-
Diagnostics (Basel, Switzerland) Aug 2023Deep learning (DL), often called artificial intelligence (AI), has been increasingly used in Pathology thanks to the use of scanners to digitize slides which allow us to... (Review)
Review
Deep learning (DL), often called artificial intelligence (AI), has been increasingly used in Pathology thanks to the use of scanners to digitize slides which allow us to visualize them on monitors and process them with AI algorithms. Many articles have focused on DL applied to prostate cancer (PCa). This systematic review explains the DL applications and their performances for PCa in digital pathology. Article research was performed using PubMed and Embase to collect relevant articles. A Risk of Bias (RoB) was assessed with an adaptation of the QUADAS-2 tool. Out of the 77 included studies, eight focused on pre-processing tasks such as quality assessment or staining normalization. Most articles ( = 53) focused on diagnosis tasks like cancer detection or Gleason grading. Fifteen articles focused on prediction tasks, such as recurrence prediction or genomic correlations. Best performances were reached for cancer detection with an Area Under the Curve (AUC) up to 0.99 with algorithms already available for routine diagnosis. A few biases outlined by the RoB analysis are often found in these articles, such as the lack of external validation. This review was registered on PROSPERO under CRD42023418661.
PubMed: 37627935
DOI: 10.3390/diagnostics13162676 -
CVIR Endovascular Aug 2023Occult gastrointestinal bleeding (GIB) is a challenge for physicians to diagnose and treat. A systematic literature search of the PubMed and Embase databases was... (Review)
Review
Occult gastrointestinal bleeding (GIB) is a challenge for physicians to diagnose and treat. A systematic literature search of the PubMed and Embase databases was conducted up to January 1, 2023. Eligible studies included primary research studies with patients undergoing provocative mesenteric angiography (PMA) for diagnosis or localization of occult GIB. Twenty-seven articles (230 patients) were included in the review. Most patients (64.8%) presented with lower GIB. The average positivity rate for provocative angiography was 48.7% (58% with heparin and 46.7% in thrombolytics). Embolization was performed in 46.4% of patients, and surgical management was performed in 37.5%. Complications were rare. PMA can be an important diagnostic and treatment tool but studies with high-level evidence and standardized protocols are needed to establish its safety and optimal use.
PubMed: 37589781
DOI: 10.1186/s42155-023-00386-7 -
European Journal of Nuclear Medicine... Dec 2023Transthyretin (ATTR) amyloidosis is a progressive protein misfolding disease with frequent cardiac involvement. This review aims to determine the value of PET in... (Review)
Review
PURPOSE
Transthyretin (ATTR) amyloidosis is a progressive protein misfolding disease with frequent cardiac involvement. This review aims to determine the value of PET in diagnosis, assessment of disease progression or treatment response and its relation to clinical outcome in follow-up of ATTR amyloid cardiomyopathy (ATTR-CM) patients.
METHODS
Medline, Cochrane Library, Embase and Web of Science databases were searched, from the earliest date available until December 2022, for studies investigating the use of PET in ATTR-CM patients. Studies containing original data were included, except for case reports. Risk of bias was assessed by QUADAS-2.
RESULTS
Twenty-one studies were included in this systematic review, investigating five different tracers: carbon-11 Pittsburgh compound B ([C]PIB), fluorine-18 Florbetaben ([F]FBB), fluorine-18 Florbetapir ([F]FBP), fluorine-18 Flutemetamol ([F]FMM) and fluorine-18 Sodium Fluoride (Na[F]F). In total 211 ATTR amyloidosis patients were included. A majority of studies concluded that [C]PIB, [F]FBP and Na[F]F can distinguish ATTR amyloidosis patients from controls, and that [C]PIB and Na[F]F, but not [F]FBP, can distinguish ATTR-CM patients from patients with cardiac light chain amyloidosis. Evidence on the performance of [F]FBB and [F]FMM was contradictory. No studies on the use of PET in follow-up were found.
CONCLUSION
[C]PIB, Na[F]F and [F]FBP can be used to diagnose cardiac amyloidosis, although [F]FBP may not be suitable for the distinction of different types of amyloid cardiomyopathy. No studies on PET in the follow-up of ATTR amyloidosis patients were found. Future research should focus on the use of these PET tracers in the follow-up of ATTR amyloidosis patients.
