Deep Learning Based Automatic Segmentation of the Thoracic Aorta from Chest Computed Tomography in Healthy Korean Adults.

OBJECTIVE: Segmenting the aorta into zones based on anatomical landmarks is a current trend to better understand interventions for aortic dissection or aneurysm. However, comprehensive reference values for aortic zones are lacking. The aim of this study was to establish reference values for aortic size using a fully automated deep learning based segmentation method. METHODS: This retrospective study included 704 healthy adults (mean age 50.6 ± 7.5 years; 407 [57.8%] males) who underwent contrast enhanced chest computed tomography (CT) for health screening. A convolutional neural network (CNN) was trained and applied on 3D CT images for automatic segmentation of the aorta based on the Society for Vascular Surgery/Society of Thoracic Surgeons classification. The CNN generated masks were reviewed and corrected by expert cardiac radiologists. RESULTS: Aortic size was significantly larger in males than in females across all zones (zones 0 - 8, all p < .001). The aortic size in each zone increased with age, by approximately 1 mm per 10 years of age, e.g., 25.4, 26.7, 27.5, 28.8, and 29.8 mm at zone 2 in men in the age ranges of 30 - < 40, 40 - < 50, 50 - < 60, 60 - < 70, and ≥ 70 years, respectively (all p < .001). CONCLUSION: The deep learning algorithm provided reliable values for aortic size in each zone, with automatic masks comparable with manually corrected ones. Aortic size was larger in males and increased with age. These findings have clinical implications for the detection of aortic aneurysms or other aortic diseases.

Staged cerebral and brachiocephalic bypass in a patient with MCA and brachiocephalic steno-occlusion.

Background: Brachiocephalic steal is a rare phenomenon that may lead to hemodynamic abnormalities in the ipsilateral vertebral and carotid arteries. Current treatment includes management of vascular risk factors, endovascular stenting, and surgical management depending on the severity of symptoms. Case Information: We describe a patient with multiple vascular risk factors who presented with mild neurological symptoms and chronic right arm weakness associated with concurrent brachiocephalic steal and right MCA M1 stenosis on neuroimaging. Use of right superficial temporal artery to middle cerebral artery bypass and aorta-right subclavian bypass resulted in improved flow to the right hemisphere on quantitative magnetic resonance angiography and single-photon emission computed tomography, significantly lowering the risk of catastrophic ischemic stroke. Conclusion: Complex steal phenomena increase stroke risk. In cases of high-grade occlusion or advanced symptoms, endovascular or surgical management should be considered for optimal stroke prevention.

Evaluation of multiple deep neural networks for detection of intracranial dural arteriovenous fistula on susceptibility weighted angiography imaging.

BACKGROUND: The natural history of intracranial dural arteriovenous fistula (DAVF) is variable and early diagnosis is crucial in order to positively impact the clinical course of aggressive DAVF. Artificial intelligence (AI) based techniques can be promising in this regard, and in this study, we used various deep neural network (DNN) architectures to determine whether DAVF could be reliably identified on susceptibility-weighted angiography images (SWAN). MATERIALS AND METHODS: A total of 3965 SWAN image slices from 30 digital subtraction angiographically proven DAVF patients and 4380 SWAN image slices from 40 age-matched patients with normal MRI findings as control group were included. The images were categorized as either DAVF or normal and the data was trained using various DNN such as VGG-16, EfficientNet-B0, and ResNet-50. RESULTS: Various DNN architectures showed the accuracy of 95.96% (VGG-16), 91.75% (EfficientNet-B0), and 86.23% (ResNet-50) on the SWAN image dataset. ROC analysis yielded an area under the curve of 0.796 (p < .001), best for VGG-16 model. Criterion of seven consecutive positive slices for DAVF diagnosis yielded a sensitivity of 74.68% with a specificity of 69.15%, while setting eight slices improved the sensitivity to above 80.38%, with a decrease of specificity up to 56.38%. Based on seven consecutive positive slices criteria, EfficientNet-B0 yielded a sensitivity of 73.21% with a specificity of 45.92% and ResNet-50 yielded a sensitivity of 72.39% with a specificity of 67.42%. CONCLUSION: This study shows that DNN can extract discriminative features of SWAN for the classification of DAVF from normal with good accuracy, reasonably good sensitivity and specificity.

