A joint convolutional-recurrent neural network with an attention mechanism for detecting intracranial hemorrhage on noncontrast head CT.

To investigate the performance of a joint convolutional neural networks-recurrent neural networks (CNN-RNN) using an attention mechanism in identifying and classifying intracranial hemorrhage (ICH) on a large multi-center dataset; to test its performance in a prospective independent sample consisting of consecutive real-world patients. All consecutive patients who underwent emergency non-contrast-enhanced head CT in five different centers were retrospectively gathered. Five neuroradiologists created the ground-truth labels. The development dataset was divided into the training and validation set. After the development phase, we integrated the deep learning model into an independent center's PACS environment for over six months for assessing the performance in a real clinical setting. Three radiologists created the ground-truth labels of the testing set with a majority voting. A total of 55,179 head CT scans of 48,070 patients, 28,253 men (58.77%), with a mean age of 53.84 ± 17.64 years (range 18-89) were enrolled in the study. The validation sample comprised 5211 head CT scans, with 991 being annotated as ICH-positive. The model's binary accuracy, sensitivity, and specificity on the validation set were 99.41%, 99.70%, and 98.91, respectively. During the prospective implementation, the model yielded an accuracy of 96.02% on 452 head CT scans with an average prediction time of 45 ± 8 s. The joint CNN-RNN model with an attention mechanism yielded excellent diagnostic accuracy in assessing ICH and its subtypes on a large-scale sample. The model was seamlessly integrated into the radiology workflow. Though slightly decreased performance, it provided decisions on the sample of consecutive real-world patients within a minute.

Inter-vendor performance of deep learning in segmenting acute ischemic lesions on diffusion-weighted imaging: a multicenter study.

There is little evidence on the applicability of deep learning (DL) in the segmentation of acute ischemic lesions on diffusion-weighted imaging (DWI) between magnetic resonance imaging (MRI) scanners of different manufacturers. We retrospectively included DWI data of patients with acute ischemic lesions from six centers. Dataset A (n = 2986) and B (n = 3951) included data from Siemens and GE MRI scanners, respectively. The datasets were split into the training (80%), validation (10%), and internal test (10%) sets, and six neuroradiologists created ground-truth masks. Models A and B were the proposed neural networks trained on datasets A and B. The models subsequently fine-tuned across the datasets using their validation data. Another radiologist performed the segmentation on the test sets for comparisons. The median Dice scores of models A and B were 0.858 and 0.857 for the internal tests, which were non-inferior to the radiologist's performance, but demonstrated lower performance than the radiologist on the external tests. Fine-tuned models A and B achieved median Dice scores of 0.832 and 0.846, which were non-inferior to the radiologist's performance on the external tests. The present work shows that the inter-vendor operability of deep learning for the segmentation of ischemic lesions on DWI might be enhanced via transfer learning; thereby, their clinical applicability and generalizability could be improved.

Machine learning-based quantitative texture analysis of conventional MRI combined with ADC maps for assessment of IDH1 mutation in high-grade gliomas.

PURPOSE: To assess the performance of texture analysis of conventional magnetic resonance imaging (MRI) and apparent diffusion coefficient (ADC) maps in predicting IDH1 status in high-grade gliomas (HGG). MATERIALS AND METHODS: A total of 142 patients with HGG were included in the study. IDH1 mutation was present in 48 of 142 HGG (33.8%). Patients were randomly divided into the training cohort (n = 96) and the validation cohort (n = 46). Texture features were extracted via regions of interest on axial T2WI FLAIR, post-contrast T1WI, and ADC maps covering the whole volume of the tumors. The training cohort was used to train the random forest classifier, and the diagnostic performance of the pre-trained model was tested on the validation cohort. RESULTS: The random forest model of conventional MRI sequences and ADC images achieved diagnostic accuracy of 82.2% and 80.4% in predicting IDH1 status in the validation cohorts, respectively. The combined model of T2WI FLAIR, post-contrast T1WI, and ADC images exhibited the highest diagnostic accuracy equating 86.94% in the validation cohort. CONCLUSION: Texture analysis of conventional MRI sequences enhanced by ML analysis can accurately predict the IDH1 status of HGG. Adding textural analysis of ADC maps to conventional MRI results in incremental diagnostic performance.