Herramientas de corte para optimizar parámetros de clasificación de especies maderables con redes neuronales convolucionales / Cutting tools to optimize classification parameters of timber species with convolutional neural networks

Introducción: La gran diversidad de especies maderables tropicales demanda el desarrollo de nuevas tecnologías de identificación con base en sus patrones o características anatómicas. La aplicación de redes neuronales convolucionales (CNN) para el reconocimiento de especies maderables tropicales se ha incrementado en los últimos años por sus resultados prometedores. Objetivo: Evaluamos la calidad de las imágenes macroscópicas con tres herramientas de corte para mejorar la visualización y distinción de las características anatómicas en el entrenamiento del modelo CNN. Métodos: Recolectamos las muestras entre el 2020 y 2021 en áreas de explotación forestal y aserraderos de Selva Central, Perú. Luego, las dimensionamos y, previo a la identificación botánica y anatómica, las cortamos en secciones transversales. Generamos una base de datos de imágenes macroscópicas de la sección transversal de la madera, a través del corte, con tres herramientas para ver su rendimiento en el laboratorio, campo y puesto de control. Resultados: Usamos tres herramientas de corte para obtener una alta calidad de imágenes transversales de la madera; obtuvimos 3 750 imágenes macroscópicas con un microscopio portátil que corresponden a 25 especies maderables. El cuchillo ''Tramontina'' es duradero, pero pierde el filo con facilidad y se necesita una herramienta para afilar, el cúter retráctil ''Pretul'' es adecuado para madera suave y dura en muestras pequeñas de laboratorio; el cuchillo ''Ubermann'' es apropiado para el campo, laboratorio y puesto de control, porque tiene una envoltura duradera y láminas intercambiables en caso de pérdida de filo. Conclusiones: La calidad de las imágenes es decisiva en la clasificación de especies maderables, porque permite una mejor visualización y distinción de las características anatómicas en el entrenamiento con los modelos de red neuronal convolucional EfficientNet B0 y Custom Vision, lo cual se evidenció en las métricas de precisión.

Introduction: The great diversity of tropical timber species demands the development of new technologies capable of identifying them based on their patterns or anatomical characteristics. The application of convolutional neural networks (CNN) for the recognition of tropical timber species has increased in recent years due to the promising results of CNNs. Objective: To evaluate the quality of macroscopic images with three cutting tools to improve the visualization and distinction of anatomical features in the CNN model training. Methods: Samples were collected from 2020 to 2021 in areas of logging and sawmills in the Central Jungle, Peru. They were later sized and, after botanical and anatomical identification, cut in cross sections. A database of macroscopic images of the cross-section of wood was generated through cutting with three different tools and observing its performance in the laboratory, field, and checkpoint. Results: Using three cutting tools, we obtained high quality images of the cross section of wood; 3 750 macroscopic images were obtained with a portable microscope and correspond to 25 timber species. We found the ''Tramontina'' knife to be durable, however, it loses its edge easily and requires a sharpening tool, the ''Pretul'' retractable cutter is suitable for cutting soft and hard wood in small laboratory samples and finally the ''Ubermann'' knife is suitable for use in the field, laboratory, and checkpoint, because it has a durable sheath and interchangeable blades in case of dullness. Conclusion: The quality of the images is decisive in the classification of timber species, because it allows a better visualization and distinction of the anatomical characteristics in training with the EfficientNet B0 and Custom Vision convolutional neural network models, which was evidenced in the precision metrics.

Sex Determination by the Machine Learning Algorithms Through Using Morphometric Measurements of the Carpal, Metacarpal, and Phalangeal Bones / Determinación del Sexo Mediante Algoritmos de Aprendizaje Automático Mediante el Uso de Mediciones Morfométricas de los Huesos del Carpo, el Metacarpo y las Falanges

SUMMARY: In the study, it was aimed to predict sex from hand measurements using machine learning algorithms (MLA). Measurements were made on MR images of 60 men and 60 women. Determined parameters; hand length (HL), palm length (PL), hand width (HW), wrist width (EBG), metacarpal I length (MIL), metacarpal I width (MIW), metacarpal II length (MIIL), metacarpal II width (MIIW), metacarpal III length (MIIL), metacarpal III width (MIIIW), metacarpal IV length (MIVL), metacarpal IV width (MIVW), metacarpal V length (MVL), metacarpal V width (MVW), phalanx I length (PILL), measured as phalanx II length (PIIL), phalanx III length (PIIL), phalanx IV length (PIVL), phalanx V length (PVL). In addition, the hand index (HI) was calculated. Logistic Regression (LR), Random Forest (RF), Linear Discriminant Analysis (LDA), K-nearest neighbour (KNN) and Naive Bayes (NB) were used as MLAs. In the study, the KNN algorithm's Accuracy, SEN, F1 and Specificity ratios were determined as 88 %. In this study using MLA, it is understood that the highest accuracy belongs to the KNN algorithm. Except for the hand's MIIW, MIIIW, MIVW, MVW, HI variables, other variables were statistically significant in terms of sex difference.

