Terms Of Diagnostic Accuracy Research Articles

Doppler ultrasound and contrast-enhanced ultrasound in detection of stent stenosis after iliac vein stenting

BackgroundTo compare the diagnostic accuracy of Doppler ultrasound (DUS) with contrast-enhanced ultrasound (CEUS) for detection of iliac vein stent stenosis using multidetector computed tomography venography (MDCTV) as the reference method.MethodsPatients with iliac vein obstructive disease treated with nitinol stents (Smart Control, Cordis, USA) between January 2016 and December 2017 were consecutively included in this study. DUS, CEUS, and MDCTV were carried out in all patients within one week of each other at 1 year post stenting to investigate the presence of stent compression and in-stent restenosis (ISR).ResultsThe study included 139 patients (87 females; mean age 58 ± 15 years). For detecting stent compression, the kappa coefficient between the ultrasound modality of gray-scale imaging and MDCTV was 0.901, indicating very good agreement between these two modalities. ISR was detected in 50, 61, and 65 patients by DUS, CEUS, and MDCTV, respectively. DUS and CEUS (kappa = 0.449) and DUS and MDCTV (kappa = 0.516) had moderate agreement for ISR diagnosis, while for which CEUS and MDCTV (kappa 0.884) had very good agreement. The sensitivity and specificity of DUS and CEUS for diagnosing ISR were 63.1% and 90.8%, 87.8% and 97.3%, respectively.ConclusionsCEUS is probably superior to DUS in terms of diagnostic accuracy for the follow-up of patients with iliac vein stent stenosis.

CAFD: Context-Aware Fault Diagnostic Scheme towards Sensor Faults Utilizing Machine Learning.

Sensors’ existence as a key component of Cyber-Physical Systems makes it susceptible to failures due to complex environments, low-quality production, and aging. When defective, sensors either stop communicating or convey incorrect information. These unsteady situations threaten the safety, economy, and reliability of a system. The objective of this study is to construct a lightweight machine learning-based fault detection and diagnostic system within the limited energy resources, memory, and computation of a Wireless Sensor Network (WSN). In this paper, a Context-Aware Fault Diagnostic (CAFD) scheme is proposed based on an ensemble learning algorithm called Extra-Trees. To evaluate the performance of the proposed scheme, a realistic WSN scenario composed of humidity and temperature sensor observations is replicated with extreme low-intensity faults. Six commonly occurring types of sensor fault are considered: drift, hard-over/bias, spike, erratic/precision degradation, stuck, and data-loss. The proposed CAFD scheme reveals the ability to accurately detect and diagnose low-intensity sensor faults in a timely manner. Moreover, the efficiency of the Extra-Trees algorithm in terms of diagnostic accuracy, F1-score, ROC-AUC, and training time is demonstrated by comparison with cutting-edge machine learning algorithms: a Support Vector Machine and a Neural Network.

Feature fusion of Raman chemical imaging and digital histopathology using machine learning for prostate cancer detection.

The diagnosis of prostate cancer is challenging due to the heterogeneity of its presentations, leading to the over diagnosis and treatment of non-clinically important disease. Accurate diagnosis can directly benefit a patient's quality of life and prognosis. Towards addressing this issue, we present a learning model for the automatic identification of prostate cancer. While many prostate cancer studies have adopted Raman spectroscopy approaches, none have utilised the combination of Raman Chemical Imaging (RCI) and other imaging modalities. This study uses multimodal images formed from stained Digital Histopathology (DP) and unstained RCI. The approach was developed and tested on a set of 178 clinical samples from 32 patients, containing a range of non-cancerous, Gleason grade 3 (G3) and grade 4 (G4) tissue microarray samples. For each histological sample, there is a pathologist labelled DP-RCI image pair. The hypothesis tested was whether multimodal image models can outperform single modality baseline models in terms of diagnostic accuracy. Binary non-cancer/cancer models and the more challenging G3/G4 differentiation were investigated. Regarding G3/G4 classification, the multimodal approach achieved a sensitivity of 73.8% and specificity of 88.1% while the baseline DP model showed a sensitivity and specificity of 54.1% and 84.7% respectively. The multimodal approach demonstrated a statistically significant 12.7% AUC advantage over the baseline with a value of 85.8% compared to 73.1%, also outperforming models based solely on RCI and mean and median Raman spectra. Feature fusion of DP and RCI does not improve the more trivial task of tumour identification but does deliver an observed advantage in G3/G4 discrimination. Building on these promising findings, future work could include the acquisition of larger datasets for enhanced model generalization.