Blood perfusion assessment by near-infrared fluorescence angiography of epiploic appendages in prevention of anastomotic leakage after laparoscopic intersphincteric resection for ultra-low rectal cancer: a case-matched study.

BACKGROUND: The role of intraoperative near-infrared fluorescence angiography with indocyanine green in reducing anastomotic leakage (AL) has been demonstrated in colorectal surgery, however, its perfusion assessment mode, and efficacy in reducing anastomotic leakage after laparoscopic intersphincteric resection (LsISR) need to be further elucidated. AIM: Aim was to study near-infrared fluorescent angiography to help identify bowel ischemia to reduce AL after LsISR. MATERIAL AND METHODS: A retrospective case-matched study was conducted in one referral center. A total of 556 consecutive patients with ultra-low rectal cancer including 140 patients with fluorescence angiography of epiploic appendages (FAEA)were enrolled. Perfusion assessment by FAEA in the monochrome fluorescence mode. Patients were divided into two groups based on perfusion assessment by FAEA. The primary endpoint was the AL rate within 6 months, and the secondary endpoint was the structural sequelae of anastomotic leakage (SSAL). RESULTS: After matching, the study group (n = 109) and control group (n = 190) were well-balanced. The AL rate in the FAEA group was lower before (3.6% vs. 10.1%, P = 0.026) and after matching (3.7% vs. 10.5%, P = 0.036). Propensity scores matching analysis (OR 0.275, 95% CI 0.035-0.937, P 0.039), inverse probability of treatment weighting (OR 0.814, 95% CI 0.765-0.921, P 0.002), and regression analysis (OR 0.298, 95% CI 0.112-0.790, P = 0.015), showed that FAEA was an independent protector factor for AL. This technique can significantly shorten postoperative hospital stay [9 (6-13) vs. 10 (8-13), P = 0.024] and reduce the risk of SSAL (1.4% vs. 6.0%, P = 0.029). CONCLUSIONS: Perfusion assessment by FAEA can achieve better visualization in LsISR and reduce the incidence of AL, subsequently avoiding SSAL after LsISR.

Association between glycemia and multi-vessel lesion in participants undergoing coronary angiography: a cross-sectional study.

Background: This study aims to elucidate the association between glycemia and the occurrence of multi-vessel lesions in participants undergoing coronary angiography. Methods: We analyzed 2,533 patients with coronary artery disease who underwent coronary angiography. Of these, 1,973 patients, identified by the endpoint of multi-vessel lesions, were examined using univariate and multivariate logistic regression analyses to determine the relationship between glycemia levels and multi-vessel lesion occurrence. Results: The analysis included 1,973 participants, among whom 474 patients were identified with coronary multi-vessel lesions. Univariate logistic regression analysis demonstrated a positive correlation between glycemia and the occurrence of coronary multi-vessel lesions (OR 1.04; 95% CI 1.01-1.08; p = 0.02). The adjusted model indicated that for each unit increase in glycemia, the risk of developing coronary multi-vessel lesions increased by 4%, showing a significant correlation (p < 0.05). Subgroup analyses revealed that the impact of glycemia on multi-vessel lesions in patients with PCI varied according to gender, age, and smoking status, with the effect being more pronounced in men, older patients, and smokers. Conclusion: Our findings establish a significant association between glycemia and the incidence of multi-vessel lesions, particularly pronounced in male patients, individuals over 45, and smokers.

Clinical and computed tomography angiography characteristics of infected vs. non-infected abdominal aortic aneurysm: a comparative study.