En el estudio, el objetivo era predecir el sexo a partir de mediciones manuales utilizando algoritmos de aprendizaje automático (MLA). Las mediciones se realizaron en imágenes de RM de 60 hombres y 60 mujeres. Parámetros determinados; longitud de la mano (HL), longitud de la palma (PL), ancho de la mano (HW), ancho de la muñeca (EBG), longitud del metacarpiano I (MIL), ancho del metacarpiano I (MIW), longitud del metacarpiano II (MIIL), ancho del metacarpiano II (MIIW), longitud del metacarpiano III (MIIL), ancho del metacarpiano III (MIIIW), longitud del metacarpiano IV (MIVL), ancho del metacarpiano IV (MIVW), longitud del metacarpiano V (MVL), ancho del metacarpiano V (MVW), longitud de la falange I (PILL), medido como longitud de la falange II (PIIL), longitud de la falange III (PIIL), longitud de la falange IV (PIVL), longitud de la falange V (PVL). Además, se calculó el índice de la mano (HI). Regresión logística (LR), Random Forest (RF), Análisis discriminante lineal (LDA), K-vecino más cercano (KNN) y Naive Bayes (NB) se utilizaron como MLA. En el estudio, las proporciones de precisión, SEN, F1 y especificidad del algoritmo KNN se determinaron en un 88 %. En este estudio que utiliza MLA, se entiende que la mayor precisión pertenece al algoritmo KNN. Excepto por las variables MIIW, MIIIW, MIVW, MVW, HI de la mano, otras variables fueron estadísticamente significativas en términos de diferencia de sexo.

Inteligencia artificial en enfermedades infecciosas / Artificial intelligence in infectious diseases

Estamos asistiendo a una verdadera revolución tecnológi-ca en el campo de la salud. Los procesos basados en la aplicación de la inteligencia artificial (IA) y el aprendizaje automático (AA) están llegando progresivamente a todas las áreas disciplinares, y su aplicación en el campo de las enfermedades infecciosas es ya vertiginoso, acelerado por la pandemia de COVID-19.Hoy disponemos de herramientas que no solamente pue-den asistir o llevar adelante el proceso de toma de deci-siones basadas en guías o algoritmos, sino que también pueden modificar su desempeño a partir de los procesos previamente realizados. Desde la optimización en la identificación de microorganis-mos resistentes, la selección de candidatos a participar en ensayos clínicos, la búsqueda de nuevos agentes terapéu-ticos antimicrobianos, el desarrollo de nuevas vacunas, la predicción de futuras epidemias y pandemias, y el segui-miento clínico de pacientes con enfermedades infecciosas hasta la asignación de recursos en el curso de manejo de un brote son actividades que hoy ya pueden valerse de la inteligencia artificial para obtener un mejor resultado. El desarrollo de la IA tiene un potencial de aplicación expo-nencial y sin dudas será uno de los determinantes principa-les que moldearán la actividad médica del futuro cercano.Sin embargo, la maduración de esta tecnología, necesaria para su inserción definitiva en las actividades cotidianas del cuidado de la salud, requiere la definición de paráme-tros de referencia, sistemas de validación y lineamientos regulatorios que todavía no existen o son aún solo inci-pientes

We are in the midst of a true technological revolution in healthcare. Processes based upon artificial intelligence and machine learning are progressively touching all disciplinary areas, and its implementation in the field of infectious diseases is astonishing, accelerated by the COVID-19 pandemic. Today we have tools that can not only assist or carry on decision-making processes based upon guidelines or algorithms, but also modify its performance from the previously completed tasks. From optimization of the identification of resistant pathogens, selection of candidates for participating in clinical trials, the search of new antimicrobial therapeutic agents, the development of new vaccines, the prediction of future epidemics and pandemics, the clinical follow up of patients suffering infectious diseases up to the resource allocation in the management of an outbreak, are all current activities that can apply artificial intelligence in order to improve their final outcomes.This development has an exponential possibility of application, and is undoubtedly one of the main determinants that will shape medical activity in the future.Notwithstanding the maturation of this technology that is required for its definitive insertion in day-to-day healthcare activities, should be accompanied by definition of reference parameters, validation systems and regulatory guidelines that do not exist yet or are still in its initial stages