Intermittent endoscopic ultrasound guided fine-needle aspiration for the diagnosis of solid pancreatic lesions. A pilot study.

endoscopic ultrasound guided fine-needle aspiration (EUS-FNA) is the method of choice for sampling pancreatic solid lesions. However, there is significant heterogeneity in terms of the technique used. Intermittent aspiration has not been evaluated in pancreatic solid lesions and could improve the diagnostic performance. a single-blind, non-inferiority pilot study was performed. Patients with solid pancreatic lesions and an indication for EUS-FNA were prospectively included. Patients were randomly assigned to intermittent (IS) or continuous (CS) suction techniques. Diagnostic performance, cellularity, blood contamination and the number of passes required to reach a diagnosis were evaluated. thirty-three patients were assigned to CS (16 patients) or IS (17 patients). Diagnostic performance was 87.5 % for CS and 94.1 % for IS (OR 2.29, 95 % CI: 0.19-27.99, p = 0.51). In the IS group, samples had a higher cellularity (OR 1.83, 95 % CI: 0.48-6.91, p = 0.37) and lower blood contamination (OR 0.38, 95 % CI: 0.09-1.54, p = 0.18). The number of passes required to reach a diagnosis was 2.12 for CS and 1.94 for IS (p = 0.64). Liquid cytology was obtained in 73.3 % of IS and 61.5 % of CS (OR 1.72, 95 % CI: 0.35-8.50). the IS technique was not inferior to CS in terms of diagnostic accuracy in the evaluation of pancreatic solid lesions, with a tendency to obtain higher cellularity, lower blood contamination and the frequent presence of cell block.

A Pancreatic Cystic Neoplasms Drama. Accuracy of Diagnostic Presumption and Surgical Indication on a Ten-year Experience

Purpose: Pancreatic cystic neoplasms (PCN) represent a challenging diagnosis despite imaging improvement and guidelines updates. Endoscopic Ultra Sound (EUS) is able to add valuable diagnostic information such as fine needle aspiration (FNA) but the accuracy of this procedure alone is extremely variable and not surely superior to classic MRI and CT-scan. The main aim of this study is to assess the importance of EUS and main cystic features in PCN preoperative work-up in terms of diagnostic accuracy, analyzing at the same those patients with any kind of PCN who could have performed, in retrospect, a close surveillance instead of surgical resection. Method: Patients with a preoperative or pathological diagnosis of PCN were retrospectively reviewed from a prospective database. Data included cases operated from 2010 to 2019 directly or under the supervision of an experienced pancreatic surgery team in three different institutions with a radiologic-oriented diagnostic approach or mixed (radiologic and endoscopic). Preoperative diagnosis was matched with final histology to confirm the diagnostic accuracy of the PCN. Patients with a benign histological who did not show absolute indication on preoperative evaluation, as defined by current guidelines, were categorized as “unrequested surgery”. Results: 587 patients were retrospectively analyzed and in 109/587 cases (18.6%) final histology did not confirm preoperative diagnosis. When performing EUS a significantly lower incidence of diagnostic fallacy was observed (p= 0.024) but risk of overtreatment was similar regardless instrumental diagnostic path (p= 0.356). MPD dilatation and cytologic sampling were the only variables independently related to a correct diagnosis (p <0.001 and p= 0.036). Based on final histology, pancreatic resection could be avoided in 145/461 patients (31.5%). The rate of “unrequested surgery” changed significantly over the years (p= 0.033) and was related with age (p <0.001) , an elevated Ca 19-9 (p= 0.012), cyst dimension (p= 0.023) and type of operation (p <0.001). Conclusions: EUS is an effective procedure in the diagnostic management of PCN, influencing in this series the diagnostic accuracy. Nevertheless, it does not seem to be able to reduce surgical overtreatment. Risk of “unrequested surgery” reduced over the years and it seems correlated with the type of pancreatic resection, being higher for distal pancreatectomies.Tabled 1RadiologicDiagnosis n/tot (%)Radiologic + Endoscopic diagnosis n/tot (%)pDiagnostic accuracyAll cohort148/194 (76.3)330/393 (84)0.024IPMN83/90 (92.2)265/274 (96.7)0.071SCA18/32 (56.3)18/30 (60)0.765MCN29/39 (74.4)25/39 (64.1)0.326Other18/33 (54.5)22/50 (44)0.377Unrequested SurgeryAll cohort46/133 (34.6)99/328 (30.2)0.356IPMN26/90 (28.9)73/274 (26.6)0.678SCA13/32 (40.6)5/30 (16.7)0.038Other7/11 (63.6)21/24 (87.5)0.101 Open table in a new tab Rate of diagnostic accuracy and “unrequested surgery”