PURPOSE: The aim of this study was to investigate the clinical and multi-slice spiral computed tomography angiography (MSCTA) characteristics for the diagnosis of infected AAA. METHODS: This retrospective comparative study included patients who were diagnosed with AAA at our hospital between January 2014 and May 2023. RESULTS: A total of 40 patients were included, comprising 20 with infected AAA and 20 with non-infected AAA. Patients with infected AAA were more likely to be younger (62.9 ± 10.1 vs. 70.0 ± 4.4 years, P = 0.007) and to present with fever [7 (35%) vs. 1 (5%), P = 0.026], pain [15 (75%) vs. 2 (10%), P < 0.001], higher C-reactive protein levels (60.4 ± 57.0 vs. 4.1 ± 2.9 mg/l, P = 0.005), and higher erythrocyte sedimentation rates (47.7 ± 23.4 vs. 15.2 ± 8.3 mm/h, P < 0.001) compared to those with non-infected AAA. Moreover, those with infected AAA exhibited significantly more eccentric saccular morphology [17 (85%) vs. 1 (5%), P = 0.002], a smaller longitudinal-transverse ratio (1.12 ± 0.33 vs. 2.33 ± 0.54, P = 0.001), thicker peri-aneurysmal soft tissue (2.29 ± 1.48 vs. 0.73 ± 0.55 cm, P < 0.001), more lobulated margins [18 (90%) vs. 1 (5%), P = 0.001], lower aortic calcification scores (49 vs. 56, P < 0.001), more pneumatosis [6 (30%) vs. 0 (0%), P = 0.014], more ruptures [15 (75%) vs. 5 (20%), P = 0.002], more blurred peri-abdominal aortic fat spaces [16 (80%) vs. 2 (10%), P = 0.001], more adjacent bone destruction [5 (25%) vs. 0 (0%), P = 0.025], more involvement of the psoas major muscle [8 (40%) vs. 1 (5%), P = 0.005], more lymphadenectasis [8 (40%) vs. 1 (5%), P = 0.020], and less tortuous aortas [2 (10%) vs. 9 (45%), P = 0.034] compared with those with non-infected AAA. CONCLUSION: The clinical manifestations and MSCTA characteristics may differ between infected and non-infected AAA.

Revolutionizing Aneurysm detection: The role of artificial intelligence in reducing rupture rates.

Cerebral aneurysms, affecting 2-5% of the global population, are often asymptomatic and commonly located within the Circle of Willis. A recent study in Neurosurgical Review highlights a significant reduction in the annual rupture rates of unruptured cerebral aneurysms (UCAs) in Japan from 2003 to 2018. By analyzing age-adjusted mortality rates of subarachnoid hemorrhage (SAH) and the number of treated ruptured cerebral aneurysms (RCAs), researchers found a substantial decrease in rupture rates-from 1.44 to 0.87% and from 0.92 to 0.76%, respectively (p < 0.001). This 88% reduction was largely attributed to improved hypertension management. Recent advancements in artificial intelligence (AI) and machine learning (ML) further support these findings. The RAPID Aneurysm software demonstrated high accuracy in detecting cerebral aneurysms on CT Angiography (CTA), while ML algorithms showed promise in predicting aneurysm rupture risk. A meta-analysis indicated that ML models could achieve 83% sensitivity and specificity in rupture prediction. Additionally, deep learning techniques, such as the PointNet + + architecture, achieved an AUC of 0.85 in rupture risk prediction. These technological advancements in AI and ML are poised to enhance early detection and risk management, potentially contributing to the observed reduction in UCA rupture rates and improving patient outcomes.

Successful chemical ablation for intraventricular septal ventricular tachycardia-An alternative approach to the septal branch.

Trans-coronary ethanol ablation for ventricular tachycardia originating from the ventricular septum is effective, but there are cases with no septal perforator from left anterior descending artery. CT and angiography can reveal the optimal vessel.

Cavitary Consolidation in Pulmonary Infarct Secondary to Embolism: A Rare Presentation.

Obstruction of the pulmonary artery or one of its branches, often due to thrombi from the deep veins of the lower extremities can result in a life-threatening condition known as pulmonary embolism. Pulmonary infarction, an unusual complication of pulmonary embolism occurs when the blood supply to lung tissue is obstructed, leading to tissue necrosis. An 80-year-old man presented with a cough, breathlessness, and generalized weakness. He was vitally stable with no oxygen requirement, which could have suggested an infective etiology like pneumonia or tuberculosis. However, the presence of calf tenderness prompted us to perform a venous Doppler ultrasonography, which revealed deep venous thrombosis. This, combined with right atrial and ventricular dilation and moderate pulmonary artery hypertension observed on transthoracic echocardiography (2D ECHO), led us to recommend a CT pulmonary angiography. The angiography revealed an uncommon presentation of pulmonary embolism with multiple pulmonary infarcts. Here, we chronicle an unusual case of pulmonary infarction secondary to pulmonary embolism, which presented radiologically as consolidation with an aseptic cavity, a rare and atypical triple occurrence.

