BACKGROUND:Adhesive capsulitis is a debilitating shoulder disorder characterized by pain and restricted motion. While arthroscopic capsular release is a well-established surgical treatment, transarterial musculoskeletal embolization (TAME) has recently emerged as a minimally invasive alternative. The aim of this study is to compare the short-term clinical outcomes of arthroscopic capsular release and TAME in patients with adhesive capsulitis of the shoulder. MATERIAL AND METHODS:We retrospectively analyzed 40 patients with refractory adhesive capsulitis who underwent either arthroscopic capsular release (n = 19) or TAME (n = 21) between August 2021 and August 2024. Pain severity (numeric rating scale), range of motion (ROM), and Constant-Murley scores were assessed at baseline and at 3 and 6 months postprocedure. The minimal clinically important difference was calculated to evaluate meaningful clinical improvement. RESULTS:Both groups demonstrated significant improvement in pain, ROM, and Constant-Murley scores at 3 and 6 months compared with baseline. At 3 months, the TAME group showed significantly lower numeric rating scale scores (2.6 vs. 4.7, P < .001) and higher rates of minimal clinically important difference achievement in pain, anterior forward elevation, and Constant-Murley score. However, by 6 months, no significant differences were observed in any outcome measures between groups. CONCLUSION:Both arthroscopic capsular release and TAME were associated with meaningful improvements in pain, ROM, and function in patients with adhesive capsulitis. In this cohort, TAME may be associated with earlier pain relief and earlier attainment of clinically meaningful improvements.

Objective: To assess the effect of transarterial embolization (TAE) for adhesive capsulitis (AC) by evaluating clinical outcomes and changes in inflammation using magnetic resonance imaging (MRI). Materials and Methods: Patients who had undergone TAE between August 2020 and August 2023 for AC refractory to conservative treatments without any invasive procedures for more than 3 months, and had undergone baseline and 3-month post-AC follow-up contrast-enhanced MRI evaluations, were included. A suspension mixture of 500 mg imipenem/cilastatin in 10 mL of iodinated contrast agent was used for TAE. MRI results were analyzed to assess periarticular capsule/ligament inflammation. Clinical assessments included pain scores using the numeric rating scale (NRS) and functional scores using the quick disabilities of the arm, shoulder, and hand (Quick DASH) questionnaire. Results: Twenty-five patients (female:male, 14:11; age, 54.9 +/- 7.1 years) were included. Significant reductions in average NRS pain scores as well as improvements in Quick DASH scores and range of motion, including anterior flexion and abduction, were observed at 1, 3, and 6 months after TAE (all P < 0.001). MRI analyses revealed that TAE significantly decreased the grades of axillary recess capsule enhancement, rotator interval (RI) capsule T2 signal intensity, and RI capsule enhancement (all P <= 0.004). Conclusion: TAE may be an effective and safe therapeutic approach for AC refractory to conservative treatments, alleviating pain and supporting functional recovery. The observed MRI findings suggest that the effectiveness of TAE for AC may be attributed to the reduction of inflammation and the elimination of angiogenesis.

Background/Objectives: Accurate prognostication for interstitial lung disease (ILD) remains challenging, limiting clinicians’ ability to optimize patient management strategies. This study aimed to evaluate the prognostic value of baseline quantitative CT-derived indices, including total lung volume (TLVcm3), normal lung volume% (NLV%), and fibrotic lung volume% (FLV%), for predicting three-year mortality in interstitial lung disease (ILD) patients. Methods: A total of 101 ILD patients were retrospectively enrolled. Baseline CT-derived indices, including TLVcm3, NLV% (−950 to −700 HU), and FLV% (−600 to +50 HU), were measured on chest CT. Baseline forced vital capacity(FVC)% predicted and diffuse capacity of lungs for carbon monoxide (DLCO)% predicted were collected. Survival analysis used Kaplan–Meier’s curves and log-rank tests. Uni- and multivariate Cox’s proportional hazards regression were performed. Pearson’s correlation was used between CT-derived indices, FVC% predicted, and DLCO% predicted. Results: During 3-year follow-up, 30 of 101 patients (29.70%) died. Deceased patients had a significantly lower baseline NLV% (59.27% ± 7.61% vs. 65.02% ± 7.82%, p = 0.001) and a higher FLV% (17.64% ± 7.98% vs. 13.34% ± 7.48%, p = 0.011) compared with survivors. Multivariate analysis identified baseline NLV% (adjusted hazard ratio 0.88, 95% CI: 0.78–0.99, p = 0.034) and DLCO% predicted (adjusted hazard ratio 0.97, 95% CI: 0.95–0.99, p = 0.007) as independent predictors of three-year mortality. Patients with NLV% ≤ 64.15 and FLV% ≥ 14.12 showed significantly worse survival outcomes (21.78% vs. 7.92%, p < 0.001;19.80% (20/101) vs. 9.90% (10/101), p < 0.001). CT-derived indices moderately correlated with FVC% predicted and DLCO% predicted. Conclusions: Baseline FLV% ≥ 14.12 and NLV% ≤ 64.15 can effectively stratify and differentiate outcomes in ILD patients. Baseline NLV% has the potential as a prognostic indicator for 3-year survival in ILD.