Crista Galli Morphometry and Morphology: An Anatomical, Radiological, and Machine Learning Application Study / Crista Galli Morfometría y Morfología: Un Estudio de Aplicación Anatómica, Radiológica y de Aprendizaje Automático

SUMMARY: The study purposed to examine the morphometry and morphology of crista galli in cone beam computed tomography (CBCT) and apply a new analysis, supervised Machine Learning techniques to find the answers to research questions "Can sex be determined with crista galli morphometric measurements?" or "How effective are the crista galli morphometric measurements in determining sex?". Crista galli dimensions including anteroposterior, superoinferior, and laterolateral were measured and carried out on 200 healthy adult subjects (98 females; 102 males) aged between 18-79 years. Also, crista galli was classified with two methods called morphological types and Keros classification. In this study, the Chi-square test, Student's t-test, and Oneway ANOVA were performed. Additionally, Machine Learning techniques were applied. The means of the CGH, CGW, and CGL were found as 14.96 mm; 3.96 mm, and 12.76 mm in males, respectively. The same values were as 13.54 mm; 3.51 mm and 11.59±1.61 mm in females, respectively. The CG morphometric measurements of males were higher than those of females. There was a significant difference between sexes in terms of morphological classification type. Also, when the sex assignment of JRip was analyzed, out of 102 male instances 62 of them were correctly predicted, and for 98 female instances, 70 of them were correctly predicted according to their CG measurements. The JRip found the following classification rule for the given dataset: "if CGH<=14.4 then sex is female, otherwise sex is male". The accuracy of this rule is not high, but it gives an idea about the relationship between CG measurements and sex. Although the issue that CG morphometric measurements can be used in sex determination is still controversial, it was concluded in the analysis that CG morphometric measurements can be used in sex determination. Also, Machine Learning Techniques give an idea about the relationship between CG measurements and sex.

En el estudio se propuso examinar la morfometría y la morfología de la crista galli del hueso etmoides usando tomografía computarizada de haz cónico (CBCT) y aplicar un nuevo análisis, técnicas de aprendizaje automático supervisado para encontrar las respuestas a las preguntas de investigación "¿Se puede determinar el sexo con mediciones morfométricas de la crista galli?" o "¿Qué tan efectivas son las medidas morfométricas de la crista galli para determinar el sexo?". Las dimensiones de la crista galli, incluidas los diámetros anteroposterior, superoinferior y laterolateral, se midieron y realizaron en 200 sujetos adultos sanos (98 mujeres; 102 hombres) con edades comprendidas entre los 18 y los 79 años. La crista galli se clasificó con dos métodos llamados tipos morfológicos y clasificación de Keros. En este estudio, se realizaron la prueba de Chicuadrado, la prueba t de Student y ANOVA de una vía. Adicionalmente, se aplicaron técnicas de Machine Learning. Las medias de CGH, CGW y CGL se encontraron en 14,96 mm; 3,96 mm y 12,76 mm en hombres, respectivamente. Los mismos valores fueron 13,54 mm; 3,51 mm y 11,59 ± 1,61 mm en mujeres, respectivamente. Las medidas morfométricas del CG de los hombress fueron más altas que las de las mujeres. Hubo una diferencia significativa entre sexos en cuanto al tipo de clasificación morfológica. Además, cuando se analizó la asignación de sexo de JRip, de 102 instancias masculinas, 62 de ellas se predijeron correctamente, y de 98 instancias femeninas, 70 de ellas se predijeron correctamente de acuerdo con las mediciones de CG. El JRip encontró la siguiente regla de clasificación para el conjunto de datos dado: "si CGH<=14.4, por tanto el sexo es femenino, de lo contrario, el sexo es masculino". La precisión de esta regla no es alta, pero da una idea de la relación entre las medidas del CG y el sexo. Aunque la pregunta si las medidas morfométricas CG se pueden usar en la determinación del sexo sigue aún siendo controvertida. Se concluyó en el análisis que las medidas morfométricas CG se pueden usar en la determinación del sexo. Además, las técnicas de aprendizaje automático dan una idea de la relación entre las medidas de CG y el sexo.

