Flouroscopic versus conventional epidural blood patch in obstetrics: A retrospective cohort study.

BACKGROUND: Epidural blood patch is a common effective treatment for postdural puncture headache after accidental dural puncture during labor and may be done in conventional or fluoroscopy-guided methods. The aim of this study was to compare intensity of headache at the time of discharge from the hospital and to compare blood volumes injected in conventional epidural blood patches versus fluoroscopic-guided blood patches and evaluate the side effects of both method of treatment. METHODS: Between the years 2010 and 2020, 84 patients who were diagnosed with postdural puncture headache received either a conventional epidural blood patch or a fluoroscopic-guided blood patch. Blood volumes were compared and evaluation of side effects was made based on data collected during and after the procedure. RESULTS: Eighty-four patients were included in this study. Fifty-two women in the conventional epidural blood patch group and 32 in the fluoroscopic-guided blood patch group. Women in the conventional epidural blood patch group received statistically significantly higher doses of blood than women in the fluoroscopic-guided blood patch group: conventional method 29 ml IQR [23-36] versus fluoroscopic method 16 ml, IQR [12-18], p < .001 with no difference in headache pain intensity at hospital release. There was no difference between groups in hospital length of stay, or persistent PDPH. There was also no difference chronic headache or backache between the two groups. CONCLUSIONS: Women who received fluoroscopic epidural blood patch required a much lower volume of blood injected while there was no difference between groups in headache pain intensity at discharge.

Deep learning for noninvasive liver fibrosis classification: A systematic review.

BACKGROUND AND AIMS: While biopsy is the gold standard for liver fibrosis staging, it poses significant risks. Noninvasive assessment of liver fibrosis is a growing field. Recently, deep learning (DL) technology has revolutionized medical image analysis. This technology has the potential to enhance noninvasive fibrosis assessment. We systematically examined the application of DL in noninvasive liver fibrosis imaging. METHODS: Embase, MEDLINE, Web of Science, and IEEE Xplore databases were used to identify studies that reported on the accuracy of DL for classification of liver fibrosis on noninvasive imaging. The search keywords were "liver or hepatic," "fibrosis or cirrhosis," and "neural or DL networks." Risk of bias and applicability were evaluated using the QUADAS-2 tool. RESULTS: Sixteen studies were retrieved. Imaging modalities included ultrasound (n = 10), computed tomography (n = 3), and magnetic resonance imaging (n = 3). The studies analyzed a total of 40 405 radiological images from 15 853 patients. All but two of the studies were retrospective. In most studies the "ground truth" reference was the METAVIR score for pathological staging (n = 9.56%). The majority of the studies reported an accuracy >85% when compared to histopathology. Fourteen studies (87.5%) had a high risk of bias and concerns regarding applicability. CONCLUSIONS: Deep learning has the potential to play an emerging role in liver fibrosis classification. Yet, it is still limited by a relatively small number of retrospective studies. Clinicians should facilitate the use of this technology by sharing databases and standardized reports. This may optimize the noninvasive evaluation of liver fibrosis on a large scale.

Deep learning for wireless capsule endoscopy: a systematic review and meta-analysis.

BACKGROUND AND AIMS: Deep learning is an innovative algorithm based on neural networks. Wireless capsule endoscopy (WCE) is considered the criterion standard for detecting small-bowel diseases. Manual examination of WCE is time-consuming and can benefit from automatic detection using artificial intelligence (AI). We aimed to perform a systematic review of the current literature pertaining to deep learning implementation in WCE. METHODS: We conducted a search in PubMed for all original publications on the subject of deep learning applications in WCE published between January 1, 2016 and December 15, 2019. Evaluation of the risk of bias was performed using tailored Quality Assessment of Diagnostic Accuracy Studies-2. Pooled sensitivity and specificity were calculated. Summary receiver operating characteristic curves were plotted. RESULTS: Of the 45 studies retrieved, 19 studies were included. All studies were retrospective. Deep learning applications for WCE included detection of ulcers, polyps, celiac disease, bleeding, and hookworm. Detection accuracy was above 90% for most studies and diseases. Pooled sensitivity and specificity for ulcer detection were .95 (95% confidence interval [CI], .89-.98) and .94 (95% CI, .90-.96), respectively. Pooled sensitivity and specificity for bleeding or bleeding source were .98 (95% CI, .96-.99) and .99 (95% CI, .97-.99), respectively. CONCLUSIONS: Deep learning has achieved excellent performance for the detection of a range of diseases in WCE. Notwithstanding, current research is based on retrospective studies with a high risk of bias. Thus, future prospective, multicenter studies are necessary for this technology to be implemented in the clinical use of WCE.