Transarterial microembolization (TAME) is an innovative approach to treating chronic musculoskeletal pain. During the procedure, intraprocedural vasospasm, presenting juxta-catheter stenosis, and poor distal artery perfusion and flow through the anastomosis commonly pose challenges. Difficulty of the catheter reaching the target vessel and restricted drug delivery are possible consequences. To address these issues, transcatheter nitroglycerin (NTG) infusion at the extremity’s small-sized artery has been introduced. We investigated patients who underwent the TAME procedure wherein NTG was employed and sourced from two institutional settings. Transcatheter NTG infusion was administered in seven instances of chronic musculoskeletal pain manifesting with intraprocedural vasospasm during TAME procedures. Patient profiles, preprocedural imaging, fluoroscopic findings, adverse events, and Numerical Rating Scale (NRS) scores were evaluated. As a result, all seven cases experiencing intraprocedural vasospasm exhibited rapid responses to transcatheter NTG infusion. Angiography conducted pre- and post-infusion demonstrated increased vessel size, resolved proximal catheter stenosis, and improved distal perfusion. One case presented an adverse effect of self-limited transient hypotension. The NRS scores decreased following the procedure. Transcatheter NTG infusion at the extremity’s small-sized muscular artery can be an effective technique for resolving intraprocedural vasospasm in TAME procedures, irrespective of the target diseases and approach vessels.

BackgroundTransarterial microembolization (TAME) is a minimally invasive treatment for chronic musculoskeletal disorders. Identifying angiogenesis and the supplying vessels of the target joint is important but challenging. Although magnetic resonance imaging (MRI) is commonly used to diagnose musculoskeletal diseases, it typically excludes vascular imaging. Dynamic contrast-enhanced magnetic resonance angiography (DCE-MRA) has the ability to visualize lesion angiogenesis, identify supplying vessels, and evaluate the vasculature anatomy. We propose that incorporating DCE-MRA into pre-procedural assessments can help identify the culprit vessels, arterial anatomy, and variant assessment of the target joint before TAME.Materials and methodsWe investigated six cases, each presenting pain in different body parts: shoulder adhesive capsulitis, trapezius myalgia, combined tennis and golf elbow, knee osteoarthritis, refractory knee pain after osteotomy, and plantar fasciitis. All patients underwent MRI with DCE-MRA before undergoing TAME. DCE-MRA was performed using either 1.5 T or 3 T MRI scanners, employing 3D-TRICKS or 4D-TRAK XD techniques. The numerical rating scale for pain was evaluated at one, three, and six months after the procedure, and any adverse events were recorded over the entire six-month follow-up period.ResultsPre-procedural DCE-MRA helped to visualize angiogenesis at the lesion site in all patients and identify the supplying vessels, arterial vasculature anatomy, and branching variants. These findings corroborated the subsequent digital subtraction angiography (DSA) findings obtained during TAME. All patients experienced pain reduction and functional improvement after TAME without any complications. The average pain score reduced significantly after TAME treatment (p < 0.05). Two patients underwent a second MRI and DCE-MRA at the six-month follow-up and showed a significant reduction in angiogenesis.ConclusionDCE-MRA offers a valuable pre-procedural assessment tool for TAME procedures by facilitating the visualization of angiogenesis at the lesion site, supplying vessels, and arterial anatomic variants, including the variable orifice of the supplying branches. This information can potentially improve patient selection and pre-procedural planning, leading to better outcomes and reduced risk of complications.

