OBJECTIVE

To develop and evaluate the diagnostic performance of an automated machine learning (AutoML) model for the detection of referable diabetic retinopathy (refDR) in ultrawide field (UWF) retinal images from local Philippine retinal image datasets.

A Google AutoML Vision model was trained using 2000 UWF images with a 50/50 ratio of refDR/non-refDR. Images were labeled according to the Early Treatment Diabetic Retinopathy Study (ETDRS) severity grading. RefDR was defined as moderate nonproliferative DR or worse. The dataset was split with 80% for training, 10% for validation, and 10% for testing. Two sets of published UWF image sets were used for external validation. Sensitivity and specificity were calculated in accordance with United States Food and Drug Administration (US FDA) performance requirements of 0.85 and 0.825, respectively.

The area under the precision-recall curve was 0.998. External validation against two datasets showed a sensitivity/specificity of 0.88/0.83 (95% CI 0.80-0.94/0.74-0.89) and 0.83/0.80 (95% CI 0.74-0.89/0.72-0.86), respectively. Positive and negative predictive values were 0.81/0.89 (95% CI 0.73-0.89/0.82-0.94) and 0.75/0.86 (95% CI 0.66-0.83/0.79-0.91), respectively.

The pilot performance of the custom AutoML model constructed using local Philippine data approaches US FDA requirements for the diagnosis of referable DR. The ease of use and intuitiveness of the platform, combined with its performance, support the potential of no-code AI in the detection of refDR.

Humans ; Female ; Breast Neoplasms/pathology* ; Multiomics ; Trastuzumab ; Receptor, ErbB-2/genetics* ; Machine Learning

Using machine learning algorithms to construct an early prediction model of brucellosis to improve the diagnosis efficiency of Brucellosis. This study was a case-control study. 2 381 brucellosis patients from Beijing Ditan Hospital affiliated to Capital Medical University were retrospectively collected as case group, and healthy people from Beijing Chaoyang Hospital affiliated to Capital Medical University were collected as control group from May 9, 2011 to November 29, 2021. The relevant clinical information and full blood count results of 13 257 data were collected and five algorithms of machine learning were used to construct an early predication model of brucellosis by using machine learning: random forest, Naive Bayes, decision tree, logistic regression and support vector machine;14 074 data (2 143 cases incase group and 11 931 cases in control group) were used to establish the early predication model of brucellosis, and 1 564 (238 cases in case group and 1 326 cases in control group) data were used to test the predication efficiency of the brucellosis model. The results showed that the support vector machine algorithm has the best predication performance by comparing the five machine learning models. The area under receiver curve (AUC) of receiver operating characteristic (ROC) was 0.991, and the accuracy, precision, specificity and Recall were 95.6%, 95.5%, 95.4% and 95.9%, respectively. Based on the SHAP plot, platelet distribution width (PDW) and basophil relative value (BASO%) results were low, and men with high coefficient of variation (R-CV), erythrocyte hemoglobin concentration (MCHC), and platelet volume (MPV) were predicted to be at high risk of brucellosis. Platelet distribution width (PDW) contributed the most to the prediction model, followed by red blood cell distribution width coefficient of variation (R-CV). In conclusion, the establishment of a high-precision early predication method of brucellosis based on machine learning may be of great significance for the early detection and treatment of brucellosis patients.
Male ; Humans ; Retrospective Studies ; Case-Control Studies ; Bayes Theorem ; Algorithms ; Machine Learning

