-
Genes Aug 2023Mucopolysaccharidosis-plus syndrome (MPSPS) is an autosomal-recessive disorder caused by c.1492C>T (p.R498W) in the gene. MPSPS is a severe disorder that causes a short...
Mucopolysaccharidosis-plus syndrome (MPSPS) is an autosomal-recessive disorder caused by c.1492C>T (p.R498W) in the gene. MPSPS is a severe disorder that causes a short lifespan in patients. Currently, there is no specific treatment for patients. The Yakut population is more prone to this disease than others. Diagnosing MPSPS relies on clinical manifestations, and genetic testing (GT) is used to confirm the diagnosis. In this research, we examined two pregnancy cases, one of which involved a prenatal diagnosis for MPSPS. Notably, neither pregnant woman had a known family history of the disorder. During their pregnancies, both women underwent prenatal ultrasonography, which revealed increased prenasal thickness during the second trimester. In the first case, ultrasonography indicated increased prenasal thickness in the second trimester, but a definitive diagnosis was not made at that time. The patient was eventually diagnosed with MPSPS at 11 months of age. On the contrary, in the second case, GT uncovered that the parents were carriers of MPSPS. Consequently, a placental biopsy was performed, leading to an early diagnosis of MPSPS. This study emphasizes the importance of ultrasonography findings in prenatal MPSPS diagnosis. Combining ultrasonography with GT can be a valuable approach to confirming MPSPS at an early stage, allowing for the appropriate planning of delivery methods and medical care. Ultimately, this comprehensive approach can significantly enhance the quality of life of both affected patients and their parents.
Topics: Pregnancy; Humans; Female; Quality of Life; Placenta; Prenatal Diagnosis; Genetic Testing; Mucopolysaccharidoses
PubMed: 37628632
DOI: 10.3390/genes14081581 -
Frontiers in Big Data 2023The comorbidity network represents multiple diseases and their relationships in a graph. Understanding comorbidity networks among critical care unit (CCU) patients can...
IMPORTANCE
The comorbidity network represents multiple diseases and their relationships in a graph. Understanding comorbidity networks among critical care unit (CCU) patients can help doctors diagnose patients faster, minimize missed diagnoses, and potentially decrease morbidity and mortality.
OBJECTIVE
The main objective of this study was to identify the comorbidity network among CCU patients using a novel application of a machine learning method (graphical modeling method). The second objective was to compare the machine learning method with a traditional pairwise method in simulation.
METHOD
This cross-sectional study used CCU patients' data from Medical Information Mart for the Intensive Care-3 (MIMIC-3) dataset, an electronic health record (EHR) of patients with CCU hospitalizations within Beth Israel Deaconess Hospital from 2001 to 2012. A machine learning method (graphical modeling method) was applied to identify the comorbidity network of 654 diagnosis categories among 46,511 patients.
RESULTS
Out of the 654 diagnosis categories, the graphical modeling method identified a comorbidity network of 2,806 associations in 510 diagnosis categories. Two medical professionals reviewed the comorbidity network and confirmed that the associations were consistent with current medical understanding. Moreover, the strongest association in our network was between "poisoning by psychotropic agents" and "accidental poisoning by tranquilizers" (logOR 8.16), and the most connected diagnosis was "disorders of fluid, electrolyte, and acid-base balance" (63 associated diagnosis categories). Our method outperformed traditional pairwise comorbidity network methods in simulation studies. Some strongest associations between diagnosis categories were also identified, for example, "diagnoses of mitral and aortic valve" and "other rheumatic heart disease" (logOR: 5.15). Furthermore, our method identified diagnosis categories that were connected with most other diagnosis categories, for example, "disorders of fluid, electrolyte, and acid-base balance" was associated with 63 other diagnosis categories. Additionally, using a data-driven approach, our method partitioned the diagnosis categories into 14 modularity classes.
