A Deep Learning Model for Classifying Histological Types of Colorectal Polyps.

In this study a deep learning architecture based on a convolutional neural network has been evaluated for the classification of white light images of colorectal polyps acquired during the process of a colonoscopy, to estimate the accuracy of the optical recognition of histologic types of polyps. Convolutional neural networks (CNNs), a subclass of artificial neural networks that have gained dominance in several computer vision tasks, are gaining popularity in many medical fields, including endoscopy. The TensorFlow framework was used for implementing EfficientNetB7, which was trained with 924 images, drawn from 86 patients. 55% of the polyps were adenomas, 22% were hyperplastic, and 17% were lesions with sessile serrations. The validation loss, accuracy, and AUC ROC were 0.4845, 0.7778, and 0.8881 respectively.

Convolutional Neural Networks for Optical Discrimination Between Histological Types of Colorectal Polyps Based on White Light Endoscopic Images.

The objective of this study was to compare different convolutional neural networks (CNNs), as employed in a Python-produced deep learning process, used on white light images of colorectal polyps acquired during the process of a colonoscopy, in order to estimate the accuracy of the optical recognition of particular histologic types of polyps. The TensorFlow framework was used for Inception V3, ResNet50, DenseNet121, and NasNetLarge, which were trained with 924 images, drawn from 86 patients.

Using Machine Learning Techniques to Predict Hospital Admission at the Emergency Department.

Introduction: One of the most important tasks in the Emergency Department (ED) is to promptly identify the patients who will benefit from hospital admission. Machine Learning (ML) techniques show promise as diagnostic aids in healthcare. Aim of the study: Our objective was to find an algorithm using ML techniques to assist clinical decision-making in the emergency setting. Material and methods: We assessed the following features seeking to investigate their performance in predicting hospital admission: serum levels of Urea, Creatinine, Lactate Dehydrogenase, Creatine Kinase, C-Reactive Protein, Complete Blood Count with differential, Activated Partial Thromboplastin Time, DDi-mer, International Normalized Ratio, age, gender, triage disposition to ED unit and ambulance utilization. A total of 3,204 ED visits were analyzed. Results: The proposed algorithms generated models which demonstrated acceptable performance in predicting hospital admission of ED patients. The range of F-measure and ROC Area values of all eight evaluated algorithms were [0.679-0.708] and [0.734-0.774], respectively. The main advantages of this tool include easy access, availability, yes/no result, and low cost. The clinical implications of our approach might facilitate a shift from traditional clinical decision-making to a more sophisticated model. Conclusions: Developing robust prognostic models with the utilization of common biomarkers is a project that might shape the future of emergency medicine. Our findings warrant confirmation with implementation in pragmatic ED trials.

Using Machine Learning for Predicting the Hospitalization of Emergency Department Patients.

Artificial intelligence processes are increasingly being used in emergency medicine, notably for supporting clinical decisions and potentially improving healthcare services. This study investigated demographics, coagulation tests, and biochemical markers routinely used for patients seen in the Emergency Department (ED) concerning hospitalization. This retrospective observational study included 13,991 emergency department visits of patients who had undergone biomarker testing to a tertiary public hospital in Greece during 2020. After applying five well-known classifiers of the caret package for machine learning of the R programming language in the whole data set and to each ED unit separately, the best performance regarding AUC ROC was observed in the Pulmonology ED unit. Furthermore, among the five classification techniques evaluated, a random forest classifier outperformed other models.

Discovering Association Rules in Antimicrobial Resistance in Intensive Care Unit.

Multidrug resistant infections in intensive care units represent a worldwide problem with adverse health effects and negative economic implications. As artificial intelligence techniques are increasingly applied in diagnosing, treating, and preventing antimicrobial resistance, in this study, we explore the use of association rule mining in the antibiotic resistance profile of critically ill patients suffering from multidrug resistant infections.

Exploratory Clustering for Emergency Department Patients.