Challenges and Successes in Introducing Coronary CT Angiography in an Emergency Medicine-Run Observation Unit.

This study was designed to analyze the departmental changes in transitioning the Emergency Department (ED)-run Observation Medicine Unit's routine noninvasive cardiac evaluation from the traditional standard-of-care procedures to coronary computed tomography angiography (CCTA).  While the routine use of CCTA for the evaluation of chest pain has been deemed feasible and safe, provider confidence appears apprehensive, and ordering patterns appear reluctant to change.  We conducted a retrospective analysis of data from two risk-matched cohorts of ED patients who presented with symptoms suggestive of acute coronary syndrome (ACS) but without ischemic electrocardiogram (ECG) changes or positive troponin. Endpoints included length of stay, major adverse cardiovascular event (MACE) rates at 28 days, recidivism rate, and downstream findings on coronary catheterization.  The adoption of CCTA led to a significant reduction in the length of stay for patients in the ED-run Observation Medicine Unit. Provider and nursing education initiatives were crucial in overcoming initial resistance and improving the implementation of CCTA. Post-education, there was a marked increase in the volume of CCTA performed and a decrease in the length of stay, enhancing overall departmental throughput.  The results suggest that CCTA offers a reliable and efficient diagnostic alternative to traditional noninvasive tests, with high diagnostic accuracy contributing to faster decision-making and reduced need for invasive procedures. Continuous education for providers and nursing staff was essential to ensure adherence to the new protocol and improve clinical outcomes.  Transitioning to CCTA for routine noninvasive cardiac evaluation in the ED-run Observation Medicine Unit demonstrated significant efficiency and diagnostic accuracy benefits. Successful implementation requires targeted educational efforts to ensure competency and confidence among healthcare providers. The findings support the integration of CCTA into standard clinical practice for the evaluation of chest pain in the emergency setting, with future research needed to validate these results in broader patient populations and assess long-term outcomes.

Performance and clinical utility of an artificial intelligence-enabled tool for pulmonary embolism detection.

PURPOSE: Diagnosing pulmonary embolism (PE) is still challenging due to other conditions that can mimic its appearance, leading to incomplete or delayed management and several inter-observer variabilities. This study evaluated the performance and clinical utility of an artificial intelligence (AI)-based application designed to assist clinicians in the detection of PE on CT pulmonary angiography (CTPA). PATIENTS AND METHODS: CTPAs from 230 US cities acquired on 57 scanner models from 6 different vendors were retrospectively collected. Three US board certified expert radiologists defined the ground truth by majority agreement. The same cases were analyzed by CINA-PE, an AI-driven algorithm capable of detecting and highlighting suspected PE locations. The algorithm's performance at a per-case and per-finding level was evaluated. Furthermore, cases with PE not mentioned in the clinical report but correctly detected by the algorithm were analyzed. RESULTS: A total of 1204 CTPAs (mean age 62.1 years ± 16.6[SD], 44.4 % female, 14.9 % positive) were included in the study. Per-case sensitivity and specificity were 93.9 % (95%CI: 89.3 %-96.9 %) and 94.8 % (95%CI: 93.3 %-96.1 %), respectively. Per-finding positive predictive value was 89.5 % (95%CI: 86.7 %-91.9 %). Among the 196 positive cases, 29 (15.6 %) were not mentioned in the clinical report. The algorithm detected 22/29 (76 %) of these cases, leading to a reduction in the miss rate from 15.6 % to 3.8 % (7/186). CONCLUSIONS: The AI-based application may improve diagnostic accuracy in detecting PE and enhance patient outcomes through timely intervention. Integrating AI tools in clinical workflows can reduce missed or delayed diagnoses, and positively impact healthcare delivery and patient care.

Effectiveness of the proximal medial branch in DIEP flaps for breast reconstruction using Hartrampf zone IV.