CONCLUSION AND RELEVANCE
Our graphical modeling method inferred a logical comorbidity network whose associations were consistent with current medical understanding and outperformed traditional network methods in simulation. Our comorbidity network method can potentially assist CCU doctors in diagnosing patients faster and minimizing missed diagnoses.
PubMed: 37663273
DOI: 10.3389/fdata.2023.846202 -
PloS One 2023Coronaviruses have affected the lives of people around the world. Increasingly, studies have indicated that the virus is mutating and becoming more contagious. Hence,...
Coronaviruses have affected the lives of people around the world. Increasingly, studies have indicated that the virus is mutating and becoming more contagious. Hence, the pressing priority is to swiftly and accurately predict patient outcomes. In addition, physicians and patients increasingly need interpretability when building machine models in healthcare. We propose an interpretable machine framework(KISM) that can diagnose and prognose patients based on blood test datasets. First, we use k-nearest neighbors, isolated forests, and SMOTE to pre-process the original blood test datasets. Seven machine learning tools Support Vector Machine, Extra Tree, Random Forest, Gradient Boosting Decision Tree, eXtreme Gradient Boosting, Logistic Regression, and ensemble learning were then used to diagnose and predict COVID-19. In addition, we used SHAP and scikit-learn post-hoc interpretability to report feature importance, allowing healthcare professionals and artificial intelligence models to interact to suggest biomarkers that some doctors may have missed. The 10-fold cross-validation of two public datasets shows that the performance of KISM is better than that of the current state-of-the-art methods. In the diagnostic COVID-19 task, an AUC value of 0.9869 and an accuracy of 0.9787 were obtained, and ultimately Leukocytes, platelets, and Proteina C reativa mg/dL were found to be the most indicative biomarkers for the diagnosis of COVID-19. An AUC value of 0.9949 and an accuracy of 0.9677 were obtained in the prognostic COVID-19 task and Age, LYMPH, and WBC were found to be the most indicative biomarkers for identifying the severity of the patient.
Topics: Humans; COVID-19; Artificial Intelligence; Prognosis; Machine Learning; Blood Platelets; COVID-19 Testing
PubMed: 37733828
DOI: 10.1371/journal.pone.0291961 -
Nature Medicine Feb 2024Although advances in deep learning systems for image-based medical diagnosis demonstrate their potential to augment clinical decision-making, the effectiveness of...
Although advances in deep learning systems for image-based medical diagnosis demonstrate their potential to augment clinical decision-making, the effectiveness of physician-machine partnerships remains an open question, in part because physicians and algorithms are both susceptible to systematic errors, especially for diagnosis of underrepresented populations. Here we present results from a large-scale digital experiment involving board-certified dermatologists (n = 389) and primary-care physicians (n = 459) from 39 countries to evaluate the accuracy of diagnoses submitted by physicians in a store-and-forward teledermatology simulation. In this experiment, physicians were presented with 364 images spanning 46 skin diseases and asked to submit up to four differential diagnoses. Specialists and generalists achieved diagnostic accuracies of 38% and 19%, respectively, but both specialists and generalists were four percentage points less accurate for the diagnosis of images of dark skin as compared to light skin. Fair deep learning system decision support improved the diagnostic accuracy of both specialists and generalists by more than 33%, but exacerbated the gap in the diagnostic accuracy of generalists across skin tones. These results demonstrate that well-designed physician-machine partnerships can enhance the diagnostic accuracy of physicians, illustrating that success in improving overall diagnostic accuracy does not necessarily address bias.
