Deep learning-enabled detection of hypoxic-ischemic encephalopathy after cardiac arrest in CT scans: a comparative study of 2D and 3D approaches.

Objective: To establish a deep learning model for the detection of hypoxic-ischemic encephalopathy (HIE) features on CT scans and to compare various networks to determine the best input data format. Methods: 168 head CT scans of patients after cardiac arrest were retrospectively identified and classified into two categories: 88 (52.4%) with radiological evidence of severe HIE and 80 (47.6%) without signs of HIE. These images were randomly divided into a training and a test set, and five deep learning models based on based on Densely Connected Convolutional Networks (DenseNet121) were trained and validated using different image input formats (2D and 3D images). Results: All optimized stacked 2D and 3D networks could detect signs of HIE. The networks based on the data as 2D image data stacks provided the best results (S100: AUC: 94%, ACC: 79%, S50: AUC: 93%, ACC: 79%). We provide visual explainability data for the decision making of our AI model using Gradient-weighted Class Activation Mapping. Conclusion: Our proof-of-concept deep learning model can accurately identify signs of HIE on CT images. Comparing different 2D- and 3D-based approaches, most promising results were achieved by 2D image stack models. After further clinical validation, a deep learning model of HIE detection based on CT images could be implemented in clinical routine and thus aid clinicians in characterizing imaging data and predicting outcome.

Combined CT-guided high-dose-rate brachytherapy (CT-HDRBT) and transarterial chemoembolization with irinotecan-loaded microspheres improve local tumor control and progression-free survival in patients with unresectable colorectal liver metastases compared with mono-CT-HDRBT.

Purpose: To compare the effectivity and toxicity of monotherapy with computed tomography-guided high-dose-rate brachytherapy (CT-HDRBT) vs. combination therapy of transarterial chemoembolization with irinotecan (irinotecan-TACE) and CT-HDRBT in patients with large unresectable colorectal liver metastases (CRLM) with a diameter of > 3 cm. Material and methods: Forty-four retrospectively matched patients with unresectable CRLM were treated either with mono-CT-HDRBT or with a combination of irinotecan-TACE and CT-HDRBT (n = 22 in each group). Matching parameters included treatment, disease, and baseline characteristics. National Cancer Institute Common Terminology Criteria for Adverse Events (version 5.0) were used to evaluate treatment toxicity and the Society of Interventional Radiology classification was applied to analyze catheter-related adverse events. Statistical analysis involved Cox regression, Kaplan-Meier estimator, log-rank test, receiver operating characteristic curve analysis, Shapiro-Wilk test, Wilcoxon test, paired sample t-test, and McNemar test. P-values < 0.05 were deemed significant. Results: Combination therapy ensued longer median progression-free survival (PFS: 5/2 months, p = 0.002) and significantly lower local (23%/68%, p < 0.001) and intrahepatic (50%/95%, p < 0.001) progress rates compared with mono-CT-HDRBT after a median follow-up time of 10 months. Additionally, tendencies for longer local tumor control (LTC: 17/9 months, p = 0.052) were found in patients undergoing both interventions. After combination therapy, aspartate and alanine aminotransferase toxicity levels increased significantly, while total bilirubin toxicity levels showed significantly higher increases after monotherapy. No catheter-associated major or minor complications were identified in each cohort. Conclusions: Combining irinotecan-TACE with CT-HDRBT can improve LTC rates and PFS compared with mono-CT-HDRBT in patients with unresectable CRLM. The combination of irinotecan-TACE and CT-HDRBT shows satisfying safety profiles.

Two-Stage Deep Learning Model for Automated Segmentation and Classification of Splenomegaly.