Emergency department (ED) overcrowding is an increasing global problem raising safety concerns for the patients. Elaborating an effective triage system that properly separates patients requiring hospital admission remains difficult. The objective of this study was to compare a clustering-related technique assignment of emergency department patients with the admission output using the k-means algorithm. Incorporating such a model into triage practice could theoretically shorten waiting times and reduce ED overcrowding.

Prediction of Hospitalization Using Machine Learning for Emergency Department Patients.

The objective of this study was to evaluate the predictive capability of five machine learning models regarding the admission or discharge of emergency department patients. A Random Forest classifier outperformed other models with respect to the area under the receiver operating characteristic curve (AUC ROC).

New considerations for colorectal cancer screening based on the demographic profile of colorectal cancer in a Greek population.

Colorectal cancer screening has long been recommended for middle age and older individuals. Recent evidence indicates increasing incidence and mortality among young adults. Therefore, the present study re-examined the current recommendations using an asymptomatic average-risk population screened by colonoscopy. A total of 716 participants of a wide age range were prospectively enrolled in an open-access endoscopic screening program based on self-referral. Comparisons between different age, gender and location groups, and receiver operating characteristic curves (ROC) curves for best age selection for detection of lesions were employed. Increased incidence of advanced lesions was observed in adults <50 years old. Although the polyp size was <1 cm in 85% of the cohort, a significant number of participants harbored advanced lesions. A disturbing incidence of lesions in women 30-49 years was located mainly in the left colon. One-third of the important pathology resides exclusively in the right colon. ROC curves demonstrated that with the current starting age of screening at 50 years, 92% of polyps and 95% of adenomas could be detected by colonoscopy, but a number of potential precancerous lesions will appear at an earlier age and therefore will be missed. The present study supported the notion that it is critical to reduce screening initiation below the currently accepted age of 50 years. Colonoscopy is a suitable method for addressing the increased prevalence of proximal lesions and the meticulous resection of smaller polyps.

Predicting Hospital Admission for Emergency Department Patients: A Machine Learning Approach.

The objective of this study was to establish a machine learning model and to evaluate its predictive capability of admission to the hospital. This observational retrospective study included 3204 emergency department visits to a public tertiary care hospital in Greece from 14 March to 4 May 2019. We investigated biochemical markers and coagulation tests that are routinely checked in patients visiting the Emergency Department (ED) in relation to the ED outcome (admission or discharge). Among the most popular classification techniques of the scikit-learn library through a 10-fold cross-validation approach, a GaussianNB model outperformed other models with respect to the area under the receiver operating characteristic curve.

Admission and Discharge Following Ambulance Transport to the Emergency Department.

Emergency ambulance use is deemed necessary for the transport of acutely ill patients to hospital emergency departments (ED). However, some patients are discharged as they present low acuity or chronic problems and should receive primary healthcare services, while the most severely ill are admitted. In the present study, we examined the descriptive epidemiology of ambulance transports for emergencies in the ED by utilizing the data of the information systems of a public tertiary general hospital in Greece. More than half of the patients transferred to the ED by an ambulance were finally admitted to the hospital (52.25%), whereas only one-third (33.74%) of those transferred by other means. A statistically significant association was detected between ambulance use and hospital admission. Age was also statistically significantly higher in the ambulance group. Higher mean values of creatinine, CRP, LDH, urea, white-blood-cell count, and neutrophils were detected in the ambulance group, in contrast to hemoglobin and lymphocyte count which were higher in the non-ambulance group.

Analyzing Acute Care Surgery Patient Flow in the Emergency Department During COVID-19 Pandemic.

During the COVID-19 pandemic, the number of visits in emergency departments (ED) worldwide decreased significantly based on several studies. This study aims to compare the patient flow in the emergency surgery department during the COVID-19 pandemic and a control period in the emergency department of a public tertiary care hospital in Greece. The overall patient flow reduction regarding the ED visits between the two examined periods was 49.07%. The emergency surgery department's corresponding visits were 235 and 552, respectively, which indicated an overall patient flow decrease of 57.43%. Chi-square analysis showed that age groups and ambulance use had statistically significant associations with the periods examined. An independent samples t-test was applied and deduced that the average patient's age was statistically significantly higher in the COVID-19 pandemic than in the non-pandemic period. By analyzing hospital information system data, useful conclusions can be drawn to prepare a surgical emergency unit better and optimize resource allocation in a healthcare facility in similar critical situations.