Topics: Humans; Deep Learning; Skin Pigmentation; Skin Diseases; Algorithms; Diagnosis, Differential
PubMed: 38317019
DOI: 10.1038/s41591-023-02728-3 -
International Wound Journal Jan 2024Necrotizing fasciitis is a clinical, surgical emergency characterized by an insidious onset, rapid progression, and a high mortality rate. The disease's mortality rate... (Review)
Review
Necrotizing fasciitis is a clinical, surgical emergency characterized by an insidious onset, rapid progression, and a high mortality rate. The disease's mortality rate has remained high for many years, mainly because of its atypical clinical presentation, which prevents many cases from being diagnosed early and accurately, resulting in patients who may die from uncontrollable septic shock and multi-organ failure. But unfortunately, no diagnostic indicator can provide a certain early diagnosis of NF, and clinical judgement of NF is still based on the results of various ancillary tests combined with the patient's medical history, clinical manifestations, and the physician's experience. This review provides a brief overview of the epidemiological features of NF and then discusses the most important laboratory indicators and scoring systems currently employed to diagnose NF. Finally, the latest progress of several imaging techniques in the early diagnosis of NF and their combined application with other diagnostic indices are highlighted. We point out promising research directions based on an objective evaluation of the advantages and shortcomings of different methods, which provide a basis for further improving the early diagnosis of NF.
Topics: Humans; Fasciitis, Necrotizing; Shock, Septic; Early Diagnosis; Retrospective Studies
PubMed: 37679292
DOI: 10.1111/iwj.14379 -
Academic Radiology Oct 2023This study aimed to investigate the reliability and accuracy of high-resolution ultrasonography (US) for diagnosing periapical lesions and differentiating radicular...
RATIONALE AND OBJECTIVES
This study aimed to investigate the reliability and accuracy of high-resolution ultrasonography (US) for diagnosing periapical lesions and differentiating radicular cysts from granulomas.
MATERIALS AND METHODS
This study included 109 teeth with periapical lesions of endodontic origin from 109 patients scheduled for apical microsurgery. Ultrasonic outcomes were analyzed and categorized after thorough clinical and radiographic examinations using US. B-mode US images reflected the echotexture, echogenicity, and lesion margin, while color Doppler US assessed the presence and features of blood flow of interested areas. Pathological tissue samples were obtained during apical microsurgery and subjected to histopathological examination. Fleiss' κ was used to measure interobserver reliability. Statistical analyses were performed to assess the diagnostic validity and the overall agreement between US and histological findings. The reliability of US compared to histopathological examinations was assessed based on Cohen's κ.
RESULTS
The percent accuracy of US for diagnosing cysts, granulomas, and cysts with infection based on histopathological findings was 89.9%, 89.0%, and 97.2%, respectively. The sensitivity of US diagnoses was 95.1% for cysts, 84.1% for granulomas, and 80.0% for cysts with infection. The specificity of US diagnoses was 86.8% for cysts, 95.7% for granulomas, and 98.1% for cysts with infection. The reliability for US compared to histopathological examinations was good (κ = 0.779).
CONCLUSION
The echotexture characteristics of lesions in US images correlated with their histopathological features. US can provide accurate information on the nature of periapical lesions based on the echotexture of their contents and the presence of vascularity. It can help improve clinical diagnosis and avoid overtreatment of patients with apical periodontitis.
Topics: Humans; Radicular Cyst; Periapical Granuloma; Reproducibility of Results; Granuloma; Ultrasonography
PubMed: 37394410
DOI: 10.1016/j.acra.2023.05.039 -
Archives of Pathology & Laboratory... Mar 2024Immunohistochemistry has become a valuable ancillary tool for the accurate classification of pleuropulmonary and mediastinal neoplasms necessary for therapeutic... (Review)
Review
CONTEXT.—
Immunohistochemistry has become a valuable ancillary tool for the accurate classification of pleuropulmonary and mediastinal neoplasms necessary for therapeutic decisions and predicting prognostic outcome. Diagnostic accuracy has significantly improved because of the continuous discoveries of tumor-associated biomarkers and the development of effective immunohistochemical panels.
OBJECTIVE.—
To increase the accuracy of diagnosis and classify pleuropulmonary neoplasms through immunohistochemistry.
DATA SOURCES.—
Literature review and the author's research data and personal practice experience.
