Current Medical Imaging - Volume 20, Issue 1, 2024

Neuroimaging has helped us learn about the stages of brain development from infancy to maturity. Neuroimaging helps physicians diagnose mental illnesses and find novel treatments for them. It can distinguish depression from neurodegenerative diseases or brain tumors, and it can reveal structural defects that cause psychosis. Psychosis has been linked to lesions in the frontal or temporal lobes of the brain, as well as the thalamus and hypothalamus, which can be detected using a brain scan for mental illnesses. Neuroimaging uses quantitative and computational methods to explore the central nervous system. It can detect brain injuries and psychological illnesses. Thus, a systematic review and meta-analysis of randomized controlled trials using neuroimaging to detect psychiatric disorders assessed their efficacy and benefits.

Appropriate articles were searched from PubMed, MEDLINE, and CENTRAL databases using the appropriate keywords as per the PRISMA guidelines. Randomized controlled trials and open-label studies were included as per the predefined PICOS criteria. Meta-analysis was performed using the RevMan software, and statistical parameters like odds ratio and risk difference were calculated.

Twelve randomized controlled clinical trials with a total of 655 psychiatric patients were included following the criteria from the year 2000 to 2022. We included studies that use different neuroimaging techniques for the detection of organic brain lesions that would help diagnose psychiatric disorders. The primary outcome was detecting brain abnormalities in diverse psychiatric illnesses with neuroimaging versus conventional methods. We found the odds ratio value of 2.29 (95% CI 1.49-3.51). The results were heterogeneous with a Tau² value of 0.38, chi² value of 35.48, df value of 11, I² value of 69%, the z value of 3.78, and p-value less than 0.05. The risk difference is 0.20 (95% CI 0.09 -0.31) with heterogeneity of Tau2 value of 0.03, chi2 value of 50, df value of 11, I² value of 78%, the z value of 3.49, and p-value less than 0.05.

The present meta-analysis strongly recommends the use of neuroimaging techniques for the detection of psychiatric disorders.

Extra-adrenal non-functional adrenocortical carcinoma (ACC)is an extremely rare tumor with only eight cases having been reported at different localizations.

A 60-year-old woman was presented to our hospital with abdominal pain. Magnetic resonance imaging revealed a solitary mass abutting the wall of the small bowel. She underwent resection of the mass, and the results of histopathology and immunohistochemistry were consistent with ACC.

We report the first occurrence of non-functional adrenocortical carcinoma in the wall of the small bowel in the literature. Magnetic resonance examination is sensitive enough to indicate the accurate location of the tumor and is of great help to clinical operation.

Schwannomas arise from nerve sheaths of cranial, peripheral, and spinal nerve or nerve roots. Most intracranial schwannomas arise from the cranial nerves, predominantly the vestibulocochlear nerve. In addition to cranial nerve schwannomas, intraparenchymal schwannomas of the brain and intramedullary schwannomas of the spinal cord are extremely rare.

In our case we describe the imaging findings of three diverse cases of schwannoma at different locations and unique presentations with acute neurological symptoms in the pediatric age group.

Schwannomas should be included in the differential diagnosis of intracranial or intraspinal intramedullary space-occupying lesions in pediatric patients.

The COVID-19 pandemic has resulted in a large number of deaths and has caused a significant increase in population morbidity. This viral infection has been associated with different neurological symptoms and complications that do not have a clear pathophysiological mechanism and exact implications for these patients.

A 40-year-old man with COVID-19 and co-infection with Klebsiella pneumoniae KPC presented extensive pulmonary involvement and required comprehensive management in the intensive care unit (ICU). During his hospitalization, he developed neurological symptoms with evidence of involvement of the corpus callosum, which was attributed to the cytotoxic lesion of the corpus callosum (CLOCC). After several months of interdisciplinary management in the ICU, there was a progressive improvement in his general condition, with discharge from the hospital without significant sequelae, with follow-up images showing complete involvement of the corpus callosum due to what was considered an atypical cytotoxic lesion of the corpus callosum.

Imaging features of CLOCCs are known to be temporary, but in the setting of COVID-19, it has not yet been determined if this is true and further studies are needed. Nonetheless, the one-year follow-up of our patient makes us believe that this atypical involvement of the corpus callosum described in severe SARS-CoV-2 infections is not transitory, even if there are no neurologic sequelae.