A New Structured Domain Adversarial Neural Network for Transfer Fault Diagnosis of Rolling Bearings Under Different Working Conditions

This article presents a new deep transfer learning method, named structured domain adversarial neural network (SDANN), for bearing fault diagnosis with the data collected under different working conditions. The key idea of this method is integrating the strong adaptability of domain adversarial neural network (DANN) and structured relatedness information among multiple failure modes to improve the effect of transfer learning. First, for fine-grained alignment between the data collected from different working conditions, a new loss function with a discriminative regularizer is designed for DANN by using maximum correlation entropy constraint. Second, to improve the stability of DANN on an insufficient amount of data, a relatedness matrix is introduced, and a new regularizer with symmetry constraint on this matrix is designed to capture the intrinsic similarity structure among multiple fault types. Finally, a stochastic gradient descent optimization strategy is used to train the network and establish an end-to-end diagnostic model. Comparative experiments are conducted on two widely used bearing data sets. The results show that the proposed method has good diagnosis performance on insufficient monitoring data and outperforms several state-of-the-art transfer diagnosis methods and deep learning-based diagnosis methods in terms of diagnostic accuracy and numerical stability.

A Bayesian Compressive Sensing-Based Planar Array Diagnosis Approach From Near-Field Measurements

Array diagnosis is an important tool for detecting and correcting array antenna failures. In this letter, a high-precision planar array diagnosis method based on the Bayesian compressive sensing (BCS) theory is proposed. The model of a near-field signal with a spherical wavefront is used to acquire the measured data. Then, the difference between the beam pattern of the reference array and the array under test is obtained. The array diagnosis problem involves finding the difference between the weights of the reference and of the array under test with known differences between patterns. This problem is reformulated in a Bayesian compressive sensing framework and can be efficiently solved using a fast relevance vector machine. Numerical results confirm the superiority of the proposed method in terms of diagnostic accuracy and computational efficiency than those in previous studies.

NEUTROPHIL TO LYMPHOCYTE RATIO VS PLATELETS TO LYMPHOCYTE RATIO: BIOMARKERS TO PREDICT SEVERITY OF DISEASE AND THEIR COMPARISON IN PATIENTS OF COVID-19

Objective: To determine and compare Neutrophil to Lymphocyte Ratio (NLR) and Platelet to Lymphocyte Ratio (PLR) in predicting severity of disease in patients with COVID-19.&#x0D; Study design: Descriptive comparative study.&#x0D; Place and Duration of study: Department of Medicine, CMH Thal from April to July, 2020&#x0D; Patients &amp; Methods: 61 patients of COVID-19 confirmed through polymerase chain reaction were recruited and divided into severe and non severe disease. Complete blood counts were done. Neutrophil to Lymphocyte Ratio and Platelet to Lymphocyte Ratio were calculated and analyzed.&#x0D; Results: 14 patients (23%) of severe disease had mean age of 49.93±19.42 and 47 patients (77%) with non-severe disease had mean age of 33.32±9.16. The mean Neutrophil to Lymphocyte Ratio and Platelet to Lymphocyte ratio in patients who had severe disease was 7.20±4.20 and 204.25±148.42 (p=0.001 and p=0.026) respectively. The diagnostic performance of both Neutrophil to Lymphocyte Ratio and Platelet to Lymphocyte Ratio produced statistically significant area under the curve (AUC), (p &lt;0.001). The adjusted and unadjusted area under curve for Neutrophil to Lymphocyte Ratio was 0.92 (95% CI: 0.85–1.00) and 0.923 (95% CI: 0.839-1.000) and for Platelet to Lymphocyte Ratio it was 0.883 (95% CI: 0.781–0.985) and 0.825(95% CI: 0.707-0.943) respectively.&#x0D; Conclusion: Elevated Neutrophil to Lymphocyte Ratio and Platelet to Lymphocyte Ratio are independent biomarkers which predict severity of disease in COVID-19 patients with Neutrophil to Lymphocyte Ratio being better predictor in terms of diagnostic accuracy. &#x0D; Keywords: Corona virus disease 2019 (COVID‐19), Neutrophil to Lymphocyte ratio, Platelet to Lymphocyte ratio

Evaluation of Left Ventricular Structure and Function using 3D Echocardiography.