BACKGROUND: When performing breast reconstruction using a deep inferior epigastric artery (DIEA) perforator (DIEP) flap, including Hartrampf zone IV, the bipedicled DIEP flap has been argued to be necessary to ensure stable perfusion. However, a proximal medial branch (PMB), which is the most proximal perforator of DIEA, may make it possible to obtain adequate perfusion in a unilateral DIEP flap. This study aimed to clarify the detailed anatomical characteristics of PMB and its potential clinical applications in breast reconstruction. METHODS: This retrospective study was conducted on breast reconstruction using the DIEP flap between May 2020 and July 2023. Data on PMB anatomy were collected from preoperative contrast-enhanced computed tomography angiography, and contralateral perfusion of the flap was estimated using intraoperative indocyanine green angiography. RESULTS: PMB was present in approximately 85% of the cases, arising near the lateral border of the rectus abdominis, branching caudomedially in more than half of the cases, and perforating 2.3 cm laterally and 8.8 cm caudally, on average, from the umbilicus. The average perfusion area of zones II and IV significantly expanded to 96.5% and 74.2%, respectively, when PMB was included in the DIEP flap, and 9 of 22 cases showed contrast extending to the entire zone IV. CONCLUSIONS: The use of the DIEP flap with PMB is a good option for substantial-volume breast reconstruction. When utilizing PMB, it is important to consider its specific anatomy, location of the main perforator, and pedicle length.

Evaluating local open-source large language models for data extraction from unstructured reports on mechanical thrombectomy in patients with ischemic stroke.

BACKGROUND: A study was undertaken to assess the effectiveness of open-source large language models (LLMs) in extracting clinical data from unstructured mechanical thrombectomy reports in patients with ischemic stroke caused by a vessel occlusion. METHODS: We deployed local open-source LLMs to extract data points from free-text procedural reports in patients who underwent mechanical thrombectomy between September 2020 and June 2023 in our institution. The external dataset was obtained from a second university hospital and comprised consecutive cases treated between September 2023 and March 2024. Ground truth labeling was facilitated by a human-in-the-loop (HITL) approach, with time metrics recorded for both automated and manual data extractions. We tested three models-Mixtral, Qwen, and BioMistral-assessing their performance on precision, recall, and F1 score across 15 clinical categories such as National Institute of Health Stroke Scale (NIHSS) scores, occluded vessels, and medication details. RESULTS: The study included 1000 consecutive reports from our primary institution and 50 reports from a secondary institution. Mixtral showed the highest precision, achieving 0.99 for first series time extraction and 0.69 for occluded vessel identification within the internal dataset. In the external dataset, precision ranged from 1.00 for NIHSS scores to 0.70 for occluded vessels. Qwen showed moderate precision with a high of 0.85 for NIHSS scores and a low of 0.28 for occluded vessels. BioMistral had the broadest range of precision, from 0.81 for first series times to 0.14 for medication details. The HITL approach yielded an average time savings of 65.6% per case, with variations from 45.95% to 79.56%. CONCLUSION: This study highlights the potential of using LLMs for automated clinical data extraction from medical reports. Incorporating HITL annotations enhances precision and also ensures the reliability of the extracted data. This methodology presents a scalable privacy-preserving option that can significantly support clinical documentation and research endeavors.

Inpatient outcomes of NSTEMI among patients with immune thrombocytopenia: a propensity matched national study.

Patients with immune thrombocytopenia (ITP) admitted for non-ST elevation myocardial infarction (NSTEMI) present a unique therapeutic challenge due to the increased risk of bleeding with antiplatelet and anticoagulation therapies. There is limited evidence studying hospital mortality and complications in this population. The study included a patient cohort from the 2018-2021 National Inpatient Sample database. Propensity score matched NSTEMI patients with and without ITP using a 1:1 matching ratio. Outcomes analyzed were in-hospital mortality, rates of diagnostic angiogram, percutaneous coronary intervention (PCI), acute kidney injury (AKI), congestive heart failure (CHF), cardiogenic shock, cardiac arrest, mechanical ventilation, tracheal intubation, ventricular tachycardia (VT), ventricular fibrillation (VF), major bleeding, need for blood and platelet transfusion, length of stay (LOS), and total hospitalization charges. A total of 1,699,020 patients met inclusion criteria (660,490 females [39%], predominantly Caucasian 1,198,415 (70.5%); mean [SD] age 67, [3.1], including 2,615 (0.1%) patients with ITP. Following the propensity matching, 1,020 NSTEMI patients with and without ITP were matched. ITP patients had higher rates of inpatient mortality (aOR 1.98, 95% CI 1.11-3.50, p 0.02), cardiogenic shock, AKI, mechanical ventilation, tracheal intubation, red blood cells and platelet transfusions, longer LOS, and higher total hospitalization charges. The rates of diagnostic angiogram, PCI, CHF, VT, VF, and major bleeding were not different between the two groups. Patients with ITP demonstrated higher odds of in-hospital mortality for NSTEMI and need for platelet transfusion with no difference in rates of diagnostic angiogram or PCI.