Chronic musculoskeletal pain imposes immense suffering and diminishes the quality of life for millions of patients worldwide; the pain persists despite the use of standard conservative treatments. Increases in our understanding of the pathophysiological mechanisms underlying musculoskeletal disorders indicate the involvement of inappropriate angiogenesis. Accordingly, the resulting neovessels are the target of emerging treatments for chronic musculoskeletal pain, including transarterial embolization. The use of this noninvasive procedure to treat pain refractory to standard therapy in a variety of musculoskeletal conditions is the focus of numerous recent investigations. Here, we describe the pathophysiological indications for the use of transarterial embolization and summarize the findings of studies investigating its use in a variety of histopathological conditions and anatomical sites.

Purpose: To evaluate the short-term outcomes of genicular artery embolization (GAE) for knee osteoarthritis (OA) with and without bone marrow lesion (BML) and/or subchondral insufficiency fracture of the knee (SIFK).Materials and Methods: This single-institution prospective observational pilot study analyzed 24 knees in 22 patients with mild to moderate knee OA, including 8 knees without BML, 13 knees with BML, and 3 knees with both BML and SIFK. The area and volume of BMLs on magnetic resonance images were measured before and after GAE. Baseline and postoperative pain and physical function were assessed using the visual analog scale (VAS) and Western Ontario and McMaster Uni-versities Osteoarthritis Index (WOMAC).Results: GAE significantly reduced the BML area and volume 3 months after embolization in the knees with BML (both P < .0005). GAE significantly decreased the VAS scores at 3 and 6 months after embolization in patients without BML (both P = .04) and those with BML (both P = .01). GAE also lowered the WOMAC scores 3 months after embolization in patients without and with BML (P = .02 and P = .0002, respectively). However, GAE did not significantly alter the BML area and volume (both P = .25), VAS scores (P = 1.00), and WOMAC scores (P = .08) in patients with BML and SIFK at 3 months after GAE.Conclusions: This observational pilot study suggested that GAE effectively reduces the BML area and volume and improves pain and physical function in patients with knee OA accompanied by BML but is inefficacious in those with both BML and SIFK.

We compared the accuracy of the fluoroscopic dorsal tangential view (DTV) and an ultrasound (US) examination in detecting dorsal screw penetration during volar distal radius plating. In six fresh cadaveric distal radii, seven periarticular locking screws in two rows for each plate were inserted according to the measured length using a depth gauge and then replaced with another that was 1 and 2 mm longer, respectively. The actual protruded length of each screw was determined using computed tomography (CT) images. The accuracy of US and DTV measurements was determined using the intraclass correlation coefficient (ICC), as both measurements were compared with CT measurements. The ICC of US and DTV was 0.96 and 0.75, respectively, for all screws. After excluding the data for proximal-row screws, the ICC of US remained unchanged at 0.96, and that of DTV improved to 0.86. The ICC of US was significantly higher than that of DTV (p < 0.01). US had a 100% detection rate for screw protrusion of more than 1.0 mm. US examination showed excellent consistency with CT measurements and its accuracy was not affected by screw location. US might thus be a practical tool for detecting dorsal cortex screw penetration during volar distal radius plating.

In this study, we modified the previously proposed X2CT-GAN to build a 2Dto3D-GAN of the spine. This study also incorporated the radiologist's perspective in the adjustment of input signals to prove the feasibility of the automatic production of three-dimensional (3D) structures of the spine from simulated bi-planar two-dimensional (2D) X-ray images. Data from 1012 computed tomography (CT) studies of 984 patients were retrospectively collected. We tested this model under different dataset sizes (333, 666, and 1012) with different bone signal conditions to observe the training performance. A 10-fold cross-validation and five metrics-Dice similarity coefficient (DSC) value, Jaccard similarity coefficient (JSC), overlap volume (OV), and structural similarity index (SSIM)- were applied for model evaluation. The optimal mean values for DSC, JSC, OV, SSIM anteroposterior (AP), and SSIM Lateral (Lat) were 0.8192, 0.6984, 0.8624, 0.9261, and 0.9242, respectively. There was a significant improvement in the training performance under empirically enhanced bone signal conditions and with increasing training dataset sizes. These results demonstrate the potential of the clinical implantation of GAN for automatic production of 3D spine images from 2D images. This prototype model can serve as a foundation in future studies applying transfer learning for the development of advanced medical diagnostic techniques.