Topics: Humans; Prealbumin; Follow-Up Studies; Amyloidosis; Positron-Emission Tomography; Cardiomyopathies
PubMed: 37561144
DOI: 10.1007/s00259-023-06381-3 -
Gastroenterology Apr 2024Current international guidelines recommend duodenal biopsies to confirm the diagnosis of celiac disease in adult patients. However, growing evidence suggests that... (Meta-Analysis)
Meta-Analysis
BACKGROUND & AIMS
Current international guidelines recommend duodenal biopsies to confirm the diagnosis of celiac disease in adult patients. However, growing evidence suggests that immunoglobulin A (IgA) anti-tissue transglutaminase (tTg) antibody levels ≥10 times the upper limit of normal (ULN) can accurately predict celiac disease, eliminating the need for biopsy. We performed a systematic review and meta-analysis to evaluate the accuracy of the no-biopsy approach to confirm the diagnosis of celiac disease in adults.
METHODS
We systematically searched MEDLINE, EMBASE, Cochrane Library, and Web of Science from January 1998 to October 2023 for studies reporting the sensitivity and specificity of IgA-tTG ≥10×ULN against duodenal biopsies (Marsh grade ≥2) in adults with suspected celiac disease. We used a bivariate random effects model to calculate the summary estimates of sensitivity, specificity, and positive and negative likelihood ratios. The positive and negative likelihood ratios were used to calculate the positive predictive value of the no-biopsy approach across different pretest probabilities of celiac disease. The methodological quality of the included studies was evaluated using the QUADAS-2 tool. This study was registered with PROSPERO, number CRD42023398812.
RESULTS
A total of 18 studies comprising 12,103 participants from 15 countries were included. The pooled prevalence of biopsy-proven celiac disease in the included studies was 62% (95% confidence interval [CI], 40%-83%). The proportion of patients with IgA-tTG ≥10×ULN was 32% (95% CI, 24%-40%). The summary sensitivity of IgA-tTG ≥10×ULN was 51% (95% CI, 42%-60%), and the summary specificity was 100% (95% CI, 98%-100%). The area under the summary receiver operating characteristic curve was 0.83 (95% CI, 0.77 - 0.89). The positive predictive value of the no-biopsy approach to identify patients with celiac disease was 65%, 88%, 95%, and 99% if celiac disease prevalence was 1%, 4%, 10%, and 40%, respectively. Between-study heterogeneity was moderate (I =30.3%), and additional sensitivity analyses did not significantly alter our findings. Only 1 study had a low risk of bias across all domains.
CONCLUSION
The results of this meta-analysis suggest that selected adult patients with IgA-tTG ≥10×ULN and a moderate to high pretest probability of celiac disease could be diagnosed without undergoing invasive endoscopy and duodenal biopsy.
Topics: Adult; Humans; Celiac Disease; Transglutaminases; Protein Glutamine gamma Glutamyltransferase 2; Immunoglobulin A; GTP-Binding Proteins; Biopsy; Sensitivity and Specificity; Autoantibodies
PubMed: 38176661
DOI: 10.1053/j.gastro.2023.12.023 -
European Journal of Medical Research Feb 2024An essential relationship between insulin resistance (IR) and atrial fibrillation (AF) has been demonstrated. Among the methods used to assess IR, the... (Meta-Analysis)
Meta-Analysis Review
BACKGROUND
An essential relationship between insulin resistance (IR) and atrial fibrillation (AF) has been demonstrated. Among the methods used to assess IR, the triglyceride-glucose (TyG) index is the more straightforward, dimensionless, and low-cost tool. However, the possible usage of this index in clinical practice to predict and diagnose AF has yet to be determined and consolidated.
OBJECTIVE AND RATIONALE
Herein, we performed a systematic review and meta-analysis to assess the association between the TyG index and AF.
