Canada Spends Logo

Explore our Budget 2026 analysis

Grant: Master's Award: Canada Graduate Scholarships

McMaster University

$17,500.00

Development of a computer-aided algorithm using artificial intelligence for the segmentation of ultrasonographic features of lymph nodes during endobronchial ultrasound.

FY 2018-19

Summary
Lung cancer is the most commonly diagnosed cancer and the leading cause of cancer death worldwide, warranting the need for accurate lymph node (LN) staging to determine treatment plans. A four-point, binary scale known as the Canada Lymph Node Score has previously been validated for the use of predicting malignancy in mediastinal LNs using endobronchial ultrasound (EBUS). However, as ultrasound is operator dependent, agreement between operators was only achieved 22.54% of the time using this tool. If researchers are able to train artificial intelligence (AI) to perform segmentation on images of LNs, it is believed that a computer algorithm will be able to generate an accurate and reliable score for these LNs. In order for AI to predict cancer, it must be proven that AI can segment accurately. We hypothesize that AI will be segment the ultrasonographic criteria more accurately and reliably than a clinician. A sample of 300 retrospectively collected LNs will be used for the training set for AI learning. A clinician will segment the LN images using a computer software, and NeuralSeg, an artificial intelligence program will use this derivation set to develop an algorithm. The algorithm output will be compared against the reference standard as determined by the spatial overlap of the target segmentations to determine reliability and accuracy using SORNSEN-DICE scores. Continuous measures will also be created from the segmentation images and compared against the binary score.
Program
Master's Award: Canada Graduate Scholarships
Location
Ontario, Canada
Recipient Type
Academia
Type
Grant
Source
CIHR Funding Decisions Database

As published

Id
411232
Name
Churchill Isabella F
Region
Ontario
Country
Canada
Orgtype
Institution Paid; Research Institution
Prcname2
Special Cases - Awards Programs
Projecttitle
Development of a computer-aided algorithm using artificial intelligence for the segmentation of ultrasonographic features of lymph nodes during endobronchial ultrasound.
Programtype2
Studentship Programs
Orgnameinp2
McMaster University
Cihramount2
$17,500
Cihrequipment2
$0
Programname2
Master's Award: Canada Graduate Scholarships
Keyworddelim
Artificial Intelligence; Endobronchial Ultrasound; Lung Cancer; Lymph Nodes; Machine Learning; Ultrasonographic Features
Abstract
Lung cancer is the most commonly diagnosed cancer and the leading cause of cancer death worldwide, warranting the need for accurate lymph node (LN) staging to determine treatment plans. A four-point, binary scale known as the Canada Lymph Node Score has previously been validated for the use of predicting malignancy in mediastinal LNs using endobronchial ultrasound (EBUS). However, as ultrasound is operator dependent, agreement between operators was only achieved 22.54% of the time using this tool. If researchers are able to train artificial intelligence (AI) to perform segmentation on images of LNs, it is believed that a computer algorithm will be able to generate an accurate and reliable score for these LNs. In order for AI to predict cancer, it must be proven that AI can segment accurately. We hypothesize that AI will be segment the ultrasonographic criteria more accurately and reliably than a clinician. A sample of 300 retrospectively collected LNs will be used for the training set for AI learning. A clinician will segment the LN images using a computer software, and NeuralSeg, an artificial intelligence program will use this derivation set to develop an algorithm. The algorithm output will be compared against the reference standard as determined by the spatial overlap of the target segmentations to determine reliability and accuracy using SORNSEN-DICE scores. Continuous measures will also be created from the segmentation images and compared against the binary score.
Competitiondate
201812
Approvedterm2
1 yr 0 mth
Pinamesdelim
Churchill, Isabella F
Approvedterm
1 an 0 mois
Cihrcontribution2
~ $17,500~ $0~