Using Machine Learning to Predict Antimicrobial Resistance of Acinetobacter Baumannii, Klebsiella Pneumoniae and Pseudomonas Aeruginosa Strains.

Hospital-acquired infections, particularly in ICU, are becoming more frequent in recent years, with the most serious of them being Gram-negative bacterial infections. Among them, Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa are considered the most resistant bacteria encountered in ICU and other wards. Given the fact that about 24 hours are usually required to perform common antibiotic resistance tests after the bacteria identification, the use of machine learning techniques could be an additional decision support tool in selecting empirical antibiotic treatment based on the sample type, bacteria, and patient's basic characteristics. In this article, five machine learning (ML) models were evaluated to predict antimicrobial resistance of Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa. We suggest implementing ML techniques to forecast antibiotic resistance using data from the clinical microbiology laboratory, available in the Laboratory Information System (LIS).

Endovascular management of secondary aortoduodenal fistula: the importance of gut restoration.

Secondary aortoenteric fistula (SAEF) is a rare, life-threatening complication of aortic reconstructive surgery. Graft excision associated with gut repair and extra-anatomic bypass or in situ aortic reconstruction is the best option. However, it is associated with high mortality rates, especially when undertaken in unstable patients with severe comorbidities. We present a case of SAEF successfully treated by endovascular repair and gut restoration. After laparotomy, a dense inflammatory "frozen" mass was found around the involved part of the duodenum and the aortic sac. Because of his comorbidities, the difficulty in dissection of the aortic sac and the risk of damage in adjacent organs, the initial plan for aortic reconstruction was abandoned. A surgical isolation of the third portion of the duodenum and a duodenum-jejunal anastomosis were performed. An Endurant endograft 32-16-16 mm (Medtronic, Inc., Minneapolis, MN) was implanted to achieve aortic continuity. Twenty months postdischarge, the patient remains in good general condition, afebrile, and with normal laboratory tests and inflammation markers. In high-risk patients, endovascular intervention combined with gut repair without further aortic reconstruction can be a permanent solution requiring, however, close surveillance.

Duplex Ultrasound versus Computed Tomography for the Postoperative Follow-Up of Endovascular Abdominal Aortic Aneurysm Repair. Where Do We Stand Now?

In the last decade, endovascular aneurysm repair (EVAR) has rapidly developed to be the preferred method for infrarenal abdominal aortic aneurysm repair in patients with suitable anatomy. EVAR offers the advantage of lower perioperative mortality and morbidity but carries the cost of device-related complications such as endoleak, graft migration, graft thrombosis, and structural graft failure. These complications mandate a lifelong surveillance of EVAR patients and their endografts. The purpose of this study is to review and evaluate the safety of color-duplex ultrasound (CDU) as compared with computed tomography (CT), based on the current literature, for post-EVAR surveillance. The post-EVAR follow-up modalities, CDU versus CT, are evaluated questioning three parameters: (1) accuracy of aneurysm size, (2) detection and classification of endoleaks, and (3) detection of stent-graft deformation. Studies comparing CDU with CT scan for investigation of post-EVAR complications have produced mixed results. Further and long-term research is needed to evaluate the efficacy of CDU versus CT, before CDU can be recommended as the primary imaging modality for EVAR surveillance, in place of CT for stable aneurysms.

Evaluation of apoptosis in varicose vein disease complicated by superficial vein thrombosis.