CONCLUSIONS.—
This review article highlights that appropriately selecting immunohistochemical panels enables pathologists to effectively diagnose most primary pleuropulmonary neoplasms and differentiate primary lung tumors from a variety of metastatic tumors to the lung. Knowing the utilities and pitfalls of each tumor-associated biomarker is essential to avoid potential diagnostic errors.
Topics: Humans; Mediastinal Neoplasms; Immunohistochemistry; Biomarkers, Tumor; Prognosis; Diagnosis, Differential
PubMed: 37406295
DOI: 10.5858/arpa.2022-0483-RA -
Scientific Reports Oct 2023Although the placement of an intraventricular catheter remains the gold standard method for the diagnosis of intracranial hypertension (ICH), the technique has several...
Although the placement of an intraventricular catheter remains the gold standard method for the diagnosis of intracranial hypertension (ICH), the technique has several limitations including but not limited to its invasiveness. Current noninvasive methods, however, still lack robust evidence to support their clinical use. We aimed to estimate, as an exploratory hypothesis generating analysis, the discriminative power of four noninvasive methods to diagnose ICH. We prospectively collected data from adult intensive care unit (ICU) patients with subarachnoid hemorrhage (SAH), intraparenchymal hemorrhage (IPH), and ischemic stroke (IS) in whom invasive intracranial pressure (ICP) monitoring had been placed. Measures were simultaneously collected from the following noninvasive methods: optic nerve sheath diameter (ONSD), pulsatility index (PI) using transcranial Doppler (TCD), a 5-point visual scale designed for brain Computed Tomography (CT), and two parameters (time-to-peak [TTP] and P2/P1 ratio) of a noninvasive ICP wave morphology monitor (Brain4Care[B4c]). ICH was defined as a sustained ICP > 20 mmHg for at least 5 min. We studied 18 patients (SAH = 14; ICH = 3; IS = 1) on 60 occasions with a mean age of 52 ± 14.3 years. All methods were recorded simultaneously, except for the CT, which was performed within 24 h of the other methods. The median ICP was 13 [9.8-16.2] mmHg, and intracranial hypertension was present on 18 occasions (30%). Median values from the noninvasive techniques were ONSD 4.9 [4.40-5.41] mm, PI 1.22 [1.04-1.43], CT scale 3 points [IQR: 3.0], P2/P1 ratio 1.16 [1.09-1.23], and TTP 0.215 [0.193-0.237]. There was a significant statistical correlation between all the noninvasive techniques and invasive ICP (ONSD, r = 0.29; PI, r = 0.62; CT, r = 0.21; P2/P1 ratio, r = 0.35; TTP, r = 0.35, p < 0.001 for all comparisons). The area under the curve (AUC) to estimate intracranial hypertension was 0.69 [CIs = 0.62-0.78] for the ONSD, 0.75 [95% CIs 0.69-0.83] for the PI, 0.64 [95%Cis 0.59-069] for CT, 0.79 [95% CIs 0.72-0.93] for P2/P1 ratio, and 0.69 [95% CIs 0.60-0.74] for TTP. When the various techniques were combined, an AUC of 0.86 [0.76-0.93]) was obtained. The best pair of methods was the TCD and B4cth an AUC of 0.80 (0.72-0.88). Noninvasive technique measurements correlate with ICP and have an acceptable discrimination ability in diagnosing ICH. The multimodal combination of PI (TCD) and wave morphology monitor may improve the ability of the noninvasive methods to diagnose ICH. The observed variability in non-invasive ICP estimations underscores the need for comprehensive investigations to elucidate the optimal method-application alignment across distinct clinical scenarios.
Topics: Adult; Humans; Middle Aged; Aged; Intracranial Pressure; Sensitivity and Specificity; Optic Nerve; Ultrasonography, Doppler, Transcranial; Intracranial Hypertension; Subarachnoid Hemorrhage; Ischemic Stroke; Ultrasonography
PubMed: 37891406
DOI: 10.1038/s41598-023-45834-5 -
BMC Medical Informatics and Decision... Aug 2023Differentiating between Crohn's disease (CD) and intestinal tuberculosis (ITB) with endoscopy is challenging. We aim to perform more accurate endoscopic diagnosis...