We aimed to set up an Automated Radiology Alert System (ARAS) for the detection of pneumothorax in chest radiographs by a deep learning model, and to compare its efficiency and diagnostic performance with the existing Manual Radiology Alert System (MRAS) at the tertiary medical center. This study retrospectively collected 1235 chest radiographs with pneumothorax labeling from 2013 to 2019, and 337 chest radiographs with negative findings in 2019 were separated into training and validation datasets for the deep learning model of ARAS. The efficiency before and after using the model was compared in terms of alert time and report time. During parallel running of the two systems from September to October 2020, chest radiographs prospectively acquired in the emergency department with age more than 6 years served as the testing dataset for comparison of diagnostic performance. The efficiency was improved after using the model, with mean alert time improving from 8.45 min to 0.69 min and the mean report time from 2.81 days to 1.59 days. The comparison of the diagnostic performance of both systems using 3739 chest radiographs acquired during parallel running showed that the ARAS was better than the MRAS as assessed in terms of sensitivity (recall), area under receiver operating characteristic curve, and F1 score (0.837 vs. 0.256, 0.914 vs. 0.628, and 0.754 vs. 0.407, respectively), but worse in terms of positive predictive value (PPV) (precision) (0.686 vs. 1.000). This study had successfully designed a deep learning model for pneumothorax detection on chest radiographs and set up an ARAS with improved efficiency and overall diagnostic performance.

The aim of the study was to use a previously proposed mask region–based convolutional neural network (Mask R-CNN) for automatic abnormal liver density detection and segmentation based on hepatocellular carcinoma (HCC) computed tomography (CT) datasets from a radiological perspective. Training and testing datasets were acquired retrospectively from two hospitals of Taiwan. The training dataset contained 10,130 images of liver tumor densities of 11,258 regions of interest (ROIs). The positive testing dataset contained 1,833 images of liver tumor densities with 1,874 ROIs, and negative testing data comprised 20,283 images without abnormal densities in liver parenchyma. The Mask R-CNN was used to generate a medical model, and areas under the curve, true positive rates, false positive rates, and Dice coefficients were evaluated. For abnormal liver CT density detection, in each image, we identified the mean area under the curve, true positive rate, and false positive rate, which were 0.9490, 91.99%, and 13.68%, respectively. For segmentation ability, the highest mean Dice coefficient obtained was 0.8041. This study trained a Mask R-CNN on various HCC images to construct a medical model that serves as an auxiliary tool for alerting radiologists to abnormal CT density in liver scans; this model can simultaneously detect liver lesions and perform automatic instance segmentation.

We sought to design a computer-assisted system measuring the anterior tibial translation in stress radiography, evaluate its diagnostic performance for an anterior cruciate ligament (ACL) tear, and assess factors affecting the diagnostic accuracy. Retrospective research for patients with both knee stress radiography and magnetic resonance imaging (MRI) at our institution was performed. A complete ACL rupture was confirmed on an MRI. The anterior tibial translations with four different methods were measured in 249 patients by the designed algorithm. The diagnostic accuracy of each method in patients with all successful measurements was evaluated. Univariate logistic regression analysis for factors affecting diagnostic accuracy of method four was performed. In the inclusive 249 patients, 177 patients (129 with completely torn ACLs) were available for analysis. Mean anterior tibial translations were significantly increased in the patients with a completely torn ACL by all four methods, with diagnostic accuracies ranging from 66.7% to 75.1%. The diagnostic accuracy of method four was negatively associated with the time interval between stress radiography and MRI as well as force-joint distance on stress view, and not significantly associated with age, gender, flexion angle, intercondylar distance, and force-joint angle. A computer-assisted system measuring the anterior tibial translation in stress radiography showed acceptable diagnostic performance of complete ACL injury. A shorter time interval between stress radiography and MRI as well as shorter force-joint distance were associated with higher diagnostic accuracy.

An aortic dissection has been reported a mortality of 50% within the first 48 hours and an increase of 1-2% per hour. Therefore, rapid diagnosis of intimal flap would be very important for the emergency treatment of patients. In order to accurately present the affected part of AD and reduce the time for doctors to diagnose, image segmentation is the most effective way of presentation. We used the U-Net model in this study and focus on AD (including ascending, arch, and descending part) in the detection process. Furthermore, we design the site and area regression (SAR) module. With this help of accurate prediction, we achieved slice-level sensitivity and specificity of 99.1 % and 93.2%, respectively.

