Evaluating severity of white matter lesions from computed tomography images with convolutional neural network

Abstract Purpose Severity of white matter lesion (WML) is typically evaluated on magnetic resonance images (MRI), yet the more accessible, faster, and less expensive method is computed tomography (CT). Our objective was to study whether WML can be automatically segmented from CT images using a convolutional neural network (CNN). The second aim was to compare CT segmentation with MRI segmentation. Methods The brain images from the Helsinki University Hospital clinical image archive were systematically screened to make CT-MRI image pairs. Selection criteria for the study were that both CT and MRI images were acquired within 6 weeks. In total, 147 image pairs were included. We used CNN to segment WML from CT images. Training and testing of CNN for CT was performed using 10-fold cross-validation, and the segmentation results were compared with the corresponding segmentations from MRI. Results A Pearson correlation of 0.94 was obtained between the automatic WML volumes of MRI and CT segmentations. The average Dice similarity index validating the overlap between CT and FLAIR segmentations was 0.68 for the Fazekas 3 group. Conclusion CNN-based segmentation of CT images may provide a means to evaluate the severity of WML and establish a link between CT WML patterns and the current standard MRI-based visual rating scale.

Download Full-text

Improvement in the Convolutional Neural Network for Computed Tomography Images

Applied Sciences ◽

10.3390/app11041505 ◽

2021 ◽

Vol 11 (4) ◽

pp. 1505

Author(s):

Keisuke Manabe ◽

Yusuke Asami ◽

Tomonari Yamada ◽

Hiroyuki Sugimori

Keyword(s):

Neural Network ◽

Computed Tomography ◽

Convolutional Neural Network ◽

Ct Images ◽

Classification Model ◽

Max Pooling ◽

Computed Tomography Images ◽

Input Size ◽

Filter Size

Background and purpose. This study evaluated a modified specialized convolutional neural network (CNN) to improve the accuracy of medical images. Materials and Methods. We defined computed tomography (CT) images as belonging to one of the following 10 classes: head, neck, chest, abdomen, and pelvis with and without contrast media, with 10,000 images per class. We modified the CNN based on the AlexNet with an input size of 512 × 512. We resized the filter sizes of the convolution layer and max pooling. Using these modified CNNs, various models were created and evaluated. The improved CNN was evaluated to classify the presence or absence of the pancreas in the CT images. We compared the overall accuracy, which was calculated from images not used for training, to that of the ResNet. Results. The overall accuracies of the most improved CNN and ResNet in the 10 classes were 94.8% and 89.3%, respectively. The filter sizes of the improved CNN for the convolution layer were (13, 13), (7, 7), (5, 5), (5, 5), and (5, 5) in order from the first layer, and that of max-pooling was (7, 7). The calculation times of the most improved CNN and ResNet were 56 and 120 min, respectively. Regarding the classification of the pancreas, the overall accuracies of the most improved CNN and ResNet were 75.75% and 58.25%, respectively. The calculation times of the most improved CNN and ResNet were 36 and 55 min, respectively. Conclusion. By optimizing the filter size of the convolution layer and max-pooling of 512 × 512 images, we quickly obtained a highly accurate medical image classification model. This improved CNN can be useful for classifying lesions and anatomies for related diagnostic aid applications.

Download Full-text

Human Organ Classifications from Computed Tomography Images Using Deep-Convolutional Neural Network

6th International Conference on the Development of Biomedical Engineering in Vietnam (BME6) - IFMBE Proceedings ◽

10.1007/978-981-10-4361-1_155 ◽

2017 ◽

pp. 917-923

Author(s):

Ho Thi Kieu Khanh ◽

Tran Cong Hung ◽

Viet-Hung Dang ◽

Nguyen Duc Thang

Keyword(s):

Neural Network ◽

Computed Tomography ◽

Convolutional Neural Network ◽

Deep Convolutional Neural Network ◽

Human Organ ◽

Computed Tomography Images

Download Full-text

An artificial intelligence algorithm that identifies middle turbinate pneumatisation (concha bullosa) on sinus computed tomography scans

The Journal of Laryngology & Otology ◽

10.1017/s0022215120000444 ◽

2020 ◽

Vol 134 (4) ◽

pp. 328-331 ◽

Cited By ~ 1

Author(s):

P Parmar ◽

A-R Habib ◽

D Mendis ◽

A Daniel ◽

M Duvnjak ◽

...

Keyword(s):

Neural Network ◽

Artificial Intelligence ◽

Computed Tomography ◽

Diagnostic Accuracy ◽

Convolutional Neural Network ◽

Middle Turbinate ◽

Concha Bullosa ◽

Computed Tomography Images ◽

Computed Tomography Scans ◽

Neural Network Algorithm

AbstractObjectiveConvolutional neural networks are a subclass of deep learning or artificial intelligence that are predominantly used for image analysis and classification. This proof-of-concept study attempts to train a convolutional neural network algorithm that can reliably determine if the middle turbinate is pneumatised (concha bullosa) on coronal sinus computed tomography images.MethodConsecutive high-resolution computed tomography scans of the paranasal sinuses were retrospectively collected between January 2016 and December 2018 at a tertiary rhinology hospital in Australia. The classification layer of Inception-V3 was retrained in Python using a transfer learning method to interpret the computed tomography images. Segmentation analysis was also performed in an attempt to increase diagnostic accuracy.ResultsThe trained convolutional neural network was found to have diagnostic accuracy of 81 per cent (95 per cent confidence interval: 73.0–89.0 per cent) with an area under the curve of 0.93.ConclusionA trained convolutional neural network algorithm appears to successfully identify pneumatisation of the middle turbinate with high accuracy. Further studies can be pursued to test its ability in other clinically important anatomical variants in otolaryngology and rhinology.

Download Full-text