or normal. ”Normal” is the largest category. We enhanced the main data set with additional left dominant cases and occlusions in
|
11 |
+
order to address the problem of imbalance
|
12 |
+
|
13 |
+
The real distribution dataset consists of 400 randomly selected studies (54 left dominant, and 346 right dominant) for the Philips Allura Clarity
|
14 |
+
cardiovascular imaging system. Due to this, unlike the main dataset, the real distribution dataset follows the same distribution as all the studies in
|
15 |
+
the clinic. The real distribution dataset does not have group subdivision, so it includes studies with poor quality, artifacts, and high uncertainty.
|
16 |
+
The domain shift dataset comprises 149 studies (52 left and 97 right dominant) from the Philips Azurion cardiovascular imaging system. This dataset
|
17 |
+
also does not have group subdivision. However, we have enhanced it with an additional left dominant studies
|
18 |
+
|
19 |
+
The structure of the data is as follows: The root directory contains two subdirectories: ”Left Dominance” and ”Right Dominance”, which contain
|
20 |
+
angiographic studies of left and right dominant patients, respectively. Each study consists of several angio-graphic views - gray-scale videos saved in
|
21 |
+
Numpy’s compressed array format (.npz). The study directory also contains two subdirectories named ”RCA” (for the right coronary artery) and ”LCA” (for
|
22 |
+
the left coronary artery), respectively.
|
23 |
+
|
24 |
+
The study folder follows the structure ”Study0xxxx study id”, where ”0xxxx” is the sequence number of the study and ”study id” is the study ID. The
|
25 |
+
name of the .npz file containing the angiographic view video matches the series ID. The study ID and the series ID are unique within the dataset and
|
26 |
+
follow the same format has the corresponding ID in DICOM (.dcm) files. Our data are fully anonymized; therefore the IDs in the dataset do not match
|
27 |
+
those in the original .dcm source files. The sequence number is unique within a specific dataset. We have introduced this for the convenience of
|
28 |
+
referencing, such as ”main dataset study 580” instead of ”1.1.11.11111.11.903846392438863822081872923423339322”.
|
29 |
+
We extracted the primary and secondary angles from the fields (0018,1510) and (0018,1511) in the original .dcm file. You can consult the DICOM Standard
|
30 |
+
Browser for more information on these angles. In most cases, the is collaterals tag is True when the is occlusion tag is also True.
|
31 |
+
|
32 |
+
‘‘Below is the data structure in .npz files.
|
33 |
+
|
34 |
+
{Name:”pixel_array”,
Type:”Float array, size of frames×512×512”,
Range: “0 - 255”,
Description: “angiographic 3D view”
}
|
35 |
+
|
36 |
+
{Name:”seriesid”,
Type:”string”,
Range: “consist of numbers and ‘.’ ”,
Description: “unique ID of an angiographic view”
}
|
37 |
+
|
38 |
+
{Name:”studyid”,
Type:”string”,
Range: “consist of numbers and ’.’ ”,
Description: “unique ID of an angiographic study”
}
|
39 |
+
|
40 |
+
{Name:”series_number”,
Type:”integer”,
Range: “0 - 1025”,
Description: “ID of an angiographic view”
}
|
41 |
+
|
42 |
+
{Name:”is_collaterals”,
Type:”Boolean”,
Range: “{True, False, None}”, Description: “Collaterals in LCA”
}
|
43 |
+
|
44 |
+
{Name:”primary_positioner_angle”,
Type:”Float ”,
Range: “-90 - 90, None”,
Description: “positioner angle, degree”
}
|
45 |
+
|
46 |
+
{Name:”secondary_positioner_angle”,
Type:”Float ”,
Range: “-90 - 90, None”,
Description: “positioner angle, degree”}
|
47 |
+
|
48 |
+
{Name:”is_occlusion”,
Type:”Boolean”,
Range: “ {True, False}, None”,
Description: “Occlusion in RCA”
}
|
49 |
+
|
50 |
+
{Name:”is_undefined”,
Type:”Boolean”,
Range: “ {True, False}, None”,
Description: “Studies with high uncertainty”
}
|
51 |
+
|
52 |
+
{Name:”is_artefact”,
Type:”Boolean”,
Range: “ {True, False}, None”,
Description: “The presence of artifacts”
}
|
53 |
+
|
54 |
+
{Name:”artery_type”,
Type:”String”,
Range: “{LCA, RCA}”,
Description: “Coronary artery type”
}
|
55 |
+
|
56 |
+
BACKGROUND INFORMATION AND TECHNICAL DETAILS ON THE DATASET
|
57 |
+
|
58 |
+
We release a new dataset containing invasive coronary angiograms for the coronary dominance classification task, an essential aspect in assessing the
|
59 |
+
severity of coronary artery disease. The dataset holds 1,574 studies, including X-ray multi-view videos from two different interventional angiography
|
60 |
+
systems, which allows one to test the effect of domain shift. Each study has the following tags: bad quality, artifact, high uncertainty, and occlusion.