Deep learning algorithms for automated detection of Crohn's disease ulcers by video capsule endoscopy.

BACKGROUND AND AIMS: The aim of our study was to develop and evaluate a deep learning algorithm for the automated detection of small-bowel ulcers in Crohn's disease (CD) on capsule endoscopy (CE) images of individual patients. METHODS: We retrospectively collected CE images of known CD patients and control subjects. Each image was labeled by an expert gastroenterologist as either normal mucosa or containing mucosal ulcers. A convolutional neural network was trained to classify images into either normal mucosa or mucosal ulcers. First, we trained the network on 5-fold randomly split images (each fold with 80% training images and 20% images testing). We then conducted 10 experiments in which images from n - 1 patients were used to train a network and images from a different individual patient were used to test the network. Results of the networks were compared for randomly split images and for individual patients. Area under the curves (AUCs) and accuracies were computed for each individual network. RESULTS: Overall, our dataset included 17,640 CE images from 49 patients: 7391 images with mucosal ulcers and 10,249 images of normal mucosa. For randomly split images results were excellent, with AUCs of .99 and accuracies ranging from 95.4% to 96.7%. For individual patient-level experiments, the AUCs were also excellent (.94-.99). CONCLUSIONS: Deep learning technology provides accurate and fast automated detection of mucosal ulcers on CE images. Individual patient-level analysis provided high and consistent diagnostic accuracy with shortened reading time; in the future, deep learning algorithms may augment and facilitate CE reading.

Comparison of deep learning models for natural language processing-based classification of non-English head CT reports.

PURPOSE: Natural language processing (NLP) can be used for automatic flagging of radiology reports. We assessed deep learning models for classifying non-English head CT reports. METHODS: We retrospectively collected head CT reports (2011-2018). Reports were signed in Hebrew. Emergency department (ED) reports of adult patients from January to February for each year (2013-2018) were manually labeled. All other reports were used to pre-train an embedding layer. We explored two use cases: (1) general labeling use case, in which reports were labeled as normal vs. pathological; (2) specific labeling use case, in which reports were labeled as with and without intra-cranial hemorrhage. We tested long short-term memory (LSTM) and LSTM-attention (LSTM-ATN) networks for classifying reports. We also evaluated the improvement of adding Word2Vec word embedding. Deep learning models were compared with a bag-of-words (BOW) model. RESULTS: We retrieved 176,988 head CT reports for pre-training. We manually labeled 7784 reports as normal (46.3%) or pathological (53.7%), and 7.1% with intra-cranial hemorrhage. For the general labeling, LSTM-ATN-Word2Vec showed the best results (AUC = 0.967 ± 0.006, accuracy 90.8% ± 0.01). For the specific labeling, all methods showed similar accuracies between 95.0 and 95.9%. Both LSTM-ATN-Word2Vec and BOW had the highest AUC (0.970). CONCLUSION: For a general use case, word embedding using a large cohort of non-English head CT reports and ATN improves NLP performance. For a more specific task, BOW and deep learning showed similar results. Models should be explored and tailored to the NLP task.

Convolutional Neural Networks for Radiologic Images: A Radiologist's Guide.

Deep learning has rapidly advanced in various fields within the past few years and has recently gained particular attention in the radiology community. This article provides an introduction to deep learning technology and presents the stages that are entailed in the design process of deep learning radiology research. In addition, the article details the results of a survey of the application of deep learning-specifically, the application of convolutional neural networks-to radiologic imaging that was focused on the following five major system organs: chest, breast, brain, musculoskeletal system, and abdomen and pelvis. The survey of the studies is followed by a discussion about current challenges and future trends and their potential implications for radiology. This article may be used as a guide for radiologists planning research in the field of radiologic image analysis using convolutional neural networks.

Antibiotic treatment for invasive nonpregnancy-associated listeriosis and mortality: a retrospective cohort study.