METHODS
Databases (PubMed, Embase, Scopus, and Web of Science) were systematically searched for studies evaluating the TyG index in AF. The inclusion criteria were observational studies investigating AF and TyG index correlation in individuals older than 18 years, while preclinical studies and those without the relevant data were excluded. Random effect meta-analyses comparing TyG levels between AF and non-AF cases, AF recurrence after radiofrequency ablation, and post-procedural AF were performed using standardized mean differences (SMD) with their matching 95% confidence intervals (CIs).
RESULTS
Our screening identified nine studies to be analyzed, including 6,171 participants including 886 with AF. The meta-analysis demonstrated that the TyG index resulted higher in patients with AF than non-AF counterparts (SMD 1.23, 95% CI 0.71 to 1.75, I 98%, P < 0.001). Subgroup analysis showed the same results for post-procedure AF (SMD 0.99, 95% CI 0.78 to 1.20, I 10%, P < 0.001) and post-ablation AF (SMD 1.25, 95% CI 1.07 to 1.43, I 46%, P < 0.001), while no difference was found in population-based cohorts (SMD 1.45, 95% CI - 0.41 to 3.31, I 100%, P = 0.13). Publication year (P = 0.036) and sample size (P = 0.003) showed significant associations with the effect size, using multivariable meta-regression.
CONCLUSION
The TyG index is an easy-to-measure surrogate marker of IR in patients with AF. Further clinical studies are warranted to demonstrate its ability for routine clinical use and as a screening tool.
Topics: Humans; Atrial Fibrillation; Glucose; Triglycerides; Biomarkers; Catheter Ablation; Insulin Resistance
PubMed: 38347644
DOI: 10.1186/s40001-024-01716-8 -
Journal of Medical Internet Research Jul 2023Tuberculosis (TB) was the leading infectious cause of mortality globally prior to COVID-19 and chest radiography has an important role in the detection, and subsequent... (Meta-Analysis)
Meta-Analysis
BACKGROUND
Tuberculosis (TB) was the leading infectious cause of mortality globally prior to COVID-19 and chest radiography has an important role in the detection, and subsequent diagnosis, of patients with this disease. The conventional experts reading has substantial within- and between-observer variability, indicating poor reliability of human readers. Substantial efforts have been made in utilizing various artificial intelligence-based algorithms to address the limitations of human reading of chest radiographs for diagnosing TB.
OBJECTIVE
This systematic literature review (SLR) aims to assess the performance of machine learning (ML) and deep learning (DL) in the detection of TB using chest radiography (chest x-ray [CXR]).
METHODS
In conducting and reporting the SLR, we followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. A total of 309 records were identified from Scopus, PubMed, and IEEE (Institute of Electrical and Electronics Engineers) databases. We independently screened, reviewed, and assessed all available records and included 47 studies that met the inclusion criteria in this SLR. We also performed the risk of bias assessment using Quality Assessment of Diagnostic Accuracy Studies version 2 (QUADAS-2) and meta-analysis of 10 included studies that provided confusion matrix results.
RESULTS
Various CXR data sets have been used in the included studies, with 2 of the most popular ones being Montgomery County (n=29) and Shenzhen (n=36) data sets. DL (n=34) was more commonly used than ML (n=7) in the included studies. Most studies used human radiologist's report as the reference standard. Support vector machine (n=5), k-nearest neighbors (n=3), and random forest (n=2) were the most popular ML approaches. Meanwhile, convolutional neural networks were the most commonly used DL techniques, with the 4 most popular applications being ResNet-50 (n=11), VGG-16 (n=8), VGG-19 (n=7), and AlexNet (n=6). Four performance metrics were popularly used, namely, accuracy (n=35), area under the curve (AUC; n=34), sensitivity (n=27), and specificity (n=23). In terms of the performance results, ML showed higher accuracy (mean ~93.71%) and sensitivity (mean ~92.55%), while on average DL models achieved better AUC (mean ~92.12%) and specificity (mean ~91.54%). Based on data from 10 studies that provided confusion matrix results, we estimated the pooled sensitivity and specificity of ML and DL methods to be 0.9857 (95% CI 0.9477-1.00) and 0.9805 (95% CI 0.9255-1.00), respectively. From the risk of bias assessment, 17 studies were regarded as having unclear risks for the reference standard aspect and 6 studies were regarded as having unclear risks for the flow and timing aspect. Only 2 included studies had built applications based on the proposed solutions.