BACKGROUND: The factors contributing to superficial vein thrombosis (SVT) in patients with varicose vein disease are unclear. Differences in vein wall apoptotic activity could be associated with the pathogenesis of SVT. The aim of the study is to address the role of the programmed cell death in the vein wall by comparing varicose veins with history of SVT to uncomplicated varicose veins. PATIENTS AND METHODS: Vein segments from the proximal part of the great saphenous vein (GSV), the distal part of the vein and from a varicose tributary, from 16 patients with varicose vein disease and one episode of SVT, were evaluated for the immunohistochemical expression of pro-apoptotic (Bax, p53, Caspase 3, BCL-6, BCL-xs), anti-apoptotic (BCL-xl and BCL-2) and proliferation (Ki-67) markers. The results of this study were compared to the results from the evaluation of 19 patients suffering from uncomplicated varicose vein disease and 10 healthy GSVs as controls. RESULTS: Overall, there was increased apoptosis in the distal part of GSV compared to the proximal part documented by increased expression of Bax (p < 0.01), Caspase 3 (p = 0.01), BCL-xs (p < 0.01). The comparisons of the markers' expression between patients with varicose veins and patients with a history of SVT showed significant differences among the three different anatomic locations. In the proximal GSV, only BCL-xs was higher in patients with SVT (p = 0.029). In the tributaries, Bax, BCL-xl and Ki-67 were higher in patients with SVT (p < 0.01). In the distal GSV, increased Bax, BCL-xs, BCL-xl and Ki-67 staining was observed in the thrombosis group compared to uncomplicated veins (p < 0.01). CONCLUSIONS: The vein wall in SVT shows increased pro-apoptotic activity compared to uncomplicated disease and normal veins. Whether increased vein wall cell apoptosis is a causative factor for SVT in varicose veins disease or a repairing mechanism of the thrombosis itself needs further research.

Management of complicated and uncomplicated acute type B dissection. A systematic review and meta-analysis.

BACKGROUND: The management of acute type B dissection represents a clinical challenge. We undertook a systematic review of the available literature regarding medical, surgical and endovascular treatments of acute type B aortic dissection and combined the eligible studies into a meta-analysis. METHODS: An extensive electronic health database search was performed on all articles published from January 2006 up to November 2013 describing the management of acute type B aortic dissection. Studies including less than 15 patients were excluded. RESULTS: ACUTE COMPLICATED TYPE B DISSECTION: overall, 2,531 patients were treated with endovascular repair (TEVAR) and the pooled rate for 30-day/in-hospital mortality was 7.3%. The pooled estimates for cerebrovascular events, spinal cord ischemia (SCI) and total neurologic events were 3.9%, 3.1% and 7.3%, respectively. A total of 1,276 patients underwent open surgical repair and the pooled rate for 30-day/in-hospital mortality was 19.0%. The pooled rate for cerebrovascular events was 6.8%, for SCI 3.3% and for total neurologic complications 9.8%. Acute uncomplicated type B dissection: outcome of 2,347 patients who underwent conservative medical management were analyzed. The pooled 30-day/in-hospital mortality rate was 2.4%. The pooled rate for cerebrovascular events was 1%, for SCI 0.8% and for overall neurologic complications 2%. CONCLUSIONS: Endovascular repair provides a superior 30-day/in-hospital survival for acute complicated type B aortic dissection compared to surgical aortic reconstruction. However, open repair still has a significant role as endovascular repair is not applicable in all patients and there remains concerns regarding the durability of this technique. TEVAR seems to have a more favorable outcome regarding aortic remodeling and the aortic-specific survival rate when compared with medical therapy alone. Randomized controlled trials focusing on the prognostic factors of early and late complications in uncomplicated type B dissections are needed.

Urinary cystatin C as an early biomarker of acute kidney injury after open and endovascular abdominal aortic aneurysm repair.