Three-dimensional (3D) quantification of the left ventricle (LV) provides significant added value in terms of diagnostic accuracy and precise risk stratification in various cardiac disorders. Recently, 3D echocardiography became available in routine cardiology practice; however, high-quality image acquisition and subsequent analysis have a steep learning curve. The present article aims to guide the reader through a detailed 3D protocol by presenting tips and tricks and also by highlighting the potential pitfalls to facilitate the widespread but technically sound use of this important technique concerning the LV. First and foremost, we show the acquisition of a high-quality 3D dataset with optimal spatial and temporal resolution. Then, we present the analytical steps toward a detailed quantification of the LV by using one of the most widely applied built-in software. We will quantify LV volumes, sphericity, mass and also systolic function by measuring ejection fraction and myocardial deformation (longitudinal and circumferential strain). We will discuss and provide clinical examples about the essential scenarios where the transition from a conventional echocardiographic approach to a 3D-based quantification is highly recommended.

The Added Value of Radiographs in Diagnosing Knee Osteoarthritis Is Similar for General Practitioners and Secondary Care Physicians; Data from the CHECK Early Osteoarthritis Cohort.

Objective: The purpose of this study was to evaluate the added value of radiographs for diagnosing knee osteoarthritis (KOA) by general practitioners (GPs) and secondary care physicians (SPs). Methods: Seventeen GPs and nineteen SPs were recruited to evaluate 1185 knees from the CHECK cohort (presenters with knee pain in primary care) for the presence of clinically relevant osteoarthritis (OA) during follow-up. Experts were required to make diagnoses independently, first based on clinical data only and then on clinical plus radiographic data, and to provide certainty scores (ranging from 1 to 100, where 1 was “certainly no OA” and 100 was “certainly OA”). Next, experts held consensus meetings to agree on the final diagnosis. With the final diagnosis as gold standard, diagnostic indicators were calculated (sensitivity, specificity, positive/negative predictive value, accuracy and positive/negative likelihood ratio) for all knees, as well as for clinically “certain” and “uncertain” knees, respectively. Student paired t-tests compared certainty scores. Results: Most diagnoses of GPs (86%) and SPs (82%) were “consistent” after assessment of radiographic data. Diagnostic indicators improved similarly for GPs and SPs after evaluating the radiographic data, but only improved relevantly in clinically “uncertain” knees. Radiographs added some certainty to “consistent” OA knees (GP 69 vs. 72, p < 0.001; SP 70 vs. 77, p < 0.001), but not to the consistent no OA knees (GP 21 vs. 22, p = 0.16; SP 20 vs. 21, p = 0.04). Conclusions: The added value of radiographs is similar for GP and SP, in terms of diagnostic accuracy and certainty. Radiographs appear to be redundant when clinicians are certain of their clinical diagnosis.

Slow-pull technique yields better quality smears: prospective comparison of slow-pull and standard suction techniques of endoscopic ultrasound-guided fine-needle aspiration

Background Diagnostic accuracy and quality of smears obtained by endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) are influenced by characteristics of suction and examined organ. Aims and methods Efficiency of EUS-FNA and quality of smears obtained by slow-pull (SP) and standard suction (SS) techniques was prospectively compared in the sampling of pancreatic (N = 56) and extrapancreatic (N = 145) tumors. Results SS technique resulted in a higher number of smear pairs both in pancreatic (1.74 vs. 3.19; p < 0.001) and extrapancreatic tumors (1.62 vs. 3.28; p < 0.001); however, it decreased the proportion of diagnostic smears (46.69% vs. 36.52%; p = 0.002 and 49.17% vs. 30.67%; p < 0.001) and increased the bloodiness (1.51 vs. 2.07; p < 0.001 and 1.48 vs. 2.05; p < 0.001). In pancreatic cancers, no difference was observed in terms of diagnostic accuracy (81.38% vs. 83.45%) and cellularity (1.44 vs. 1.27; p = 0.067); however, they were substantially higher in extrapancreatic tumors using SP technique (71.41% vs. 60.71% and 1.34 vs. 0.77; p < 0.001). Only SP technique resulted in a significant difference between examiners in terms of technical success rate and quality of smears without any decrease of diagnostic accuracy. Conclusions SP technique yields better quality smears independently from tumors characteristics; however, it shows significant examiner-dependency. SS technique reduces the diagnostic accuracy of sampling in extrapancreatic tumors.