CONCLUSIONS
Findings from this SLR confirm the high potential of both ML and DL for TB detection using CXR. Future studies need to pay a close attention on 2 aspects of risk of bias, namely, the reference standard and the flow and timing aspects.
TRIAL REGISTRATION
PROSPERO CRD42021277155; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=277155.
Topics: Humans; Artificial Intelligence; COVID-19; Deep Learning; Radiography; Reproducibility of Results; Tuberculosis; X-Rays
PubMed: 37399055
DOI: 10.2196/43154 -
Cureus Oct 2023Barrett's esophagus (BE) remains a significant precursor to esophageal adenocarcinoma, requiring accurate and efficient diagnosis and management. The increasing... (Review)
Review
Barrett's esophagus (BE) remains a significant precursor to esophageal adenocarcinoma, requiring accurate and efficient diagnosis and management. The increasing application of machine learning (ML) technologies presents a transformative opportunity for diagnosing and treating BE. This systematic review evaluates the effectiveness and accuracy of machine learning technologies in BE diagnosis and management by conducting a comprehensive search across PubMed, Scopus, and Web of Science databases up to the year 2023. The studies were organized into five categories: computer-aided systems, natural language processing and text-based systems, deep learning on histology and biopsy images, real-time and video analysis, and miscellaneous studies. Results indicate high sensitivity and specificity across machine learning applications. Specifically, computer-aided systems showed sensitivities ranging from 84% to 100% and specificities from 64% to 90.7%. Natural language processing and text-based systems achieved an accuracy as high as 98.7%. Deep learning techniques applied to histology and biopsy images displayed sensitivities up to greater than 90% and a specificity of 100%. Furthermore, real-time and video analysis technologies demonstrated high performance with assessment speeds of up to 48 frames per second (fps) and a mean average precision of 75.3%. Overall, the reviewed literature underscores the growing capability and efficiency of machine learning technologies in diagnosing and managing Barrett's esophagus, often outperforming traditional diagnostic methods. These findings highlight the promising future role of machine learning in enhancing clinical practice and improving patient care for Barrett's esophagus.
PubMed: 38021699
DOI: 10.7759/cureus.47755 -
Nutrition, Metabolism, and... Aug 2023The Controlling Nutritional Status (CONUT) score is a tool for assessing the risk of malnutrition (undernutrition) that can be calculated from albumin concentration,... (Meta-Analysis)
Meta-Analysis
AIMS
The Controlling Nutritional Status (CONUT) score is a tool for assessing the risk of malnutrition (undernutrition) that can be calculated from albumin concentration, total peripheral lymphocyte count, and total cholesterol concentration. CONUT score has been proposed as a promising prognostic marker in several clinical settings; however, a consensus on its prognostic value in patients with stroke is lacking. The aim of this systematic review and meta-analysis was to evaluate the relationship between CONUT score and clinical outcomes in patients with stroke based on all current available studies.
DATA SYNTHESIS
Systematic research on PubMed, Scopus and Web of Science from inception to February 2023 was performed on the association between CONUT score and clinical outcomes in patients with stroke. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses were followed. Methodological quality was evaluated using the Newcastle-Ottawa Scale quality assessment tool. Pooled effect estimation was calculated by a random-effect model. Through the initial literature search, 15 studies (all high-quality) including 16 929 patients were found to be eligible and analysed in the meta-analysis. A significant risk of malnutrition (in most studies defined by a CONUT score ≥5) was directly associated with mortality, higher risk of poor functional outcome according to the modified Rankin Scale and total infection development. Evidence was consistent for acute ischaemic stroke and preliminary for acute haemorrhagic stroke.
CONCLUSION
CONUT score is an independent prognostic indicator, and it is associated with major disability and infection development during hospitalisation.
PROSPERO ID
CRD42022306560.