This study aimed to evaluate the feasibility of automatic Stanford classification of classic aortic dissection (AD) using a 2-step hierarchical neural network. Between 2015 and 2019, 130 arterial phase series (57 type A, 43 type B, and 30 negative cases) in aortic CTA were collected for the training and validation. A 2-step hierarchical model was built including the first step detecting AD and the second step predicting the probability (0–1) of Stanford types. The model’s performance was evaluated with an off-line prospective test in 2020. The sensitivity and specificity for Stanford type A, type B, and no AD (Sens A, B, N and Spec A, B, N, respectively) and Cohen’s kappa were reported. Of 298 cases (22 with type A, 29 with type B, and 247 without AD) in the off-line prospective test, the Sens A, Sens B, and Sens N were 95.45% (95% confidence interval [CI], 77.16–99.88%), 79.31% (95% CI, 60.28–92.01%), and 93.52% (95% CI, 89.69–96.25%), respectively. The Spec A, Spec B, and Spec N were 98.55% (95% CI, 96.33–99.60%), 94.05% (95% CI, 90.52–96.56%), and 94.12% (95% CI, 83.76–98.77%), respectively. The classification rate achieved 92.28% (95% CI, 88.64–95.04%). The Cohen’s kappa was 0.766 (95% CI, 0.68–0.85; p < 0.001). Stanford classification of classic AD can be determined by a 2-step hierarchical neural network with high sensitivity and specificity of type A and high specificity in type B and no AD. • The Stanford classification for aortic dissection is widely adopted and divides it into Stanford type A and type B based on the ascending thoracic aorta dissected or not. • The 2-step hierarchical neural network for Stanford classification of classic aortic dissection achieved high sensitivity (95.45%) and specificity (98.55%) of type A and high specificity in type B and no aortic dissection (94.05% and 94.12%, respectively) in 298 test cases. • The 2-step hierarchical neural network demonstrated moderate agreement (Cohen’s kappa: 0.766, p < 0.001) with cardiovascular radiologists in detection and Stanford classification of classic aortic dissection in 298 test cases.

This paper reports an innovative approach to the classification of Stanford Type A and Type B aortic dissection using 3D CNN in conjunction with a novel Guided Attention (GA) mechanism. Recently, Computerized Tomography (CT) scan is increasingly applied for diagnoses of aortic dissection, and AI-assisted technology has been proven effective in increasing the productivity of radiologists. However, the general 3D CNN method even takes advantage of spatial continuity is not able to focus on the torn region of the aorta. In contrast, we propose an innovative approach termed the “External Guided Attention” (EGA), which is capable of focusing on both global and local features and guiding the model to learn key representative of the lesion. The scheme has been modified such that inputs of grayscale images combining with EGA channels can be trained and fine-tuned like regular RGB image inputs, so the pre-trained model on RGB video sequences can be utilized. Finally, we demonstrate that our new approach significantly outperforms other attention methods on categorizing Stanford Type-A and Type-B aortic dissection where the accuracy of 0.991 and an AUC of 0.994 are achieved in our untrimmed test dataset.

Prospective electrocardiography (ECG)-gated cardiac computed tomography angiography (CTA) is widely used for pediatric patients with congenital heart disease (CHD) due to the lower radiation dose compared with the ECG-gated technique. However, functional parameters acquired using ECG-gated cardiac CT to predict pulmonary hypertension (PH) in children with CHD have not yet been reported. This study aimed to investigate the potential of diastolic and systolic right ventricular diameters (RVD) on prospective ECG-gated cardiac CTA to predict PH in children with CHD. A total of 44 children with CHD were divided into two groups: CHD with PH (n = 22) and CHD without PH (n = 22). The association between ECG-gated CTA parameters and PH was evaluated by logistic regression. The receiver operating characteristic curve (ROC) was used to find the best cut-off point for the parameters measured by Youden's index. Patients with higher RVD-BSA [aOR (95% CI) diastolic: 2.76 (1.23-6.23); systolic: 6.15 (1.72-22.06)] had higher risk of PH after adjusting for age and patent ductus arteriosus. The area under the curve (AUC) of D-RVD-BSA was 0.907 and the AUC of S-RVD-BSA was 0.917. Logistic regression showed that patients with D-RVD-BSA over 6.86 or S-RVD-BSA over 5.87 had significantly higher risk of PH after adjustments (aOR = 23.52, 95% CI = 2.89-191.03; aOR = 31.14, 95% CI = 2.75-352.85). In conclusion, in children with CHD, measurements of diastolic or systolic BSA-modified RVDs on prospective ECG gated CTA are non-invasive markers of PH. BSA-modified D-RVD of 6.86 or BSA-modified S-RVD of 5.87 may be used to identify PH in children with CHD.