COVID-19 has become a worldwide epidemic disease and a new challenge for all mankind. The potential advantages of chest X-ray images on COVID-19 were discovered. We proposed a lightweight and effective Convolution Neural Network framework based on chest X-ray images for the diagnosis of COVID-19, named AMResNet.

COVID-19 has become a worldwide epidemic disease and a new challenge for all mankind. The potential advantages of chest X-ray images on COVID-19 were discovered.

A lightweight and effective Convolution Neural Network framework based on chest X-ray images for the diagnosis of COVID-19.

By introducing the channel attention mechanism and image spatial information attention mechanism, a better level can be achieved without increasing the number of model parameters.

In the collected data sets, we achieved an average accuracy rate of more than 92%, and the sensitivity and specificity of specific disease categories were also above 90%.

The convolution neural network framework can be used as a novel method for artificial intelligence to diagnose COVID-19 or other diseases based on medical images.

Whether deep learning-based CT reconstruction could improve lesion conspicuity on abdominal CT when the radiation dose is reduced is controversial.

To determine whether DLIR can provide better image quality and reduce radiation dose in contrast-enhanced abdominal CT compared with the second generation of adaptive statistical iterative reconstruction (ASiR-V).

This study aims to determine whether deep-learning image reconstruction (DLIR) can improve image quality.

In this retrospective study, a total of 102 patients were included, who underwent abdominal CT using a DLIR-equipped 256-row scanner and routine CT of the same protocol on the same vendor's 64-row scanner within four months. The CT data from the 256-row scanner were reconstructed into ASiR-V with three blending levels (AV30, AV60, and AV100), and DLIR images with three strength levels (DLIR-L, DLIR-M, and DLIR-H). The routine CT data were reconstructed into AV30, AV60, and AV100. The contrast-to-noise ratio (CNR) of the liver, overall image quality, subjective noise, lesion conspicuity, and plasticity in the portal venous phase (PVP) of ASiR-V from both scanners and DLIR were compared.

The mean effective radiation dose of PVP of the 256-row scanner was significantly lower than that of the routine CT (6.3±2.0 mSv vs. 2.4±0.6 mSv; p< 0.001). The mean CNR, image quality, subjective noise, and lesion conspicuity of ASiR-V images of the 256-row scanner were significantly lower than those of ASiR-V images at the same blending factor of routine CT, but significantly improved with DLIR algorithms. DLIR-H showed higher CNR, better image quality, and subjective noise than AV30 from routine CT, whereas plasticity was significantly better for AV30.

DLIR can be used for improving image quality and reducing radiation dose in abdominal CT, compared with ASIR-V.

The dimensions of the liver, spleen, and kidneys either change in primary diseases related to these organs or in secondary diseases that indirectly affect them, such as diseases of the cardiovascular system. Therefore, we aimed to investigate the normal dimensions of the liver, kidneys, and spleen and their correlations with body mass index in healthy Turkish adults.

A total of 1,918 adults older than 18 years of age underwent ultrasonographic (USG) examinations. Participants’ age, sex, height, weight, BMI, liver, spleen, and kidney dimensions, biochemistry and haemogram results were recorded. The relationships between organ measurements and these parameters were examined.

A total of 1,918 patients participated in the study. Of these, 987 (51.5%) were female and 931 (48.5%) were male. The mean age of the patients was 40.74± 15.95 years. The liver length (LL) for men was found to be greater than that for women. The effect of the sex factor on the LL value was statistically significant (p = 0.000). The difference between men and women in terms of liver depth (LD) was statistically significant (p=0.004). The difference between BMI groups in terms of splenic length (SL) was not statistically significant (p=0.583). The difference between BMI groups in terms of splenic thickness (ST) was statistically significant (p=0.016).

We obtained the mean normal standard values of the liver, spleen, and kidneys in a healthy Turkish adult population. Consequently, values exceeding those in our findings will guide clinicians in the diagnosis of organomegaly and will contribute to filling the gap in this regard.

The artifacts produced by calcification on coronary computed tomographic angiography (CCTA) have a great influence on the diagnosis of coronary stenosis. The purpose of this study is to investigate the value of corrected coronary opacification (CCO) difference in the diagnosis of stenosis in diffusely calcified coronary arteries (DCCAs).