Topics: Humans; Nutritional Status; Brain Ischemia; Stroke; Malnutrition; Prognosis; Retrospective Studies; Nutrition Assessment
PubMed: 37336716
DOI: 10.1016/j.numecd.2023.05.012 -
Translational Vision Science &... Jul 2023The purpose of this study was to perform a systematic review and meta-analysis to synthesize evidence from studies using deep learning (DL) to predict cardiovascular... (Meta-Analysis)
Meta-Analysis
PURPOSE
The purpose of this study was to perform a systematic review and meta-analysis to synthesize evidence from studies using deep learning (DL) to predict cardiovascular disease (CVD) risk from retinal images.
METHODS
A systematic literature search was performed in MEDLINE, Scopus, and Web of Science up to June 2022. We extracted data pertaining to predicted outcomes, model development, and validation and model performance metrics. Included studies were graded using the Quality Assessment of Diagnostic Accuracies Studies 2 tool. Model performance was pooled across eligible studies using a random-effects meta-analysis model.
RESULTS
A total of 26 studies were included in the analysis. There were 42 CVD risk-related outcomes predicted from retinal images were identified, including 33 CVD risk factors, 4 cardiac imaging biomarkers, 2 CVD risk scores, the presence of CVD, and incident CVD. Three studies that aimed to predict the development of future CVD events reported an area under the receiver operating curve (AUROC) between 0.68 and 0.81. Models that used retinal images as input data had a pooled mean absolute error of 3.19 years (95% confidence interval [CI] = 2.95-3.43) for age prediction; a pooled AUROC of 0.96 (95% CI = 0.95-0.97) for gender classification; a pooled AUROC of 0.80 (95% CI = 0.73-0.86) for diabetes detection; and a pooled AUROC of 0.86 (95% CI = 0.81-0.92) for the detection of chronic kidney disease. We observed a high level of heterogeneity and variation in study designs.
CONCLUSIONS
Although DL models appear to have reasonably good performance when it comes to predicting CVD risk, further work is necessary to evaluate the real-world applicability and predictive accuracy.
TRANSLATIONAL RELEVANCE
DL-based CVD risk assessment from retinal images holds great promise to be translated to clinical practice as a novel approach for CVD risk assessment, given its simple, quick, and noninvasive nature.
Topics: Humans; Cardiovascular Diseases; Deep Learning
PubMed: 37440249
DOI: 10.1167/tvst.12.7.14 -
ESC Heart Failure Aug 2023The clinical value of cardiopulmonary exercise testing (CPET) in cardiac amyloidosis (CA) is uncertain. Due to the growing prevalence of the disease and the current... (Meta-Analysis)
Meta-Analysis
BACKGROUND
The clinical value of cardiopulmonary exercise testing (CPET) in cardiac amyloidosis (CA) is uncertain. Due to the growing prevalence of the disease and the current availability of disease-modifying drugs, prognostic stratification is becoming fundamental to optimizing the cost-effectiveness of treatment, patient phenotyping, follow-up, and management. Peak VO and VE/VCO slope are currently the most studied CPET variables in clinical settings, and both demonstrate substantial, independent prognostic value in several cardiovascular diseases. We aim to study the association of peak VO and VE/VCO slope with prognosis in patients with CA.
METHODS AND RESULTS
We performed a systematic review and searched for clinical studies performing CPET for prognostication in patients with transthyretin-CA and light-chain-CA. Studies reporting hazard ratio (HR) for mortality and peak VO or VE/VCO slope were further selected for quantitative analysis. HRs were pooled using a random-effect model. Five studies were selected for qualitative and three for quantitative analysis. A total of 233 patients were included in the meta-analysis. Mean peak VO resulted consistently depressed, and VE/VCO slope was increased. Our pooled analysis showed peak VO (pooled HR 0.89, 95% CI 0.84-0.94) and VE/VCO slope (pooled HR 1.04, 95% CI 1.01-1.07) were significantly associated with the risk of death in CA patients, with no significant statistical heterogeneity for both analyses.
CONCLUSIONS
CPET is a valuable tool for prognostic stratification in CA, identifying patients at increased risk of death. Large prospective clinical trials are needed to confirm this exploratory finding.
Topics: Humans; Exercise Test; Prospective Studies; Oxygen Consumption; Prognosis; Amyloidosis; Cardiomyopathies
PubMed: 37264762
DOI: 10.1002/ehf2.14406