BACKGROUND
Differentiating between Crohn's disease (CD) and intestinal tuberculosis (ITB) with endoscopy is challenging. We aim to perform more accurate endoscopic diagnosis between CD and ITB by building a trustworthy AI differential diagnosis application.
METHODS
A total of 1271 electronic health record (EHR) patients who had undergone colonoscopies at Peking Union Medical College Hospital (PUMCH) and were clinically diagnosed with CD (n = 875) or ITB (n = 396) were used in this study. We build a workflow to make diagnoses with EHRs and mine differential diagnosis features; this involves finetuning the pretrained language models, distilling them into a light and efficient TextCNN model, interpreting the neural network and selecting differential attribution features, and then adopting manual feature checking and carrying out debias training.
RESULTS
The accuracy of debiased TextCNN on differential diagnosis between CD and ITB is 0.83 (CR F1: 0.87, ITB F1: 0.77), which is the best among the baselines. On the noisy validation set, its accuracy was 0.70 (CR F1: 0.87, ITB: 0.69), which was significantly higher than that of models without debias. We also find that the debiased model more easily mines the diagnostically significant features. The debiased TextCNN unearthed 39 diagnostic features in the form of phrases, 17 of which were key diagnostic features recognized by the guidelines.
CONCLUSION
We build a trustworthy AI differential diagnosis application for differentiating between CD and ITB focusing on accuracy, interpretability and robustness. The classifiers perform well, and the features which had statistical significance were in agreement with clinical guidelines.
Topics: Humans; Crohn Disease; Diagnosis, Differential; Tuberculosis, Gastrointestinal; Colonoscopy
PubMed: 37582768
DOI: 10.1186/s12911-023-02257-6 -
Annals of Medicine Dec 2023The persistent spread of SARS-CoV-2 makes diagnosis challenging because COVID-19 symptoms are hard to differentiate from those of other respiratory illnesses. The...
OBJECTIVE
The persistent spread of SARS-CoV-2 makes diagnosis challenging because COVID-19 symptoms are hard to differentiate from those of other respiratory illnesses. The reverse transcription-polymerase chain reaction test is the current golden standard for diagnosing various respiratory diseases, including COVID-19. However, this standard diagnostic method is prone to erroneous and false negative results (10% -15%). Therefore, finding an alternative technique to validate the RT-PCR test is paramount. Artificial intelligence (AI) and machine learning (ML) applications are extensively used in medical research. Hence, this study focused on developing a decision support system using AI to diagnose mild-moderate COVID-19 from other similar diseases using demographic and clinical markers. Severe COVID-19 cases were not considered in this study since fatality rates have dropped considerably after introducing COVID-19 vaccines.
METHODS
A custom stacked ensemble model consisting of various heterogeneous algorithms has been utilized for prediction. Four deep learning algorithms have also been tested and compared, such as one-dimensional convolutional neural networks, long short-term memory networks, deep neural networks and Residual Multi-Layer Perceptron. Five explainers, namely, Shapley Additive Values, Eli5, QLattice, Anchor and Local Interpretable Model-agnostic Explanations, have been utilized to interpret the predictions made by the classifiers.
RESULTS
After using Pearson's correlation and particle swarm optimization feature selection, the final stack obtained a maximum accuracy of 89%. The most important markers which were useful in COVID-19 diagnosis are Eosinophil, Albumin, T. Bilirubin, ALP, ALT, AST, HbA1c and TWBC.
CONCLUSION
The promising results suggest using this decision support system to diagnose COVID-19 from other similar respiratory illnesses.
Topics: Humans; COVID-19; Artificial Intelligence; SARS-CoV-2; COVID-19 Vaccines; COVID-19 Testing
PubMed: 37436038
DOI: 10.1080/07853890.2023.2233541