This study used machine learning techniques combined with routine blood cell analysis parameters to build preliminary prediction models, helping differentiate patients with depression from healthy controls, or patients with anxiety. A multicenter study was performed by collecting blood cell analysis data of Beijing Chaoyang Hospital and the First Hospital of Hebei Medical University from 2020 to 2021. Machine learning techniques, including support vector machine, decision tree, naïve Bayes, random forest and multi-layer perceptron were explored to establish a prediction model of depression. The results showed that based on the blood cell analysis results of healthy controls and depression group, the accuracy of prediction model reached as high as 0.99, F1 was 0.975. Receiver operating characteristic curve area and average accuracy were 0.985 and 0.967, respectively. Platelet parameters contributed mostly to depression prediction model. While, to random forest differential diagnosis model based on the data from depression and anxiety groups, prediction accuracy reached 0.68 and AUC 0.622. Age, platelet parameters, and average volume of red blood cells contributed the most to the model. In conclusion, the study researched on the prediction model of depression by exploring blood cell analysis parameters, revealing that machine learning models were more objective in the evaluation of mental illness.
Humans ; Depression ; Bayes Theorem ; Machine Learning ; Support Vector Machine ; Blood Cell Count

Kidney disease affects a large number of people around the world, imposing a significant burden to people's health and life. If early prediction, rapid diagnosis and prognosis prediction of kidney disease can be carried out, the health of patients will be better protected. Machine learning belongs to the category of artificial intelligence, which can be divided into supervised learning, unsupervised learning and reinforcement learning. With the increasing requirements for the processing and analyzing large-scale and high-dimensional data, machine learning is playing an increasingly important role in the medical domain, and the field of kidney disease is no exception. This article presents a comprehensive overview of the application progress of machine learning in kidney disease, aiming to make medical staff's decision-making in kidney disease more early, accurate and rapid, and better escort the life and health of patients.
Humans ; Artificial Intelligence ; Machine Learning ; Kidney ; Kidney Diseases/diagnosis*

Humans ; Parkinson Disease/diagnosis* ; Machine Learning

Humans ; Monocytes ; Machine Learning ; HIV Infections/genetics*

The present study aims to establish comprehensive evaluation models of physical fitness of the elderly based on machine learning, and provide an important basis to monitor the elderly's physique. Through stratified sampling, the elderly aged 60 years and above were selected from 10 communities in Nanchang City. The physical fitness of the elderly was measured by the comprehensive physical assessment scale based on our previous study. Fuzzy neural network (FNN), support vector machine (SVM) and random forest (RF) models for comprehensive physical evaluation of the elderly people in communities were constructed respectively. The accuracy, sensitivity and specificity of the comprehensive physical fitness evaluation models constructed by FNN, SVM and RF were above 0.85, 0.75 and 0.89, respectively, with the FNN model possessing the best prediction performance. FNN, RF and SVM models are valuable in the comprehensive evaluation and prediction of physical fitness, which can be used as tools to carry out physical evaluation of the elderly.
Aged ; Humans ; Physical Fitness ; Neural Networks, Computer ; Exercise ; Machine Learning

Humans ; Aged ; Built Environment ; Exercise ; Machine Learning ; China

This study aimed to establish a method based on machine learning technology for accurately predicting the commodity specifications of Fritillariae Cirrhosae Bulbus and explore the application of data augmentation technology in the field of drug analysis. The correlation optimized warping(COW) algorithm was used to perform peak calibration on the UPLC-QDA multi-channel superimposed data of 30 batches of samples, and the data were normalized. Through unsupervised learning methods such as clustering analysis, principal component analysis(PCA), and correlation analysis, the general characteristics of the data were understood. Then, the logistic regression algorithm was used for supervised learning on the data, and the condition tabular generative adversarial networks(CTGAN) was used to generate a large amount of data. Logistic regression classification models were trained separately using the real data and the data generated by CTGAN, and these models were evaluated. The logistic regression model trained with real data achieved cross-validation and test set accuracies of 0.95 and 1.00, respectively, while the logistic regression model trained with both real and CTGAN-generated data achieved cross-validation and test set accuracies of 0.99 and 1.00, respectively. The results indicate that machine learning can accurately predict the classification of Songbei, Qingbei, and Lubeibased on UPLC-QDA detection data. CTGAN-generated data can partially compensate for the lack of data in drug analysis, improving the accuracy and predictive ability of machine learning models.
Drugs, Chinese Herbal ; Fritillaria ; Technology ; Machine Learning ; Plant Roots