Little evidence exists addressing the clinical value of adding gentamicin to ampicillin for invasive listeriosis. A multicenter retrospective observational study of nonpregnant adult patients with invasive listeriosis (primary bacteremia, central nervous system (CNS) disease, and others) in 11 hospitals in Israel between the years 2008 and 2014 was conducted. We evaluated the effect of penicillin-based monotherapy compared with early combination therapy with gentamicin, defined as treatment started within 48 h of culture results and continued for a minimum of 7 days. Patients who died within 48 h of the index culture were excluded. The primary outcome was 30-day all-cause mortality. A total of 190 patients with invasive listeriosis were included. Fifty-nine (30.6%) patients were treated with early combination therapy, 90 (46.6%) received monotherapy, and 44 (22.8%) received other treatments. Overall 30-day mortality was 20.5% (39/190). Factors associated with mortality included lower baseline functional status, congestive heart failure, and higher sequential organ failure assessment score. Source of infection, treatment type, and time from culture taken date to initiation of effective therapy were not associated with mortality. In multivariable analysis, monotherapy was not significantly associated with increased 30-day mortality compared with early combination therapy (OR 1.947, 95% CI 0.691-5.487). Results were similar in patients with CNS disease (OR 3.037, 95% CI 0.574-16.057) and primary bacteremia (OR 2.983, 95% CI 0.575-15.492). In our retrospective cohort, there was no statistically significant association between early combination therapy and 30-day mortality. A randomized controlled trial may be necessary to assess optimal treatment.

Gram-negative bloodstream infections in hemodialysis patients: A retrospective study
.

BACKGROUND: Bloodstream infections (BSIs) are an important cause of hospitalizations and mortality among hemodialysis (HD) patients. Epidemiology of these infections is changing, with increasing rates of Gram-negative pathogens, including resistant ones. Few studies have focused on the characteristics and outcomes of these infections. OBJECTIVE: We aimed to document the causative pathogens of BSIs in HD patients and their clinical outcomes during 2008 - 2015, and to compare risk factors, clinical features, appropriateness of therapy, and outcomes between patients with Gram-negative vs. Gram-positive BSIs. MATERIALS AND METHODS: A single-center retrospective cohort study. Charts of 120 HD patients hospitalized with first BSI were reviewed. RESULTS: A total of 120 patients were included, 61 episodes of Gram-negative (51.8%) and 59 episodes of Gram-positive bacteria (49.2%). Source of infection was significantly more likely to be urinary or abdominal among patients with Gram-negative infection. Otherwise, no statistically significant differences were documented between groups in terms of baseline characteristics, presentation of infection and outcomes. Most Gram-negative BSIs were caused by enterobacteriaceae, followed by Pseudomonas spp. Previous clinical or surveillance cultures added little to accurate prediction of the causative organism. CONCLUSION: In a cohort of HD patients with BSI, no significant differences were found between Gram-negative and Gram-positive BSIs, besides source of infection. A urinary or abdominal source strongly suggests a Gram-negative pathogen. Otherwise, patient's characteristics, clinical presentation, and previous cultures, all cannot accurately predict the causative pathogen of BSI, and broad-spectrum antibiotics should be used empirically.
.

Artificial Intelligence for Interstitial Lung Disease Analysis on Chest Computed Tomography: A Systematic Review.

RATIONALE AND OBJECTIVES: High-resolution computed tomography (HRCT) is paramount in the assessment of interstitial lung disease (ILD). Yet, HRCT interpretation of ILDs may be hampered by inter- and intra-observer variability. Recently, artificial intelligence (AI) has revolutionized medical image analysis. This technology has the potential to advance patient care in ILD. We aimed to systematically evaluate the application of AI for the analysis of ILD in HRCT. MATERIALS AND METHODS: We searched MEDLINE/PubMed databases for original publications of deep learning for ILD analysis on chest CT. The search included studies published up to March 1, 2021. The risk of bias evaluation included tailored Quality Assessment of Diagnostic Accuracy Studies and the modified Joanna Briggs Institute Critical Appraisal checklist. RESULTS: Data was extracted from 19 retrospective studies. Deep learning techniques included detection, segmentation, and classification of ILD on HRCT. Most studies focused on the classification of ILD into different morphological patterns. Accuracies of 78%-91% were achieved. Two studies demonstrated near-expert performance for the diagnosis of idiopathic pulmonary fibrosis (IPF). The Quality Assessment of Diagnostic Accuracy Studies tool identified a high risk of bias in 15/19 (78.9%) of the studies. CONCLUSION: AI has the potential to contribute to the radiologic diagnosis and classification of ILD. However, the accuracy performance is still not satisfactory, and research is limited by a small number of retrospective studies. Hence, the existing published data may not be sufficiently reliable. Only well-designed prospective controlled studies can accurately assess the value of existing AI tools for ILD evaluation.