Risk scores and coronary artery disease in patients with suspected acute coronary syndrome and intermediate cardiac troponin concentrations.

BACKGROUND: Guidelines recommend the use of risk scores to select patients for further investigation after myocardial infarction has been ruled out but their utility to identify those with coronary artery disease is uncertain. METHODS: In a prospective cohort study, patients with intermediate high-sensitivity cardiac troponin I concentrations (5 ng/L to sex-specific 99th percentile) in whom myocardial infarction was ruled out were enrolled and underwent coronary CT angiography (CCTA) after hospital discharge. History, ECG, Age, Risk factors, Troponin (HEART), Emergency Department Assessment of Chest Pain Score (EDACS), Global Registry of Acute Coronary Event (GRACE), Thrombolysis In Myocardial Infarction (TIMI), Systematic COronary Risk Evaluation 2 and Pooled Cohort Equation risk scores were calculated and the odds ratio (OR) and diagnostic performance for obstructive coronary artery disease were determined using established thresholds. RESULTS: Of 167 patients enrolled (64±12 years, 28% female), 29.9% (50/167) had obstructive coronary artery disease. The odds of having obstructive disease were increased for all scores with the lowest and highest increase observed for an EDACS score ≥16 (OR 2.2 (1.1-4.6)) and a TIMI risk score ≥1 (OR 12.9 (3.0-56.0)), respectively. The positive predictive value (PPV) was low for all scores but was highest for a GRACE score >88 identifying 39% as high risk with a PPV of 41.9% (30.4-54.2%). The negative predictive value (NPV) varied from 77.3% to 95.2% but was highest for a TIMI score of 0 identifying 26% as low risk with an NPV of 95.2% (87.2-100%). CONCLUSIONS: In patients with intermediate cardiac troponin concentrations in whom myocardial infarction has been excluded, clinical risk scores can help identify patients with and without coronary artery disease, although the performance of established risk thresholds is suboptimal for utilisation in clinical practice. TRIAL REGISTRATION NUMBER: NCT04549805.

Coronary CT Angiography-Based Radiomics to Predict Vessel-Specific Ischemia by Stress Dynamic CT Myocardial Perfusion Imaging.

RATIONALE AND OBJECTIVES: To investigate the predictive value of coronary CT angiography (CCTA)-based radiomics for vessel-specific ischemia by stress dynamic CT myocardial perfusion imaging (MPI). MATERIALS AND METHODS: Patients with typical angina/atypical angina/non-angina chest pain who underwent both stress dynamic CT MPI and CCTA scans were retrospectively enrolled. The following models were constructed for ischemic prediction using logistic regression and CCTA-derived quantitative and radiomic features: plaque quantitative model, lumen quantitative model, CT-fractional flow reserve (CT-FFR) model, integrative quantitative model, plaque radiomic model, peri-coronary adipose tissue (pCAT) radiomic model, integrative radiomic model, and quantitative and radiomic fusion model. A relative myocardial blood flow ≤ 0.75 on stress dynamic CT MPI was considered ischemic. The models' performances were quantified by the area under the receiver-operating characteristic curve (AUC). RESULTS: 386 coronary vessels (stenosis grade: 25%∼75%; training set: 200 [ischemia/non-ischemia=96/104]; test set:186 [ischemia/non-ischemia=79/107]) from 326 patients were included. The plaque radiomic model (training/test set: AUC=0.81/0.80) outperformed (p < .05) both the plaque quantitative (training/test set: AUC=0.71/0.68) model and the lumen quantitative (training/test set: AUC=0.69/0.65) model in identifying ischemia. The integrative radiomic model (training/test set: AUC=0.83/0.82) outperformed (p < .05) the CT-FFR model (training/test set: AUC=0.74/0.73) for ischemic prediction. The quantitative and radiomic fusion model (training/test set: AUC=0.86/0.84) outperformed (p < .05) the integrative quantitative model (training/test set: AUC=0.79/0.77) for ischemic detection. CONCLUSION: The plaque and pCAT radiomic features were superior to the plaque and pCAT quantitative features in predicting ischemia and the addition of the radiomic features to the quantitative features for ischemic identification yielded incremental discriminatory value.