A total of 84 patients were enrolled. The CCO difference across the diffuse calcification was measured through CCTA. Coronary arteries were grouped according to the extent of stenosis obtained by invasive coronary angiography (ICA). The Kruskal-Wallis H test was used to compare the CCO differences between different groups and a receiver operating characteristic (ROC) curve was used to evaluate the diagnostic efficacy of the CCO difference.

Among the 84 patients, 58 patients had one DCCA, 14 patients had 2 DCCAs, and 12 patients had 3 DCCAs. A total of 122 coronary arteries were examined, 16 showed no significant stenosis, 42 had <70% stenosis, and 64 had 70-99% stenosis. The median CCO differences among the 3 groups were 0.064, 0.117, and 0.176, respectively. There were significant differences between the group without stenosis and the group with 70-99% stenosis (H = -3.581, P = 0.001), and between the group with <70% stenosis and the group with 70-99% stenosis (H = -2.430, P = 0.045). The area under the ROC curve was 0.681 and the optimal cut-off point was 0.292. Taking the ICA results as the gold standard, the sensitivity and specificity for the diagnosis of ≥70% coronary stenosis with a cut-off point of 0.292 were 84.4% and 44.8%, respectively.

CCO difference could be useful in the diagnosis of ≥70% severe coronary stenosis in DCCA. Through this non-invasive examination, the CCO difference could be a reference for clinical treatment.

Although the essential components of pain pathways have been identified, a thorough comprehension of the interactions necessary for creating focused treatments is still lacking. Such include more standardised methods for measuring pain in clinical and preclinical studies and more representative study populations.

This review describes the essential neuroanatomy and neurophysiology of pain nociception and its relation with currently available neuroimaging methods focused on health professionals responsible for treating pain.

Conduct a PubMed search of pain pathways using pain-related search terms, selecting the most relevant and updated information.

Current reviews of pain highlight the importance of their study in different areas from the cellular level, pain types, neuronal plasticity, ascending, descending, and integration pathways to their clinical evaluation and neuroimaging. Advanced neuroimaging techniques such as fMRI, PET, and MEG are used to better understand the neural mechanisms underlying pain processing and identify potential targets for pain therapy.

The study of pain pathways and neuroimaging methods allows physicians to evaluate and facilitate decision-making related to the pathologies that cause chronic pain. Some identifiable issues include a better understanding of the relationship between pain and mental health, developing more effective interventions for chronic pain's psychological and emotional aspects, and better integrating data from different neuroimaging modalities for the clinical efficacy of new pain therapies.

Larger sample volume can be obtained in one needle pass using an aspiration-type semi-automatic cutting biopsy needle (STARCUT^® aspiration-type needle; TSK Laboratory, Tochigi, Japan) in comparison to the conventional semi-automatic cutting biopsy needle.

To evaluate and compare the safety and effectiveness of aspiration-type semi-automatic cutting biopsy needles and non-aspiration-type biopsy needles when performing computed tomography (CT)-guided core needle biopsies (CNBs).

A total of 106 patients underwent CT-guided CNB for chest lesions between June 2013 and March 2020 at our hospital. Non-aspiration-type cutting biopsy needles were used in 47 of these patients, while aspiration-type needles were used in the remaining 59 patients. All needles used were 18- or 20-gauge biopsy needles. Parameters, like forced expiratory volume in 1-second percent (FEV1.0%), the maximum size of the target lesion, puncture pathway distance in the lung, number of needle passes, procedure time, diagnostic accuracy, and incidence of complications, were measured. Comparisons were made between the needle-type groups.

No significant difference was observed in terms of diagnostic accuracy. However, the procedure time was shorter and a lesser number of needle passes were required with the aspiration-type cutting biopsy needle compared to the non-aspiration-type needle. Pneumothorax and pulmonary hemorrhage were the complications encountered, however, their incidence was not significantly different between the two types of needles.

The aspiration-type semi-automatic cutting biopsy needle had similar diagnostic accuracy as the non-aspiration-type biopsy needle, with added advantages of a lesser number of needle passes and shorter procedure time.

Chronic kidney disease (CKD) is a long-term risk to one’s health that can result in kidney failure. CKD is one of today's most serious diseases, and early detection can aid in proper treatment. Machine learning techniques have proven to be reliable in the early medical diagnosis.

The paper aims to perform CKD prediction using machine learning classification approaches. The dataset used for the present study for detecting CKD was obtained from the machine learning repository at the University of California, Irvine (UCI).

