Development of Effective Acute Infraction Diagnostic Approach using Brain Region Evaluator
- HwangJi Hye, DongYoon Han,1Hye JinKang, Hyug-GiKim, KyungMi Lee, Jang hoon Oh, SoonchanPark, Chang-WooRyu,EuiJong Kim
-
Department of Radiology, Kyung Hee University College of Medicine, Kyung Hee University Hospital, #23 Kyungheedae-ro, Dongdaemun-gu, Seoul
02447, Republic of Korea
-
Department of Radiology, Kyung Hee University Hospital at Gangdong, Kyung Hee University College of Medicine, 892 Dongnam-ro, Gangdong-Gu,
Seoul 05278, Republic of Korea
- Department of Research, ReadBrain Inc, Republic of Korea
Abstract
Purpose
The objective of this study was to propose the effective evaluation technique of acute infarction using brain region evaluator.
Methods
A total of 1,600 subjects, which included 800 normal and 800 acute infarction subjects each, underwent DWI scan. DWI was independently
labeled as two steps- 1) upper, middle, and lower brain regions to evaluate lesion according to brain position, and 2) acute infarction or
normal. The datasets were split into training (70%), validation (15%) and test (15%). We implemented optimized convolutional neural network
(CNN) software for acute infraction evaluation (v. 1.0, ReadBrain Inc, RB-Stroke, Korea). The performance was conducted the quantitative
value using the area under curve (AUC) and the lesion detection was evaluated qualitatively by radiologists.
Results
The performance showed 0.88 AUC for not split brain region and 0.94 AUC for split brain region. The model with brain region evaluator can
provide lesion location as well as more reasonable lesion detection than not split brain.
Conclusion
Recognition of acute infarction with brain region evaluation model showed higher accuracy and the result of reasonable lesion detection than
not split the brain region. Deep learning with MR imaging can be used as an adjunct to diagnose the acute infarction.
Keywords
Acute Infarction; Deep Learning; Machine Learning; Convolutional neural network (CNN); Lesion Detection.
Clinical relevance/application
Deep learning method with brain region evaluator can demonstrate to be used to evaluate acute infarction indicating it as a practical method
for early and precision diagnosis.
Improvement diagnostic accuracy of sinusitis recognition in paranasal sinus X-ray using multiple deep learning models
- Hyug-Gi Kim, Kyung Mi Lee, Eui Jong Kim, Jin San Lee
-
Department of Radiology, 2Department of Neurology, Kyung Hee University College of Medicine, Kyung Hee University Hospital, Seoul,
Republic of Korea
Abstract
Background
Sinus X-ray imaging is still used in the initial evaluation of paranasal sinusitis, which is diagnosed by the opacification or air/fluid
level in the sinuses and best seen in the Waters’ view of the paranasal sinus (PNS). The objective of this study was to investigate the
feasibility of recognizing the maxillary sinusitis features using PNS X-ray images, as well as to propose the most effective method of
determining a reasonable consensus using multiple deep learning models.
Methods
A total of 4,860 patients, which included 2,430 normal and maxillary sinusitis subjects each, underwent Waters’ view PNS X-ray scan. The
datasets were randomly split into training (70%), validation (15%), and test (15%) subsets. We implemented a majority decision algorithm to
determine a reasonable consensus using three multiple convolutional neural network (CNN) models: VGG-16, VGG-19, and ResNet-101. The
performance of sinusitis detection was evaluated with quantitative accuracy (ACC) and activation maps.
Results
We compared the results of our approaches with ACC and activation maps. ACC [and area under the curve (AUC)] of the internal test dataset
was evaluated as 87.4% (0.891), 90.8% (0.891), 93.7% (0.937), and 94.1% (0.948) for VGG-16, VGG-19, ResNet-101, and the majority decision,
respectively. ACC (and AUC) of the external test dataset was evaluated as 87.58% (0.877), 87.58% (0.877), 92.12% (0.929), and 94.12% (0.942)
for VGG-16, VGG-19, ResNet-101, and the majority decision, respectively. Majority decision algorithms can detect missing and correct lesions
using a compensation function of the majority decision.
Conclusion
The majority decision algorithm showed high accuracy and significantly more accurate lesion detection compared with those of individual CNN
models. The proposed deep learning method with PNS X-ray images can be used as an adjunct to classify maxillary sinusitis.
Keywords
Sinusitis; deep learning; convolutional neural network (CNN); paranasal sinus (PNS) X-ray; majority decision
개인 맞춤형 뇌질병 진단 및 상태 판정을 위한 의료 영상 처리 시스템 및 방법
개인 맞춤형 뇌질병 진단 및 상태 판정을 위한 의료 영상 처리 시스템 및 방법
개인 맞춤형 뇌질병 진단 및 상태 판정을 위한 의료 영상 처리 시스템 및 방법
개인 맞춤형 뇌질병 진단 및 상태 판정을 위한 의료 영상 처리 시스템 및 방법
블록 기반 유연한 AI 모델을 이용한 지능형 의료진단 및 진료 시스템
인공지능 기반 커리큘럼 학습 방법을 이용한 의료 영상 처리 시스템 및 지능형 의료 진단 및 진료 시스템
인공지능 기반 커리큘럼 학습 방법을 이용한 의료 영상처리 시스템, 지능형 의료 진단 및 진료 시스템 및 블록기반 유연한 AI 모델을 이용한
지능형 의료 진단 및 진료 시스템