Interactive content-based image retrieval with deep learning for CT abdominal organ recognition.

Objective.Recognizing the most relevant seven organs in an abdominal computed tomography (CT) slice requires sophisticated knowledge. This study proposed automatically extracting relevant features and applying them in a content-based image retrieval (CBIR) system to provide similar evidence for clinical use.Approach.A total of 2827 abdominal CT slices, including 638 liver, 450 stomach, 229 pancreas, 442 spleen, 362 right kidney, 424 left kidney and 282 gallbladder tissues, were collected to evaluate the proposed CBIR in the present study. Upon fine-tuning, high-level features used to automatically interpret the differences among the seven organs were extracted via deep learning architectures, including DenseNet, Vision Transformer (ViT), and Swin Transformer v2 (SwinViT). Three images with different annotations were employed in the classification and query.Main results.The resulting performances included the classification accuracy (94%-99%) and retrieval result (0.98-0.99). Considering global features and multiple resolutions, SwinViT performed better than ViT. ViT also benefited from a better receptive field to outperform DenseNet. Additionally, the use of hole images can obtain almost perfect results regardless of which deep learning architectures are used.Significance.The experiment showed that using pretrained deep learning architectures and fine-tuning with enough data can achieve successful recognition of seven abdominal organs. The CBIR system can provide more convincing evidence for recognizing abdominal organs via similarity measurements, which could lead to additional possibilities in clinical practice.

Predictive stroke risk model with vision transformer-based Doppler features.

BACKGROUND: Acute stroke is the leading cause of death and disability globally, with an estimated 16 million cases each year. The progression of carotid stenosis reduces blood flow to the intracranial vasculature, causing stroke. Early recognition of ischemic stroke is crucial for disease treatment and management. PURPOSE: A computer-aided diagnosis (CAD) system was proposed in this study to rapidly evaluate ischemic stroke in carotid color Doppler (CCD). METHODS: Based on the ground truth from the clinical examination report, the vision transformer (ViT) features extracted from all CCD images (513 stroke and 458 normal images) were combined in machine learning classifiers to generate the likelihood of ischemic stroke for each image. The pretrained weights from ImageNet reduced the time-consuming training process. The accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve were calculated to evaluate the stroke prediction model. The chi-square test, DeLong test, and Bonferroni correction for multiple comparisons were applied to deal with the type-I error. Only p values equal to or less than 0.00125 were considered to be statistically significant. RESULTS: The proposed CAD system achieved an accuracy of 89%, a sensitivity of 94%, a specificity of 84%, and an area under the receiver operating characteristic curve of 0.95, outperforming the convolutional neural networks AlexNet (82%, p < 0.001), Inception-v3 (78%, p < 0.001), ResNet101 (84%, p < 0.001), and DenseNet201 (85%, p < 0.01). The computational time in model training was only 30 s, which would be efficient and practical in clinical use. CONCLUSIONS: The experiment shows the promising use of CCD images in stroke estimation. Using the pretrained ViT architecture, the image features can be automatically and efficiently generated without human intervention. The proposed CAD system provides a rapid and reliable suggestion for diagnosing ischemic stroke.

Computer-aided diagnosis of ischemic stroke using multi-dimensional image features in carotid color Doppler.

PURPOSE: Stroke is one of the leading causes of disability and mortality. Carotid atherosclerosis is a crucial factor in the occurrence of ischemic stroke. To achieve timely recognition, a computer-aided diagnosis (CAD) system was proposed to evaluate the ischemic stroke patterns in carotid color Doppler (CCD). METHODS: A total of 513 stroke and 458 normal CCD images were collected from 102 stroke and 75 normal patients, respectively. For each image, quantitative histogram, shape, and texture features were extracted to interpret the diagnostic information. In the experiment, a logistic regression classifier with backward elimination and leave-one-out cross validation was used to combine features as a prediction model. RESULTS: The performance of the CAD system using histogram, shape, and texture features achieved accuracies of 87%, 60%, and 87%, respectively. With respect to the combined features, the CAD achieved an accuracy of 89%, a sensitivity of 89%, a specificity of 88%, a positive predictive value of 89%, a negative predictive value of 88%, and Kappa = 0.77, with an area under the receiver operating characteristic curve of 0.94. CONCLUSIONS: Based on the extracted quantitative features in the CCD images, the proposed CAD system provides valuable suggestions for assisting physicians in improving ischemic stroke diagnoses during carotid ultrasound examination.

Rapid Assessment of Acute Ischemic Stroke by Computed Tomography Using Deep Convolutional Neural Networks.

Acute stroke is one of the leading causes of disability and death worldwide. Regarding clinical diagnoses, a rapid and accurate procedure is necessary for patients suffering from acute stroke. This study proposes an automatic identification scheme for acute ischemic stroke using deep convolutional neural networks (DCNNs) based on non-contrast computed tomographic (NCCT) images. Our image database for the classification model was composed of 1254 grayscale NCCT images from 96 patients (573 images) with acute ischemic stroke and 121 normal controls (681 images). According to the consensus of critical stroke findings by two neuroradiologists, a gold standard was established and used to train the proposed DCNN using machine-generated image features. Including the earliest DCNN, AlexNet, the popular Inception-v3, and ResNet-101 were proposed. To train the limited data size, transfer learning with ImageNet parameters was also used. The established models were evaluated by tenfold cross-validation and tested on an independent dataset containing 50 patients with acute ischemic stroke (108 images) and 58 normal controls (117 images) from another institution. AlexNet without pretrained parameters achieved an accuracy of 97.12%, a sensitivity of 98.11%, a specificity of 96.08%, and an area under the receiver operating characteristic curve (AUC) of 0.9927. Using transfer learning, transferred AlexNet, transferred Inception-v3, and transferred ResNet-101 achieved accuracies between 90.49 and 95.49%. Tested with a dataset from another institution, AlexNet showed an accuracy of 60.89%, a sensitivity of 18.52%, and a specificity of 100%. Transferred AlexNet, Inception-v3, and ResNet-101 achieved accuracies of 81.77%, 85.78%, and 80.89%, respectively. The proposed DCNN architecture as a computer-aided diagnosis system showed that training from scratch can generate a customized model for a specific scanner, and transfer learning can generate a more generalized model to provide diagnostic suggestions of acute ischemic stroke to radiologists.

Assessing Ischemic Stroke with Convolutional Image Features in Carotid Color Doppler.

Stroke is a leading cause of disability and death worldwide. Early and accurate recognition of acute stroke is critical for achieving a good prognosis. The novel automated system proposed in this study was based on convolutional neural networks (CNNs), which were used to identify lesion findings on carotid color Doppler (CCD) images in patients with acute ischemic stroke. An image database composed of 1032 CCD images from 106 patients with acute ischemic stroke (549 images) and from 79 normal controls (483 images) was retrospectively analyzed. Taking the consensus of two neuroradiologists as the gold standard, different CNN models with and without transfer learning were evaluated with 10-fold cross-validation. The diagnostic information provided from individual color channels was also explored. AlexNet, which was trained from scratch, achieved an accuracy of 91.67%, a sensitivity of 93.33%, a specificity of 90.20% and an area under the receiver operating characteristic curves (AUC) of 0.9432. Other transferred models achieved accuracies between 77.69% and 83.94%. In channel comparisons, the green channel had the best performance, with an accuracy of 87.50%, a sensitivity of 97.78%, a specificity of 78.43% and an AUC of 0.9507. The proposed CNN architecture, as a computer-aided diagnosis system, suggests using automatic feature extraction from CCD images to predict ischemic stroke. The developed scheme has the potential to provide diagnostic suggestions in clinical use.