In this study, twelve machine learning-based classification algorithms with full features were used. Since the CKD dataset had a class imbalance issue, the Synthetic Minority Over-Sampling technique (SMOTE) was used to alleviate the problem of class imbalance and review the performance based on machine learning classification models using the K fold cross-validation technique. The proposed work compares the results of twelve classifiers with and without the SMOTE technique, and then the top three classifiers with the highest accuracy, Support Vector Machine, Random Forest, and Adaptive Boosting classification algorithms were selected to use the ensemble technique to improve performance.

The accuracy achieved using a stacking classifier as an ensemble technique with cross-validation is 99.5%.

The study provides an ensemble learning approach in which the top three best-performing classifiers in terms of cross-validation results are stacked in an ensemble model after balancing the dataset using SMOTE. This proposed technique could be applied to other diseases in the future, making disease detection less intrusive and cost-effective.

Brain tumour detection and classification require trained radiologists for efficient diagnosis. The proposed work aims to build a Computer Aided Diagnosis (CAD) tool to automate brain tumour detection using Machine Learning (ML) and Deep Learning (DL) techniques.

Magnetic Resonance Image (MRI) collected from the publicly available Kaggle dataset is used for brain tumour detection and classification. Deep features extracted from the global pooling layer of Pretrained Resnet18 network are classified using 3 different ML Classifiers, such as Support vector Machine (SVM), K-Nearest Neighbour (KNN), and Decision Tree (DT). The above classifiers are further hyperparameter optimised using Bayesian Algorithm (BA) to enhance the performance. Fusion of features extracted from shallow and deep layers of the pretrained Resnet18 network followed by BA-optimised ML classifiers is further used to enhance the detection and classification performance. The confusion matrix derived from the classifier model is used to evaluate the system's performance. Evaluation metrics, such as accuracy, sensitivity, specificity, precision, F1 score, Balance Classification Rate (BCR), Mathews Correlation Coefficient (MCC) and Kappa Coefficient (Kp), are calculated.

Maximum accuracy, sensitivity, specificity, precision, F1 score, BCR, MCC, and Kp of 99.11%, 98.99%, 99.22%, 99.09%, 99.09%, 99.10%, 98.21%, 98.21%, respectively, were obtained for detection using fusion of shallow and deep features of Resnet18 pretrained network classified by BA optimized SVM classifier. Feature fusion performs better for classification task with accuracy, sensitivity, specificity, precision, F1 score, BCR, MCC and Kp of 97.31%, 97.30%, 98.65%, 97.37%, 97.34%, 97.97%, 95.99%, 93.95%, respectively.

The proposed brain tumour detection and classification framework using deep feature extraction from Resnet 18 pretrained network in conjunction with feature fusion and optimised ML classifiers can improve the system performance. Henceforth, the proposed work can be used as an assistive tool to aid the radiologist in automated brain tumour analysis and treatment.

In recent years, various deep learning algorithms have exhibited remarkable performance in various data-rich applications, like health care, medical imaging, as well as in computer vision. COVID-19, which is a rapidly spreading virus, has affected people of all ages both socially and economically. Early detection of this virus is therefore important in order to prevent its further spread.

COVID-19 crisis has also galvanized researchers to adopt various machine learning as well as deep learning techniques in order to combat the pandemic. Lung images can be used in the diagnosis of COVID-19.

In this paper, we have analysed the COVID-19 chest CT image classification efficiency using multilayer perceptron with different imaging filters, like edge histogram filter, colour histogram equalization filter, color-layout filter, and Garbo filter in the WEKA environment.

The performance of CT image classification has also been compared comprehensively with the deep learning classifier Dl4jMlp. It was observed that the multilayer perceptron with edge histogram filter outperformed other classifiers compared in this paper with 89.6% of correctly classified instances.

Hepatic arterial infusion chemotherapy (HAIC) has been popular for treating unresectable hepatocellular carcinoma (HCC). However, there are few reports comparing the transradial approach (TRA) and transfemoral approach (TFA) in HAIC.

This study aimed to compare the duration of the hepatic artery catheterization, fluoroscopy time (FT), radiation exposure, safety, and quality of life associated with the procedure in patients undergoing HAIC via TRA and TFA.

