Date of Presentation

5-2-2024 12:00 AM

College

Rowan-Virtua School of Osteopathic Medicine

Poster Abstract

INTRODUCTION

Deep learning is a method of artificial intelligence involving progressively layered neural networks to extrapolate patterns from data to provide predictions. Moreover, given the arduous nature required for examining CT scans for intracranial aneurysms, discovering ways to expedite this process is beneficial. The use of deep learning to evaluate CT angiograms for intracranial aneurysms has been sparsely studied. This literature review aims to determine the accuracy and reliability of deep learning to analyze CT angiograms in patients suspected to have intracranial aneurysms.

METHODS

A qualitative review of literature using PubMed, SCOPUS, and EMBASE was conducted. Inclusion criteria comprised articles using deep learning to analyze CT angiograms for detecting intracranial aneurysms. The search string was “deep learning”, CT angiogram”, and “intracranial aneurysm” with results limited from 2020-2024. After screening, eleven papers met the inclusion criteria.

RESULTS

Multiple studies demonstrated that deep learning models led to higher sensitivity when radiologists use them simultaneously. Additionally, deep learning has shown benefits in predicting future aneurysm growth based on pre-existing characteristics such as size and shape. Furthermore, intracranial aneurysms >3 mm in diameter had the highest levels of sensitivity and accuracy, while those with diameters lowest.

CONCLUSION

Equipped with deep learning models, radiologists have a superior capability to detect intracranial aneurysms on CT angiograms. Future studies should establish methods that deep learning can have to identify intracranial aneurysms <3 >mm, as well as minimizing biases arising during the model training period.

Keywords

deep learning, CT angiogram, CT Angiography, intracranial aneurysm

Disciplines

Cardiovascular Diseases | Diagnosis | Health and Medical Administration | Investigative Techniques | Medicine and Health Sciences | Nervous System Diseases | Neurology | Pathological Conditions, Signs and Symptoms | Radiology | Translational Medical Research

Document Type

Poster

DOI

10.31986/issn.2689-0690_rdw.stratford_research_day.106_2024

Share

COinS
 
May 2nd, 12:00 AM

Applications of Deep Learning With Detecting Intracranial Aneurysms on CT Angiograms: A Literature Review

INTRODUCTION

Deep learning is a method of artificial intelligence involving progressively layered neural networks to extrapolate patterns from data to provide predictions. Moreover, given the arduous nature required for examining CT scans for intracranial aneurysms, discovering ways to expedite this process is beneficial. The use of deep learning to evaluate CT angiograms for intracranial aneurysms has been sparsely studied. This literature review aims to determine the accuracy and reliability of deep learning to analyze CT angiograms in patients suspected to have intracranial aneurysms.

METHODS

A qualitative review of literature using PubMed, SCOPUS, and EMBASE was conducted. Inclusion criteria comprised articles using deep learning to analyze CT angiograms for detecting intracranial aneurysms. The search string was “deep learning”, CT angiogram”, and “intracranial aneurysm” with results limited from 2020-2024. After screening, eleven papers met the inclusion criteria.

RESULTS

Multiple studies demonstrated that deep learning models led to higher sensitivity when radiologists use them simultaneously. Additionally, deep learning has shown benefits in predicting future aneurysm growth based on pre-existing characteristics such as size and shape. Furthermore, intracranial aneurysms >3 mm in diameter had the highest levels of sensitivity and accuracy, while those with diameters lowest.

CONCLUSION

Equipped with deep learning models, radiologists have a superior capability to detect intracranial aneurysms on CT angiograms. Future studies should establish methods that deep learning can have to identify intracranial aneurysms <3>mm, as well as minimizing biases arising during the model training period.

 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.