Bayesian Analysis of Three Methods for Diagnosis of Cystic Echinococcosis in Sheep.

Diagnosis of cystic echinococcosis (CE) in sheep is essentially based on necropsy findings. Clinical symptoms can be easily overlooked, while the use of immunological tests is still not recommended for an intra vitam diagnosis. This study assessed the performances of three post-mortem laboratory methods in the diagnosis of ovine CE. In the absence of a single and accurate test as a gold standard, the results of multiple analytical tests can be combined to estimate diagnostic performance based on a Bayesian statistical approach. For this purpose, livers (n = 77), and lungs (n = 79) were sampled from adult sheep and examined using gross pathology, histopathology and molecular analyses. Data from the three diagnostic methods were analyzed using a Bayesian latent class analysis model to evaluate their diagnostic accuracy in terms of sensitivity (Se), specificity (Sp), positive predictive value (PPV) and negative predictive value (NPV). The gross pathology examination revealed excellent diagnostic capabilities in diagnosing ovine CE with an Se of 99.7 (96.7–99.8), Sp of 97.5 (90.3–99.8), PPV of 97.6 (90.5–100), and NPV of 99.7 (96.5–100). The experimental design used in this work could be implemented as a validation protocol in a quality assurance system.

L2 Mispronunciation Verification Based on Acoustic Phone Embedding and Siamese Networks

In computer-assisted pronunciation training (CAPT) system, feedback for non-native mispronunciation verification is important, for the reason that it is beneficial to the second language (L2) learners in respect of pronunciation improving. In pronunciation evaluation at the phone level, the pairwise distances between embeddings of native phones and non-native phones could be an ideal predictor of L2 learners’ proficiency. In CAPT, there are two key research issues to be addressed, one is mispronunciation verification and the other is pronunciation evaluation. Considering the positive role played by phone embedding and Siamese networks in related fields, we proposed to evaluate L2 learners’ pronunciation based on phone embedding and Siamese networks. Arbitrary-length speech segments corresponding to phones can be projected into acoustic phone embeddings space as fixed-dimensional vector representations. For system inputs, what is used is a pair of acoustic feature vectors of phone segments. The vectors are pair-wise labeled. And the Siamese networks will encode the feature vectors to phone embeddings as high-level representation. Thus, we can differentiate each type of phones through the similarities of their embeddings. As a result, in terms of diagnostic accuracy in mispronunciation verification tasks, Based on bi-directional Long Short Term Memory (LSTM) and contrastive loss, Siamese networks can be trained by a self-supervision using the pairwise labeled vectors without any mispronunciation-labeled L2 speech data in the training set. Results show that the proposed networks surpassed other methods and achieve accuracy as high as 90.69%.

COVID-19: An Opportunity to Reimagine Colorectal Cancer Diagnostic Testing—A New Paradigm Shift

ABSTRACT Colorectal cancer is the third leading cause of cancer related deaths and to date, the COVID-19 pandemic has led to major disruptions of the diagnosis and treatment pathways of this cancer, which may adversely affect survival outcomes for many years. Colorectal cancer pathways must be maintained at a near normal throughput with urgent attention to the backlog of patients to avoid a potential crisis of avoidable cancer deaths. A radical review and revamping of endoscopy services to improve efficiency during COVID-19 is urgently required. To date most endoscopy units have severely restricted their activity, except for time-critical diagnostic and therapeutic procedures. In the post- COVID-19 recovery phase, access to endoscopy services will increase but capacity will be reduced compared to before the pandemic, due to infection control measures. A pragmatic and safe strategy is urgently needed to clear the backlog of endoscopic procedures and for managing the new symptomatic referrals. We propose the use of non-invasive tests, such as FIT and mSEPT9 combined with “safety netting” to “rule out” colorectal cancer and to reduce the utilization of colonoscopy by up to 80% and limit the procedure to those patients at high risk of cancer.