This prospective, single-center, randomized, controlled study included 120 patients with unresectable HCC undergoing HAIC procedures. Patients were randomly assigned to group A (n = 60, TRA-HAIC) or group B (n = 60, TFA-HAIC). The hepatic artery catheterization time, FT, entrance surface dose (ESD), dose area product (DAP), procedure-related complications, and quality of life associated with the procedure were assessed between the two groups. Independent-sample t-test and analysis of variance (ANOVA) were used to assess differences. Statistical significance was set at P < 0.05.

HAIC procedures were successfully performed in both groups. The hepatic artery catheterization time (19.35 ± 5.84 vs. 18.93 ± 5.62 minutes, P = 0.837), FT (2.35 ± 2.23 vs. 2.25 ± 2.16 minutes, P = 0.901), ESD (259.32 ± 167.46 vs. 250.56 ± 170.58 mGy, P = 0.449), and DAP (125.37 ± 60.65 vs. 120.56 ± 64.33 Gy.cm³, P = 0.566) were comparable between the two groups. The incidence of artery occlusion (10.0% vs. 0%, P < 0.001) in the TRA group was significantly higher than that in the TFA group. TRA was associated with a statistically significant (P < 0.05) improvement in the quality of life.

TRA to HAIC was associated with greater improvement in the quality of life associated with the procedure compared with TFA. Both approaches to HAIC had similar efficiency, safety, radiation exposure, and procedure duration.

COVID-19 is a global pandemic. Currently, the predominant strain is SARS-CoV-2 Omicron subvariant BA.2 in many countries. Understanding its infection characteristics can facilitate clinical management.

This study aimed to characterize the clinical, laboratory, and high-resolution computed tomography (HRCT) findings in patients with mild or moderate infection from SARS-CoV-2 Omicron subvariant BA.2.

We performed a retrospective study on patients infected with SARS-CoV-2 Omicron subvariant BA.2 between April 4^th and April 17^th, 2022. The clinical characteristics, laboratory features, and HRCT images were reviewed.

A total of 805 patients were included (411 males and 394 females, median age 33 years old). The infection was mild, moderate, severe, and asymptomatic in 490 (60.9%), 37 (4.6%), 0 (0.0%), and 278 (34.5%) patients, respectively. Notably, 186 (23.1%), 96 (11.9%), 265 (32.9%), 11 (3.4%), 7 (0.9%), and 398 (49.4%) patients had fever, cough, throat discomfort, stuffy or runny nose, fatigue, and no complaint, respectively. Furthermore, 162 (20.1%), 332 (41.2%), and 289 (35.9%) patients had decreased white blood cell counts, reduced lymphocytes, and elevated C-reactive protein levels, respectively. HRCT revealed pneumonia in 53 (6.6%) patients. The majority of the lung involvements were ground-glass opacity (50, 94.3%) mostly in the subpleural area. The grade of lung injury was mainly mild (90.6%). Short-term follow-ups showed that most patients with pneumonia recovered.

Most patients with mild or moderate infection from SARS-CoV-2 Omicron subvariant BA.2 were adults, with fever and upper respiratory symptoms as the main clinical presentations. Lower respiratory infection was mild, with ground-glass opacity in the subpleural area as the main finding.

This study aimed to explore the imaging features and risk factors of PCLs complicating AIP, and investigate its prognosis through continuous imaging follow-up.

Patients who were diagnosed with AIP from January 2014 to December 2020 in our hospital were recruited. We analyzed the CT and MRI features of PCLs complicating AIP, and investigated its prognosis through imaging follow-up. We also compared subjects with and without PCLs using clinical, laboratory, and imaging data; the related risk factors associated with PCLs were investigated in a multivariate logistic regression analysis.

In this group, 16 patients had PCLs and 86 did not. A total of 43 PCLs larger than 5mm were found in 15 patients. Among these PCLs, 35 showed homogeneous signal (density); one, bleeding; three, linear separation; and four, small focal low signal on T2WI. Eight patients with 23 PCLs appeared for the follow-up after steroid treatment. Short-term follow-up showed that 11 PCLs disappeared, nine reduced, one unchanged and two enlarged. Of the 12 PCLs that did not disappear, 10 PCLs disappeared at long-term follow-up, except for two reduced PCLs were not re-examined. Logistic regression analysis showed that drinking history was an independent risk factor, age ≥ 65 years was an independent protective factor for PCLs complicating AIP.

