RESUMEN
Background: We compared the ischemic core and hypoperfused tissue volumes estimated by RAPID and JLK-CTP, a newly developed automated computed tomography perfusion (CTP) analysis package. We also assessed agreement between ischemic core volumes by two software packages against early follow-up infarct volumes on diffusion-weighted images (DWI). Methods: This retrospective study analyzed 327 patients admitted to a single stroke center in Korea from January 2021 to May 2023, who underwent CTP scans within 24 h of onset. The concordance correlation coefficient (ρ) and Bland-Altman plots were utilized to compare the volumes of ischemic core and hypoperfused tissue volumes between the software packages. Agreement with early (within 3 h from CTP) follow-up infarct volumes on diffusion-weighted imaging (n = 217) was also evaluated. Results: The mean age was 70.7 ± 13.0 and 137 (41.9%) were female. Ischemic core volumes by JLK-CTP and RAPID at the threshold of relative cerebral blood flow (rCBF) < 30% showed excellent agreement (ρ = 0.958 [95% CI, 0.949 to 0.966]). Excellent agreement was also observed for time to a maximum of the residue function (T max) > 6 s between JLK-CTP and RAPID (ρ = 0.835 [95% CI, 0.806 to 0.863]). Although early follow-up infarct volume showed substantial agreement in both packages (JLK-CTP, ρ = 0.751 and RAPID, ρ = 0.632), ischemic core volumes at the threshold of rCBF <30% tended to overestimate ischemic core volumes. Conclusion: JLK-CTP and RAPID demonstrated remarkable concordance in estimating the volumes of the ischemic core and hypoperfused area based on CTP within 24 h from onset.
RESUMEN
Introduction: We developed and externally validated a fully automated algorithm using deep learning to detect large vessel occlusion (LVO) in computed tomography angiography (CTA). Method: A total of 2,045 patients with acute ischemic stroke who underwent CTA were included in the development of our model. We validated the algorithm using two separate external datasets: one with 64 patients (external 1) and another with 313 patients (external 2), with ischemic stroke. In the context of current clinical practice, thrombectomy amenable vessel occlusion (TAVO) was defined as an occlusion in the intracranial internal carotid artery (ICA), or in the M1 or M2 segment of the middle cerebral artery (MCA). We employed the U-Net for vessel segmentation on the maximum intensity projection images, followed by the application of the EfficientNetV2 to predict TAVO. The algorithm's diagnostic performance was evaluated by calculating the area under the receiver operating characteristics curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Results: The mean age in the training and validation dataset was 68.7 ± 12.6; 56.3% of participants were men, and 18.0% had TAVO. The algorithm achieved AUC of 0.950 (95% CI, 0.915-0.971) in the internal test. For the external datasets 1 and 2, the AUCs were 0.970 (0.897-0.997) and 0.971 (0.924-0.990), respectively. With a fixed sensitivity of 0.900, the specificities and PPVs for the internal test, external test 1, and external test 2 were 0.891, 0.796, and 0.930, and 0.665, 0.583, and 0.667, respectively. The algorithm demonstrated a sensitivity and specificity of approximately 0.95 in both internal and external datasets, specifically for cases involving intracranial ICA or M1-MCA occlusion. However, the diagnostic performance was somewhat reduced for isolated M2-MCA occlusion; the AUC for the internal and combined external datasets were 0.903 (0.812-0.944) and 0.916 (0.816-0.963), respectively. Conclusion: We developed and externally validated a fully automated algorithm that identifies TAVO. Further research is needed to evaluate its effectiveness in real-world clinical settings. This validated algorithm has the potential to assist early-career physicians, thereby streamlining the treatment process for patients who can benefit from endovascular treatment.
RESUMEN
Background and purpose: Multiple attempts at intracranial hemorrhage (ICH) detection using deep-learning techniques have been plagued by clinical failures. We aimed to compare the performance of a deep-learning algorithm for ICH detection trained on strongly and weakly annotated datasets, and to assess whether a weighted ensemble model that integrates separate models trained using datasets with different ICH improves performance. Methods: We used brain CT scans from the Radiological Society of North America (27,861 CT scans, 3,528 ICHs) and AI-Hub (53,045 CT scans, 7,013 ICHs) for training. DenseNet121, InceptionResNetV2, MobileNetV2, and VGG19 were trained on strongly and weakly annotated datasets and compared using independent external test datasets. We then developed a weighted ensemble model combining separate models trained on all ICH, subdural hemorrhage (SDH), subarachnoid hemorrhage (SAH), and small-lesion ICH cases. The final weighted ensemble model was compared to four well-known deep-learning models. After external testing, six neurologists reviewed 91 ICH cases difficult for AI and humans. Results: InceptionResNetV2, MobileNetV2, and VGG19 models outperformed when trained on strongly annotated datasets. A weighted ensemble model combining models trained on SDH, SAH, and small-lesion ICH had a higher AUC, compared with a model trained on all ICH cases only. This model outperformed four deep-learning models (AUC [95% C.I.]: Ensemble model, 0.953[0.938-0.965]; InceptionResNetV2, 0.852[0.828-0.873]; DenseNet121, 0.875[0.852-0.895]; VGG19, 0.796[0.770-0.821]; MobileNetV2, 0.650[0.620-0.680]; p < 0.0001). In addition, the case review showed that a better understanding and management of difficult cases may facilitate clinical use of ICH detection algorithms. Conclusion: We propose a weighted ensemble model for ICH detection, trained on large-scale, strongly annotated CT scans, as no model can capture all aspects of complex tasks.