Real Time Tissue Identification From Diathermy Smoke by Differential Mobility Spectrometry

Current methods for intraoperative surgical margin assessment are inadequate in terms of diagnostic accuracy, ease-of-use, and speed of analysis. Molecular analysis of tissues could potentially overcome these issues. A system based on differential ion mobility spectrometry (DMS) analysis of surgical smoke has been proposed as one potential method, but to date, it has been able to function in a relatively slow and heavily controlled manner that is inadequate for clinical use. In this study, we present an integrated sensor system that can measure a surgical smoke sample in seconds and relay the information of the tissue type to the user in near real time in simulated surgical use. The system was validated by operating porcine adipose tissue and muscle tissue. The differentiation of these tissues based on their surgical smoke profile with a cross-validated linear discriminant analysis model produced a classification accuracy of 93.1% (N = 1059). The measurements were also classified with a convolutional neural network model, resulting in a classification accuracy of 93.2%. These results indicate that the DMS-based smoke analysis system is capable of rapid tissue identification from surgical smoke produced in freehand surgery.

Validation of the revised 2018 AAST-OIS classification and the CT severity index for prediction of operative management and survival in patients with blunt spleen and liver injuries

ObjectivesNon-operative management (NOM) is increasingly utilised in blunt abdominal trauma. The 1994 American Association of Surgery of Trauma grading (1994-AAST) is applied for clinical decision-making in many institutions. Recently, classifications incorporating contrast extravasation such as the CT severity index (CTSI) and 2018 update of the liver and spleen AAST were proposed to predict outcome and guide treatment, but validation is pending.MethodsCT images of patients admitted 2000–2016 with blunt splenic and hepatic injury were systematically re-evaluated for 1994/2018-AAST and CTSI grading. Diagnostic accuracy, diagnostic odds ratio (DOR), and positive and negative predictive values were calculated for prediction of in-hospital mortality. Correlation with treatment strategy was assessed by Cramer V statistics.ResultsSeven hundred and three patients were analysed, 271 with splenic, 352 with hepatic and 80 with hepatosplenic injury. Primary NOM was applied in 83% of patients; mortality was 4.8%. Comparing prediction of mortality in mild and severe splenic injuries, the CTSI (3.1% vs. 10.3%; diagnostic accuracy = 75.4%; DOR = 3.66; p = 0.006) and 1994-AAST (3.3% vs. 10.5%; diagnostic accuracy = 77.9%; DOR = 3.45; p = 0.010) were more accurate compared with the 2018-AAST (3.4% vs. 8%; diagnostic accuracy = 68.2%; DOR = 2.50; p = 0.059). In hepatic injuries, the CTSI was superior to both AAST classifications in terms of diagnostic accuracy (88.7% vs. 77.1% and 77.3%, respectively). CTSI and 2018-AAST correlated better with the need for surgery in severe vs. mild hepatic (Cramer V = 0.464 and 0.498) and splenic injuries (Cramer V = 0.273 and 0.293) compared with 1994-AAST (Cramer V = 0.389 and 0.255; all p < 0.001).ConclusionsThe 2018-AAST and CTSI are superior to the 1994-AAST in correlation with operative treatment in splenic and hepatic trauma. The CTSI outperforms the 2018-AAST in mortality prediction.Key Points• Non-operative management of blunt abdominal trauma is increasingly applied and correct patient stratification is crucial.• CT-based scoring systems are used to assess injury severity and guide clinical decision-making, whereby the 1994 version of the American Association of Surgery of Trauma Organ Injury Scale (AAST-OIS) is currently most commonly utilised.• Including contrast media extravasation in CT-based grading improves management and outcome prediction. While the 2018-AAST classification and the CT-severity-index (CTSI) better correlate with need for surgery compared to the 1994-AAST, the CTSI is superior in outcome-prediction to the 2018-AAST.

Rational use of CT-scanner Considering its Diagnostic Accuracy for the Management of Acute Abdomen: A single-Center Study

Aim: We aimed to put forward the efficacy of abdominal CT performed for patients diagnosed with acute abdomen and to share our clinical findings in that regard.&#x0D; Methods: 216 patients who had been hospitalized in General Surgery Department due to acute abdomen had been analyzed retrospectively between October 2018 - January 2019. Abdominal computerized tomography (CT) results had been discussed in terms of diagnostic accuracy and clinical outcomes.&#x0D; Results: 171 (79.2%) patients (M/F:95/76 (55.6% / 44.4%)) had been evaluated by computerized tomography (CT) due to various causes of acute abdomen. Patients who had been scanned more than once (n=13 (7.6%)) during theirs’ hospitalization period had longer average length of stays compared to those whom had been scanned for only once (n=158 (92.4%)) (multiple scanned:17 days vs single scanned:3 days; p=0.000). Besides, single-scanned patients had showed higher clinical accuracy compared to those with multi-scans (80.4% vs 61.5%, p=0.126). 28 ileus (84.8%) patients had been scanned by non-enteral computerized tomography (CT), and 12 (36.4%) of them had been given false radiological reviews.&#x0D; Conclusion: Improper and redundant use of computerized tomography (CT) may cause prolongation of stays at the hospital, besides performing unnecessary scans do not improve diagnostic accuracy.