Improved microvascular imaging with optical coherence tomography using 3D neural networks and a channel attention mechanism.

Skin microvasculature is vital for human cardiovascular health and thermoregulation, but its imaging and analysis presents significant challenges. Statistical methods such as speckle decorrelation in optical coherence tomography angiography (OCTA) often require multiple co-located B-scans, leading to lengthy acquisitions prone to motion artefacts. Deep learning has shown promise in enhancing accuracy and reducing measurement time by leveraging local information. However, both statistical and deep learning methods typically focus solely on processing individual 2D B-scans, neglecting contextual information from neighbouring B-scans. This limitation compromises spatial context and disregards the 3D features within tissue, potentially affecting OCTA image accuracy. In this study, we propose a novel approach utilising 3D convolutional neural networks (CNNs) to address this limitation. By considering the 3D spatial context, these 3D CNNs mitigate information loss, preserving fine details and boundaries in OCTA images. Our method reduces the required number of B-scans while enhancing accuracy, thereby increasing clinical applicability. This advancement holds promise for improving clinical practices and understanding skin microvascular dynamics crucial for cardiovascular health and thermoregulation.

In vitro and in vivo optimized reconstruction for low-keV virtual monoenergetic photon-counting detector CT angiography of lower legs.

BACKGROUND: Lower extremity peripheral artery disease frequently presents with calcifications which reduces the accuracy of computed tomography (CT) angiography, especially below-the-knee. Photon-counting detector (PCD)-CT offers improved spatial resolution and less calcium blooming. We aimed to identify the optimal reconstruction parameters for PCD-CT angiography of the lower legs. METHODS: Tubes with different diameters (1-5 mm) were filled with different iodine concentrations and scanned in a water container. Images were reconstructed with 0.4 mm isotropic resolution using a quantitative kernel at all available sharpness levels (Qr36 to Qr76) and using different levels of quantum iterative reconstruction (QIR-2-4). Noise and image sharpness were determined for all reconstructions. Additionally, CT angiograms of 20 patients, reconstructed with a medium (Qr44), sharp (Qr60), and ultrasharp (Qr72) kernel at QIR-2-4, were evaluated by three readers assessing noise, delineation of plaques and vessel walls, and overall quality. RESULTS: In the phantom study, increased kernel sharpness led to higher image noise (e.g., 16, 38, 77 HU for Qr44, Qr60, Qr72, and QIR-3). Image sharpness increased with increasing kernel sharpness, reaching a plateau at the medium-high level 60. Higher QIR levels decreased image noise (e.g., 51, 38, 25 HU at QIR-2-4 and Qr60) without reducing vessel sharpness. The qualitative in vivo results confirmed these findings: the sharp kernel (Qr60) with the highest QIR yielded the best overall quality. CONCLUSION: The combination of a sharpness level optimized reconstruction kernel (Qr60) and the highest QIR level yield the best image quality for PCD-CT angiography of the lower legs when reconstructed at 0.4-mm resolution. RELEVANCE STATEMENT: Using high-resolution PCD-CT angiography with optimized reconstruction parameters might improve diagnostic accuracy and confidence in peripheral artery disease of the lower legs. KEY POINTS: Effective exploitation of the potential of PCD-CT angiography requires optimized reconstruction parameters. Too soft or too sharp reconstruction kernels reduce image quality. The highest level of quantum iterative reconstruction provides the best image quality.