|
61 |
+
Those tags help to classify dominance classification more accurately and allow to utilize the dataset for uncertainty estimation and outlier
|
62 |
+
detection.
|
63 |
+
|
64 |
+
Coronary dominance classification is an essential step in the SYNTAX (Synergy Between Percutaneous Coronary Intervention with Taxus and Cardiac
|
65 |
+
Surgery) score estimation, which has become an crucial tool for assessing the severity of coronary artery disease. This classification is also
|
66 |
+
necessary when determining the severity of coronary lesions using the Gensini score.
|
67 |
+
Coronary dominance is determined by the coronary artery branch that supplies the posterior descending artery (PDA). There are three main categories:
|
68 |
+
left dominance, right dominance, and co-dominance [3]. Approximately 70-80% of the population has right coronary dominance, while approximately 5-10%
|
69 |
+
have left coronary dominance. In some regions, the proportion of individuals with left coronary dominance may be as high as 20%. Co-dominance occurs in
|
70 |
+
approximately 10-20% ofcases, where the PDA receives blood from the left and right coronary arteries. It is
important to note that co-dominance is not
|
71 |
+
in the SYNTAX scoring system. Therefore, when dealing with complex cases involving arteries supplying the PDA almost equally, clinicians have to choose
|
72 |
+
’right dominance’ when calculating SYNTAX scores.
|
73 |
+
|
74 |
+
Interventional cardiologists use X-ray video, known as angiographic view, to estimate the SYNTAX score and classify coronary dominance. To get these
|
75 |
+
videos,interventional cardiologists inject a contrast agent into the left (LCA) and right (RCA) coronary arteries. By capturing heart pulses from
|
76 |
+
different angles, an angiography study provides valuable information about the cardiovascular system.
The following factors may complicate the
|
77 |
+
classification of coronary dominance: a total occlusion, a small RCA diameter, a poor-quality angiogram, and artifacts. A RCA with a small diameter
|
78 |
+
means there is significant disagreement among experts about how to classify a patient. Cases with a poor-quality angiogram, which can lead to incorrect
|
79 |
+
diagnoses, typically have low-contrast medium filling. Studies that involve artifacts include those with pacing electrodes or sternal wires. Our
|
80 |
+
dataset includes all these categories and allows us to measure the impact of each factor on classification metrics.
|
81 |
+
|
82 |
+
The domain shift is an important problem when working with medical data. A change in a medical device's parameters or settings could significantly
|
83 |
+
impact the neural network's prediction accuracy. To allow ML researchers to experience this phenomenon, we present angiograms from two different
|
84 |
+
cardiovascular imaging systems in our dataset - Phillips Allura Clarity and Philips Azurion.
|
85 |
+
|
86 |
+
Coronary angiography is an essential tool for medical imaging that allows for real-time observation of the heart and blood vessels in a 3D space. Using
|
87 |
+
X-ray technology, this technique captures images on 2D planes and creates video images of the internal organs and blood vessels. These views are not
|
88 |
+
synchronized in time. The study consists of the LCA and RCA views, which typically contain between 20 and 70 frames, 512×512 each. The pixels values
|
89 |
+
are in a range from 0 to 255. The angiograms from Phillips Azurion have a frame around the image which we deleted the frame and resized the images to
|
90 |
+
512×512.
|
91 |
+
|
92 |
+
Six interventional cardiologists with experience ranging from 1 to 25 years provided ground truth information on coronary dominance. A moderator
|
93 |
+
PhD-level cardiologist with more than 15 years of experience verified their labels. We considered his labels to be a gold standard. He also labeled
|
94 |
+
studies with additional tags, such as occlusions and artifacts. The disagreement rate between the experts was 2.8%. After excluding complex categories
|
95 |
+
such as poor quality, artefacts, and small RCA diameters, the disagreement rate on the main data set dropped to 1.5%
|
96 |
+
Our dataset of angiographic studies for coronary dominance classification could be a valuable benchmark for 3D multi-view methods. Since the data come
|
97 |
+
from different cardiovascular imaging systems, one can test the effect of domain shift. The dataset is useful for estimating uncertainty and detecting
|
98 |
+
outliers. If one trains a neural network using data without artifacts and then tests the network on a subset with artifacts, one can estimate the
|
99 |
+
algorithm’s stability concerning outliers. Studies involving RCA with a small diameter are particularly interesting for uncertainty estimation since
|
100 |
+
there is significant disagreement among experts regarding dominance classification in these studies.
|
101 |
+
|
102 |
+
We recommend using normal and occlusion studies from the main dataset for model training and the other studies for testing only. For normal studies,
|
103 |
+
one could use both the RCA and LCA views for coronary dominance classification. However, in our opinion, the RCA views are more suitable for this
|
104 |
+
purpose. For studies with occlusion, only LCA views provide essential information for the dominance classification.
|
105 |
+
|