Chat GPT: origen, evolución, retos e impactos en la educación / Chat GPT: Origen, Evolution, Challenges and Impacts on Education

Introducción: Los avances de unas tecnologías y la obsolescencia de otras marchan a una velocidad inimaginable, especialmente en este siglo xxi. En los últimos meses de 2022 y primeros meses de 2023 muchas incógnitas y controversias en diferentes campos han surgido en torno a los Chat GPS, una innovación que presenta desafíos nunca pensados para la sociedad actual, así como nuevos retos que impactarán de manera directa en la formación y/o desempeño de profesores, estudiantes, profesionales de la salud, juristas, políticos, informáticos, bibliotecarios, científicos y cualquier ciudadano. Objetivo: Identificar algunas características del chat GPT y su posible impacto en el educación. Posicionamiento de los autores: Se leen en las noticias y reportajes valoraciones de especialistas; se han realizado encuentros virtuales y exposiciones; y están disponibles diversos artículos y videos sobre este tema, algunos llegan a ser elaborados con el propio asistente. Por la novedad del tema, la reciente incorporación como herramienta para el desarrollo profesional, así como por el interés mostrado en los últimos días por la comunidad de profesores de las ciencias médicas cubanas, y considerando que esta herramienta es resultado del desarrollo de la inteligencia artificial, cabe preguntarse: ¿en qué consiste? y ¿cuáles son sus perspectivas? Conclusiones: Resulta oportuno acercarse al tema desde las posibilidades y los retos que abre a la educación y el aprendizaje, en particular a la docencia médica(AU)

Introduction: The advances of some technologies and the obsolescence of others are marching at an unimaginable speed, especially in this twenty-first century. In the last months of 2022 and first months of 2023, many questions and controversies in different fields have arisen with respect to Chat GPT, an innovation that presents challenges never thought of before for today's society, as well as new challenges that will have a direct impact on the training and/or performance of professors, students, health professionals, law practitioners, politicians, computer scientists, librarians, scientists and any citizen. Objective: To identify some technological characteristics of Chat GPT. Positioning of the authors: In news and reports, assessments by specialists are read; virtual meetings and presentations have been held; and several articles and videos on this topic are available, some of them even elaborated by the assistant itself. Due to the novelty of the subject, its recent assimilation as a tool for professional development, as well as the interest shown in recent days by the community of professors of Cuban medical sciences and considering that this tool is the result of the development of artificial intelligence, it is worth wondering what it consists in and what its prospects are. Conclusions: It is appropriate to approach the subject with a focus on the possibilities and challenges that it opens to education and learning (AU)

An interpretable machine learning-based prediction model for risk of death for patients with ischemic stroke in intensive care unit / 南方医科大学学报

OBJECTIVE@#To construct an inherent interpretability machine learning model as an explainable boosting machine model (EBM) for predicting one-year risk of death in patients with severe ischemic stroke.@*METHODS@#We randomly divided the data of 2369 eligible patients with severe ischemic stroke in the MIMIC-Ⅳ(2.0) database, who were admitted in ICU in 2008 to 2019, into a training dataset (80%) and a test dataset (20%), and assessed the prognosis of the patients using the EBM model. The prediction performance of the model was evaluated by calculating the area under the receiver operating characteristic (AUC) curve. The calibration curve and Brier score were used to evaluate the degree of calibration of the model, and a decision curve was generated to assess the net clinical benefit.@*RESULTS@#The EBM model constructed in this study had good discrimination power, calibration and net benefit, with an AUC of 0.857 (95% CI: 0.831-0.887) for predicting prognosis of severe ischemic stroke. Calibration curve analysis showed that the standard curve of the EBM model was the closest to the ideal curve. Decision curve analysis showed that the model had the greatest net benefit rate at the prediction probability threshold of 0.10 to 0.80. The top 5 independent predictive variables based on the EBM model were age, SOFA score, mean heart rate, mechanical ventilation, and mean respiratory rate, whose significance scores ranged from 0.179 to 0.370.@*CONCLUSION@#This EBM model has a good performance for predicting the risk of death within one year in patients with severe ischemic stroke and allows clinicians to better understand the contributing factors of the patients' outcomes through the model interpretability.