Splenomegaly is a common cross-sectional imaging finding with a variety of differential diagnoses. This study aimed to evaluate whether a deep learning model could automatically segment the spleen and identify the cause of splenomegaly in patients with cirrhotic portal hypertension versus patients with lymphoma disease. This retrospective study included 149 patients with splenomegaly on computed tomography (CT) images (77 patients with cirrhotic portal hypertension, 72 patients with lymphoma) who underwent a CT scan between October 2020 and July 2021. The dataset was divided into a training (n = 99), a validation (n = 25) and a test cohort (n = 25). In the first stage, the spleen was automatically segmented using a modified U-Net architecture. In the second stage, the CT images were classified into two groups using a 3D DenseNet to discriminate between the causes of splenomegaly, first using the whole abdominal CT, and second using only the spleen segmentation mask. The classification performances were evaluated using the area under the receiver operating characteristic curve (AUC), accuracy (ACC), sensitivity (SEN), and specificity (SPE). Occlusion sensitivity maps were applied to the whole abdominal CT images, to illustrate which regions were important for the prediction. When trained on the whole abdominal CT volume, the DenseNet was able to differentiate between the lymphoma and liver cirrhosis in the test cohort with an AUC of 0.88 and an ACC of 0.88. When the model was trained on the spleen segmentation mask, the performance decreased (AUC = 0.81, ACC = 0.76). Our model was able to accurately segment splenomegaly and recognize the underlying cause. Training on whole abdomen scans outperformed training using the segmentation mask. Nonetheless, considering the performance, a broader and more general application to differentiate other causes for splenomegaly is also conceivable.

Evaluation of a Deep Learning Algorithm for Automated Spleen Segmentation in Patients with Conditions Directly or Indirectly Affecting the Spleen.

The aim of this study was to develop a deep learning-based algorithm for fully automated spleen segmentation using CT images and to evaluate the performance in conditions directly or indirectly affecting the spleen (e.g., splenomegaly, ascites). For this, a 3D U-Net was trained on an in-house dataset (n = 61) including diseases with and without splenic involvement (in-house U-Net), and an open-source dataset from the Medical Segmentation Decathlon (open dataset, n = 61) without splenic abnormalities (open U-Net). Both datasets were split into a training (n = 32.52%), a validation (n = 9.15%) and a testing dataset (n = 20.33%). The segmentation performances of the two models were measured using four established metrics, including the Dice Similarity Coefficient (DSC). On the open test dataset, the in-house and open U-Net achieved a mean DSC of 0.906 and 0.897 respectively (p = 0.526). On the in-house test dataset, the in-house U-Net achieved a mean DSC of 0.941, whereas the open U-Net obtained a mean DSC of 0.648 (p < 0.001), showing very poor segmentation results in patients with abnormalities in or surrounding the spleen. Thus, for reliable, fully automated spleen segmentation in clinical routine, the training dataset of a deep learning-based algorithm should include conditions that directly or indirectly affect the spleen.

Feasibility and Safety of CT-Guided High-Dose-Rate Brachytherapy Combined with Transarterial Chemoembolization Using Irinotecan-Loaded Microspheres for the Treatment of Large, Unresectable Colorectal Liver Metastases.

PURPOSE: To evaluate feasibility and safety of combined irinotecan chemoembolization and CT-guided high-dose-rate brachytherapy (HDRBT) in patients with unresectable colorectal liver metastases > 3 cm in diameter. MATERIALS AND METHODS: This prospective study included 23 patients (age, 70 y ± 11.3; 16 men) with 47 liver metastases (size, 62 mm ± 18.7). Catheter-related adverse events were reported per Society of Interventional Radiology classification, and treatment toxicities were reported per Common Terminology Criteria for Adverse Events. Liver-related blood values were analyzed by Wilcoxon test, with P < .05 as significant. Time to local tumor progression, progression-free survival (PFS), and overall survival (OS) were estimated by Kaplan-Meier method. RESULTS: No catheter-related major or minor complications were recorded. Significant differences vs baseline levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT; both P < .001), γ-glutamyltransferase (GGT; P = .013), and hemoglobin (P = .014) were recorded. After therapy, 11 of 23 patients (47.8%) presented with new grade I/II toxicities (bilirubin, n = 3 [13%]; AST, n = 16 [70%]; ALT, n = 18 [78%]; ALP, n = 12 [52%] and hemoglobin, n = 15 [65%]). Moreover, grade III/IV toxicities developed in 10 (43.5%; 1 grade IV): AST, n = 6 (26%), grade III, n = 5; grade IV, n = 1; ALT, n = 3 (13%); GGT, n = 7 (30%); and hemoglobin, n = 1 (4%). However, all new toxicities resolved within 3 months after therapy without additional treatment. Median local tumor control, PFS, and OS were 6, 4, and 8 months, respectively. CONCLUSIONS: Combined irinotecan chemoembolization and CT-guided HDRBT is safe and shows a low incidence of toxicities, which were self-resolving.