Research trends analysis of chronic hepatitis C versus nonalcoholic fatty liver disease: A literature review text-mining analysis of publications.

Background: Hepatits C virus (HCV) rates have lowered due to direct-acting antiviral treatment. Nonalcoholic steatohepatitis (NASH)/nonalcoholic fatty liver disease (NAFLD) is rising with no available therapy. We employed text-mining to analyze trends in HCV and NAFLD research from the past two decades. Materials and Methods: We queried PubMed for all HCV and NASH/NAFLD entries published between 2000 and 2020. We compared the total number of publications on both etiologies. We performed subanalyses for different terms of interest and for geographic origin. Results: Overall, 75,934 HCV-related entries and 24,987 NASH/NAFLD-related entries were published during the study period. Up to 2015, there was a linear upward slope in the number of annual HCV publications (154.9 publications/year, p < 0.001). In 2015, the yearly number of HCV publications started showing a downward slope (-242.2 publications/year, p < 0.001). The number of NASH/NAFLD publications showed a continuous upward slope during the study period. The NASH/NAFLD field lacks publications on screening and treatment methods. Conclusion: Trends in publications varied between both etiologies. They reflect the success of antiviral treatment for HCV. The growing rates of NAFLD/NASH and the lack of a targeted cure explain the rise in related publications.

Deep learning for pulmonary embolism detection on computed tomography pulmonary angiogram: a systematic review and meta-analysis.

Computed tomographic pulmonary angiography (CTPA) is the gold standard for pulmonary embolism (PE) diagnosis. However, this diagnosis is susceptible to misdiagnosis. In this study, we aimed to perform a systematic review of current literature applying deep learning for the diagnosis of PE on CTPA. MEDLINE/PUBMED were searched for studies that reported on the accuracy of deep learning algorithms for PE on CTPA. The risk of bias was evaluated using the QUADAS-2 tool. Pooled sensitivity and specificity were calculated. Summary receiver operating characteristic curves were plotted. Seven studies met our inclusion criteria. A total of 36,847 CTPA studies were analyzed. All studies were retrospective. Five studies provided enough data to calculate summary estimates. The pooled sensitivity and specificity for PE detection were 0.88 (95% CI 0.803-0.927) and 0.86 (95% CI 0.756-0.924), respectively. Most studies had a high risk of bias. Our study suggests that deep learning models can detect PE on CTPA with satisfactory sensitivity and an acceptable number of false positive cases. Yet, these are only preliminary retrospective works, indicating the need for future research to determine the clinical impact of automated PE detection on patient care. Deep learning models are gradually being implemented in hospital systems, and it is important to understand the strengths and limitations of these algorithms.

Rhabdomyosarcoma disease spread evaluation on CT scans: Association with primary tumor size and Ki-67 proliferation marker.

PURPOSE: Ki-67 is a marker of cellular proliferation that is commonly used for the assessment of rhabdomyosarcoma. The aim of this study was to investigate the associations between Ki-67 expression and primary tumor diameter with CT evidence of lymph node and solid organ metastatic spread in rhabdomyosarcoma. MATERIALS AND METHODS: An institutional review board approval was granted for this study. A retrospective search for rhabdomyosarcoma patients was conducted. Pathology reports were examined for Ki-67 expression. Chest-abdomen CT was assessed for radiological evidence of lymph node and metastatic spread. The maximal primary tumor diameter (termed tumor size) was also measured in different modalities CT, MRI, PET-CT and US. Ki-67 levels and primary tumor maximal diameters were compared to CT evidence of lymph node and organ metastatic spread. RESULTS: Twenty-four patients with rhabdomyosarcoma were included. CT evidence of lymph node spread was associated with Ki-67 levels (AUC = 0.896, p = 0.006) and to a lesser extent with tumor size (AUC = 0.790, p = 0.030). However, organ metastatic spread was associated only with tumor size (AUC = 0.854, p = 0.006) and not with Ki-67 levels (AUC = 0.604, p = 0.469). A combination of tumor size ≥50 mm and Ki-67 levels ≥60% was significantly associated with CT evidence of lymph node spread (p = 0.004). CONCLUSION: In conclusion, this study demonstrates radiological-pathological correlation in RMS. Lymph node spread detected by radiological images is associated with Ki-67 values. Lymph node and metastatic spread are associated with primary tumor size.