The imaging features of PCLs complicating AIP are various, which can be single or multiple, most of them are homogeneous, and some lesions may be accompanied by hemorrhage, separation and necrosis. Age ≥ 65 years and avoiding drinking may help to reduce the occurrence of these lesions.

Xanthogranulomatous endometritis (XGE) is a rare inflammatory disease, which can easily misdiagnose as cancer in imaging diagnosis. Diagnosis of XGE relies on histopathological examination and immunohistochemistry.

In this study, a case of a 72-year-old female with XGE and elevated CA125 is presented, which was misdiagnosed as endometrial cancer in transvaginal ultrasonography and ovarian cystadenocarcinoma in CT. However, the features of XGE on the contrast-enhanced ultrasound (CEUS) were different from that of endometrial cancer. The patient finally underwent laparoscopic hysterectomy and bilateral adnexectomy.

The histopathological examination and immunohistochemistry suggested xanthogranulomatous endometritis (histiocytic endometritis). This case report manifests that CEUS may be a new noninvasive diagnostic method for XGE, which may reduce extensive tissue sampling and unnecessary hysterectomies for patients.

It has been reported in the literature that the complication rate of percutaneous compression plate (PCCP) is the lowest among the new internal fixators for the treatment of femoral neck fracture (FNS). However, no multicenter studies of PCCP for FNS have been reported. This study aimed to evaluate the med-term effectiveness of PCCP in a multicenter mainly through radiology.

265 patients with FNF treated with PCCP fixation in our five hospitals between January 2011 and December 2020 were retrospectively analyzed. 140 men and 125 women; aged 19–79 (mean 51.6) years. The follow-up time was 2-5 years (mean 3.1). Radiological evaluation of the therapeutic effect was the main outcome, and the function was the secondary outcome.

One case of screw cutting out, 3 cases of screw back out, 25 cases of neck shortening, 2 cases of nonunion, 8 cases of delayed healing, and 29 cases of avascular necrosis (AVN). Bivariate correlation showed that shortening healing was correlated with age, Singh index, and Garden alignment index, poor healing was correlated with garden alignment index, and AVN was correlated with Pauwels and Garden classifications and operation timing. Further pairwise comparison analysis showed that age of > 65 and Singh index IV were dangerous factors for neck shortening, and the operation timing > 3 days, Pauwels II and III, and Garden III and IV were dangerous factors for AVN. The excellent and good rate of function in 198 patients who were readmitted for internal fixator removal or other surgery was 90.9%.

PCCP for FNS has satisfactory med-term efficacy with a low complication rate. The main complication is AVN, which is prone to occur in patients with displaced Pauwels II or III FNF and operation timing > 3 days. Another main complication is shortening healing, which is prone to occur in patients with an age of > 65 and Singh index IV.

Mediastinal and hilar lesions may be benign or malignant. Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is increasingly used for the diagnosis of these lesions as it is both minimally invasive and safe.

To investigate the clinical efficacy of EBUS-TBNA in the diagnosis and differential diagnosis of mediastinal and hilar lesions.

A retrospective observational study was undertaken to investigate patients diagnosed with mediastinal and hilar lymphadenopathy based on imaging at our hospital from 2020 to 2021. After evaluation, EBUS TBNA was used and data including the puncture site, postoperative pathology, and complications were recorded.

Data from 137 patients were included in the study, of which 135 underwent successful EBUS TBNA. A total of 149 lymph node punctures were performed, of which 90 punctures identified malignant lesions. The most common malignancies were small-cell lung carcinoma, adenocarcinoma, and squamous cell carcinoma. Forty-one benign lesions were identified, resulting from sarcoidosis, tuberculosis, and reactive lymphadenitis, amongst others. Follow-up findings showed that 4 cases were malignant tumors, with 1 case of pulmonary tuberculosis and 1 case of sarcoidosis). Four specimens where lymph node puncture was insufficient were subsequently confirmed by other means. The sensitivity of EBUS TBNA for malignant lesions, tuberculosis and sarcoidosis in mediastinal and hilar lesions was 94.7%, 71.4%, and 93.3%, respectively. Similarly, the negative predictive values (NPV) were 88.9%, 98.5%, and 99.2%, and the accuracy was 96.3%, 98.5%, and 99.3%.

EBUS TBNA is an effective and feasible approach for the diagnosis of mediastinal and hilar lesions that is minimally invasive and safe.