Background: changes may occur in tumor phenotype and receptor status during the progression of breast cancer. Discordance between primary and metastases has implications for further treatment and prognosis. Methods: 185 patients confirmed breast cancer metastasis were retrospectively analyzed during 1999-2019. All the pathological assessments of receptors and phenotypes of both primaries and metastases were recorded. Results: rates of receptor discordance were 18.65%, 30.57%, and 16.06% for ER, PR, and HER2, respectively and 31.62% for phenotype change. Patients with ER discordance experienced a worse OS and PMS, and those with ER loss had worse PMS compared with ER positive concordance. Patients with PR discordance experienced poorer OS and loss of PR positivity also had decreased OS and PMS when comparing with PR positive concordance. There was also significantly poorer PMS of hormon receptor (HR) discordance than HR positive concordance. In phenotype change, the luminal A type concordance group showed better PMS result. Conclusions: this study demonstrated that discordance in subtype and receptor status between primary and metastatic lesions ultimately affects the survival and has a potential impact on treatment options. (C) 2020 Asian Surgical Association and Taiwan Robotic Surgery Association. Publishing services by Elsevier B.V.

BACKGROUND:Neurenteric cyst (NEC) is a rare intradural spinal tumor, but a correct preoperative diagnosis remains challenging. A misdiagnosis of arachnoid cyst (AC) often leads to conflicting surgical management and significantly higher recurrence.CASE DESCRIPTION:We report the case of a 26-year-old woman who presented with progressive spastic quadriparesis with myelopathy below the C4 level, which was caused by a ventral intradural extramedullary cystic tumor at the C3-4 level. Magnetic resonance images showed the cystic content as identical to cerebrospinal fluid, which prompted the tentative diagnosis of spinal AC. Surgical fenestration was scheduled. However, intraoperative findings of a thick-walled cyst and severe adhesion to the neural structure without a history of trauma and inflammation were more compatible with the pathogenesis of an NEC. Because of the high recurrence rate after an incomplete resection of an NEC, we did a complete resection of the cyst with adhesive rootlets instead. Pathology analysis and immunohistochemical staining confirmed the diagnosis of an endodermal-derived NEC.CONCLUSIONS:NECs must be differentiated from ACs because they are different diseases and require different surgical management. In cases with clear cystic content, however, the diagnosis is likely to be AC, but a thick cystic wall and structural adhesions should suggest the differential diagnosis of NEC. Gross total removal of NECs should be attempted to reduce NEC recurrence.

Purpose: To examine outcomes and complications of percutaneous transhepatic biliary drainage (PTBD) performed via the right or left lobe and different entry levels (lobar, segmental, subsegmental, sub-subsegmental). Methods: The records of patients who underwent PTBD for obstructive jaundice from 2008 to 2015 were retrospectively reviewed. Logistic regression analysis was performed to determine factors associated with outcomes and complications based on entry side and entry level. PTBD success was defined as a total bilirubin decrease after catheter placement. Results: The data of 446 patients (mean age 68.4 years) were included. Multivariate logistic regression revealed a decrease of bilirubin level was associated with left lobe (vs. right) entry [adjusted odds ratio (aOR) = 2.657, 95% confidence interval (CI) 1.160, 6.087], external drainage (aOR = 2.908, 95% CI 1.226, 6.897), and liver volume undrained < 50% (aOR = 2.623, 95% CI 1.045, 6.581). PTBD success was increased with left lobe entry (aOR = 1.853, 95% CI 1.167, 2.940) and associated with entry level (subsegmental vs. lobar, aOR = 2.992, 95% CI 1.258, 7.114; sub-subsegmental vs. lobar, aOR = 3.711, 95% CI 1.383, 9.956). Complications were significantly decreased with left lobe entry (aOR = 0.450, 95% CI 0.263, 0.769) and associated with entry level (segmental vs. lobar, aOR = 0.359, 95% CI 0.148, 0.873; subsegmental vs. lobar, aOR = 0.248, 95% CI 0.10, 0.615; sub-subsegmental vs. lobar, aOR = 0.129, 95% CI 0.041, 0.411). Conclusions: The success and complications of PTBD vary with entry side and level.