Squamous Cell Carcinoma: An Update on Diagnosis and Treatment.

Squamous cell carcinoma (SCC) accounts for most nonmelanoma skin cancer-related metastatic disease and deaths. Histopathology and correct surgical excision remain the gold standard for the diagnosis and treatment of SCC; however, new diagnostic imaging techniques such as dermoscopy and reflectance confocal microscopy have increased the diagnostic accuracy in terms of early recognition, better differential diagnosis, more precise selection of areas to biopsy, and noninvasive monitoring of treatments. The therapeutic intervention in patients with severe actinic damage and multiple in situ/low-risk SCC, and the development of innovative treatments such as epidermal growth factor receptor inhibitors and immune checkpoint inhibitors for locally advanced and metastatic SCC, are improving considerably the approach to the disease. This review summarizes the up-to-date knowledge in the field of detection, treatment, and monitoring of cutaneous SCC.

A Measurement Software for Professional Training in Early Detection of Melanoma

Software systems have been long introduced as support to the early detection of melanoma through the automatic analysis of suspicious skin lesions. Nevertheless, their behavior is not yet similar to the performance exhibited by expert dermatologists in terms of diagnostic accuracy. Instead, a software system should be adopted by non-experienced dermatologists in order to improve the measurement and detection results for skin atypical patterns and the accuracy of the corresponding second opinion. This paper describes an image-based measurement and classification system able to score pigmented skin lesions according to the Seven-Point Check-list diagnostic method. Focus is devoted to the measurement procedure of biological structures more closely related to the atypical character of the nevus. Moreover, the performances of the measurement system are evaluated by considering the support to dermatologists with different experiences during the clinical activity.

0594 Can a Deep Convolutional Neural Network Extract Diagnostic Information on Obstructive Sleep Apnea from Images?

Abstract Introduction Lateral cephalometric radiography is a simple way to provide craniofacial soft/hard tissue profiles specific for patients with obstructive sleep apnea (OSA) and may thus offer diagnostic information on the disease. We hypothesized that a machine learning technology, a deep convolutional neural network (DCNN), could make it possible to detect OSA based solely on lateral cephalometric radiographs without the need for either large amounts of subjective/laboratory data or skilled analyses. Methods In this diagnostic study, a DCNN was developed (n=1,258) and tested (n=131) using data from 1,389 lateral cephalometric radiographs obtained from individuals diagnosed with severe OSA (n=867; apnea hypopnea index &amp;gt;30/hour) or non-OSA (n=522; apnea hypopnea index &amp;lt; 5) at a single center for sleep disorders from March, 2006 to February, 2017. Three kinds of data sets were prepared by changing the area of interest using a single image; original image without any modification (Full Image), image containing a facial profile, upper airway, craniofacial soft/hard tissues, and image containing part of the occipital region (upper left corner of the image; Head Only). A radiologist and an orthodontist also performed a manual cephalometric analysis of the Full Image for comparison. Observers were blinded to the patient groupings. Data analysis was performed from April, 2018 to August, 2019. When the predictive score obtained from the DCNN analysis exceeded the threshold (0.50), the patient was judged to have OSA. The primary outcome was diagnostic accuracy in terms of area under the receiver-operating characteristic curve. Results The sensitivity/specificity was 0.87/0.82 for Full Image, 0.88/0.75 for Main Region, 0.71/0.63 for Head Only, and 0.54/0.80 for the manual analysis. The area under the curve was the highest for Main Region (0.92): 0.89 for Full Image, 0.70 for Head Only, and 0.75 for the manual analysis. Conclusion A DCNN identified individuals with OSA with high accuracy. This is a useful approach that does not require any laborious analyses in a primary care setting or in remote areas where an initial specialized OSA diagnosis is not feasible. Support This study was supported in part by the Japan Society for the Promotion of Science (grant numbers 17K11793, 19K10236).