BACKGROUND: Acute kidney injury (AKI) after open repair (OR) and endovascular repair (EVAR) of abdominal aortic aneurysm (AAA) is associated with increased mortality and hospital costs. Early detection of AKI is critical to prevent its progression. Recent findings demonstrate that elevated levels of urinary cystatin C (uCysC) may reflect tubular dysfunction. We prospectively evaluated whether uCysC can detect renal dysfunction earlier than serum creatinine (sCr). METHODS: In a prospective study, 126 consecutive patients (mean age ± SD, 69.1 ± 8.66 years) with AAA (EVAR = 87, OR = 39) were enrolled. sCr and uCysC were measured preoperatively (baseline) and at 6, 24, and 48 hr postoperatively. A final measurement was made on day 5. AKI was defined according to Acute Kidney Injury Network criteria. RESULTS: The incidence of AKI was significantly higher (χ(2) test, P < 0.05) in the OR group (n = 13, 33%) than in the EVAR group (n = 15, 17%). The baseline median (interquartile range) value of uCysC was significantly higher (t-test, P < 0.05) in patients of both groups (OR-EVAR) who developed AKI from those who did not (OR/AKI group: 0.06 [0.02-0.12] mg/L, EVAR/AKI group: 0.08 [0.05-0.11] mg/L versus no-AKI subjects: 0.04 [0.02-0.07] mg/L). Subsequent analysis showed that at 6 hr postoperatively, the patients who developed AKI increased their uCysC levels significantly from baseline (OR/AKI group: 0.58 [0.42-0.70] mg/L, EVAR/AKI group: 0.59 [0.30-1.07] mg/L). The median value of uCysC in AKI patients increased at 24 hr (OR/AKI group: 1.37 [0.78-3.40] mg/L, EVAR/AKI group: 2.11 [0.70-2.42] mg/L) and peaked at 48 hr (OR/AKI group: 6.16 [1.74-10.73] mg/L, EVAR/AKI group: 2.57 [1.21-7.40] mg/L), while no increase was observed among those who did not develop AKI at the same time points (0.06 [0.04-0.14] vs. 0.08 [0.04-0.19] mg/L). The diagnostic accuracy of uCysC at 6 hr post-surgery was excellent (area under the curve - receiver-operating characteristic [AUC-ROC] = 0.968), significantly higher than sCr (AUC-ROC = 0.844) and a cutoff value set at 0.30 mg/L can diagnose AKI with a sensitivity of 85.71% and a specificity of 98.97%. CONCLUSIONS: uCysC is superior to sCr in the early diagnosis of AKI following open and endovascular AAA repair.

The chimney-graft technique for preserving supra-aortic branches: a review.

Evolution in the endovascular era has influenced the management of aortic arch pathologies. "Chimney" or "snorkel" graft technique has been used as an alternative in high risk patients unfit for open repair. We reviewed the published literature on the chimney graft technique for preservation of the supra-aortic branches in order to provide an extensive insight of its feasibility and efficacy and investigate its outcomes. 18 reports were identified, with a total of 124 patients and 136 chimney. Primary technical success was achieved in 123/124 patients (99.2%). The perioperative mortality rate was 4.8% and the stroke rate was 4%, while events of spinal cord ischemia were rare. The overall endoleak rate was 18.5%; 13 patients (10.5%) developed a type I endoleak and 10 (8%) patients a type II endoleak. During a median follow-up period of 11.4 months (range, 0.87-20.1 months) all implanted chimney grafts remained patent. From this, we conclude that endovascular aortic arch repair with chimney grafts is associated with a lower mortality rate compared to totally open or hybrid reconstruction. However, the stroke rate remains noteworthy, and requires longterm data to elucidate.

Endovascular Treatment versus Open Repair for Abdominal Aortic Aneurysms: The Influence of Fitness in Decision Making.

Two methods of repair are currently available for an abdominal aortic aneurysm (AAA), open aneurysm repair and endovascular aneurysm repair (EVAR). The decision making depends on the balance of risks and benefits. The treating physician must take into account the patient's life expectancy, the patient's fitness, the anatomic suitability that makes endovascular repair possible, and finally the patient's preference. The patient's fitness is an important variable predicting the outcome of AAA surgical reconstruction. The hypothesis is that the impact of risk factors upon perioperative mortality might differ between patients undergoing open repair and endovascular repair. The purpose of this review article is to investigate whether fitter patients with a large AAA benefit more from having endovascular rather than open repair. According to the available data, there is emerging evidence that patients at high medical risk for open repair may benefit from EVAR while in low risk patients with suitable anatomy for EVAR, both techniques have similar effects. There is rising evidence that a patient with ruptured AAA would benefit more from an endovascular procedure if eligible, and thus fitness in such emergencies is not the first priority but anatomical suitability for EVAR.