PET-CT imaging features that differentiate between positive and negative EWSR1 translocation in Ewing sarcoma.

OBJECTIVES: Ewing sarcoma breakpoint region 1 (EWSR1) translocation-negative tumors represent a minor portion of small round cell tumors consistent with Ewing sarcoma morphology. The purpose of this study was to differentiate EWSR1 translocation-positive tumors from EWSR1 translocation-negative tumors using PET-computed tomography features. MATERIALS AND METHODS: In this retrospective study 27, Ewing sarcoma patients (December 2011 to November 2016) were divided into two groups, EWSR1 translocation-positive and EWSR1 translocation-negative based on cytogenetic analysis. Pretreatment standardized uptake value maximum (SUVmax) and Hounsfield Units (HU) were measured in the primary tumor in the axial slice with the largest tumor diameter.The associations between SUVmax, HU and the presence of EWSR1 translocation were evaluated. Receiver operating characteristic curve analysis was used to determine cut-off levels of SUVmax and HU suggestive of EWSR1-negative tumors. RESULTS: Twenty-one patients were classified as EWSR1-positive and six as EWSR1-negative. Eighteen had SUVmax and 21 had HU measurements. EWSR1-negative tumors had significantly higher SUVmax values (P = 0.003) and significantly lower HU values (P = 0.008). Receiver operating characteristic curve analysis showed that SUVmax had diagnostic ability to discriminate between EWSR1-negative and EWSR1-positive tumors (area under the curve = 0.964, P = 0.006). A SUVmax of at least 10 had a sensitivity of 100% and specificity of 85.7% for EWSR1-negative tumors. HU had lower diagnostic ability than SUVmax (area under the curve = 0.881, P = 0.012). A HU up to 57 had a sensitivity of 81.3% and specificity of 80.0% for EWSR1-negative tumors. CONCLUSION: Higher SUVmax and lower HU may differentiate between EWSR1-positive and EWSR1-negative tumors. This may reflect EWSR1-negative tumor aggressiveness.

Long-Term Effectiveness of Laparoscopic Conversion of Sleeve Gastrectomy to a Biliopancreatic Diversion with a Duodenal Switch or a Roux-en-Y Gastric Bypass due to Weight Loss Failure.

BACKGROUND: Gastroesophageal reflux disease and inadequate weight loss (IWL) are long-term complications of laparoscopic sleeve gastrectomy (LSG) and indications for a laparoscopic conversion to an alternative bariatric procedure. The aim of this study is to report the long-term outcomes of biliopancreatic diversion with a duodenal switch (DS) or a Roux-en-Y gastric bypass (RYGB) as conversion procedures for weight loss failure after LSG. METHODS: The data of all patients who underwent post-LSG conversion to either a RYGB or a DS at our institution between November 2006 and May 2016 was retrospectively analyzed. Included were all patients with > 1-year follow-up who were operated due to IWL or weight regain. Patients with the indication of reflux were excluded. RESULTS: Sixty-six patients underwent conversion from LSG to RYGB, DS, or one-anastomosis gastric bypass during the study period. There were 21 revisions to DS and 18 to RYGB that met the inclusion criteria. The respective weight and body mass index (BMI) before and after LSG were 125 and 110 kg and 46 and 40.5 kg/m2 in the RYGB group and 148 and 126 kg and 53.7 and 46 kg/m2 in the DS group. At the last follow-up (> 2 years), 15 RYGB patients had a reduction in BMI of 8.5-31.9 kg/m2 and 18 DS patients had a reduction in BMI of 12.8-31.9 kg/m2. The mean follow-up was 48.5 months (range 24-76). All comorbidities improved or underwent complete remission. CONCLUSION: Conversion from SG to RYGB or DS is an efficient and effective treatment for IWL and improvement of comorbidities. Further studies are warranted to evaluate long-term weight regain.