Efficacy of machine learning models versus Cox regression model for predicting prognosis of esophagogastric junction adenocarcinoma / 南方医科大学学报

OBJECTIVE@#To compare the performance of machine learning models and traditional Cox regression model in predicting postoperative outcomes of patients with esophagogastric junction adenocarcinoma (AEG).@*METHODS@#This study was conducted among 203 AEG patients with complete clinical and follow-up data, who were treated in our hospital between September, 2015 and October, 2020. The clinicopathological data of the patients were processed for analysis using R language package and divided into training and validation datasets at the ratio of 3:1. The Cox proportional hazards regression model and 4 machine learning models were constructed for analyzing the datasets. ROC curves, calibration curves and clinical decision curves (DCA) were plotted. Internal validation of the machine learning models was performed to assess their predictive efficacy. The predictive performance of each model was evaluated by calculating the area under the curve (AUC), and the model fitting was assessed using the calibration curve.@*RESULTS@#For predicting 3-year survival based on the validation dataset, the AUC was 0.870 for Cox proportional hazard regression model, 0.901 for eXtreme Gradient Boosting (XGBoost), 0.791 for random forest, 0.832 for support vector machine, and 0.725 for multilayer perceptron; For predicting 5-year survival, the AUCs of these models were 0.915, 0.916, 0.758, 0.905, and 0.737, respectively. For internal validation, the AUCs of the 4 machine learning models decreased in the order of XGBoost (0.818), random forest (0.758), support vector machine (0.0.804), and multilayer perceptron (0.745).@*CONCLUSION@#The machine learning models show better predictive efficacy for survival outcomes of patients with AEG than Cox proportional hazard regression model, especially when proportional odds assumption or linear regression models are not applicable. XGBoost models have better performance than the other machine learning models, and the multi-layer perception model may have poor fitting results for a limited data volume.

A review on cardiovascular disease risk prediction models in the elderly / 中华流行病学杂志

Risk prediction models play an important role in the primary prevention of cardiovascular diseases (CVD) in the elderly population. There are fifteen papers about CVD risk prediction models developed for the elderly domestically and internationally, of which the definitions of disease outcome vary widely. Ten models were reported with insufficient information about study methods or results. Ten models were at high risk of bias. Thirteen models presented moderate discrimination in internal validation, and only four models have undertaken external validation. The CVD risk prediction models for the elderly differed from those for the general population in terms of model algorithm and the effect size of association between predictor and outcome, and the prediction performance of the models for the elderly attenuated. In the future, high-quality external validation researches are necessary to provide more solid evidence. Different ways, including adding new predictors, using competing risk model algorithms, machine learning methods, or joint models, and altering the prediction time horizon, should be explored to optimize the current models.

Advances in machine learning for predicting protein functions / 生物工程学报

Proteins play a variety of functional roles in cellular activities and are indispensable for life. Understanding the functions of proteins is crucial in many fields such as medicine and drug development. In addition, the application of enzymes in green synthesis has been of great interest, but the high cost of obtaining specific functional enzymes as well as the variety of enzyme types and functions hamper their application. At present, the specific functions of proteins are mainly determined through tedious and time-consuming experimental characterization. With the rapid development of bioinformatics and sequencing technologies, the number of protein sequences that have been sequenced is much larger than those can be annotated, thus developing efficient methods for predicting protein functions becomes crucial. With the rapid development of computer technology, data-driven machine learning methods have become a promising solution to these challenges. This review provides an overview of protein function and its annotation methods as well as the development history and operation process of machine learning. In combination with the application of machine learning in the field of enzyme function prediction, we present an outlook on the future direction of efficient artificial intelligence-assisted protein function research.

Progression in the application of machine learning in acute respiratory distress syndrome / 中华危重病急救医学

Acute respiratory distress syndrome (ARDS) is a clinical syndrome defined by acute onset of hypoxemia and bilateral pulmonary opacities not fully explained by cardiac failure or volume overload. At present, there is no specific drug treatment for ARDS, and the mortality rate is high. The reason may be that ARDS has rapid onset, rapid progression, complex etiology, and great heterogeneity of clinical manifestations and treatment. Compared with traditional data analysis, machine learning algorithms can automatically analyze and obtain rules from complex data and interpret them to assist clinical decision making. This review aims to provide a brief overview of the machine learning progression in ARDS clinical phenotype, onset prediction, prognosis stratification, and interpretable machine learning in recent years, in order to provide reference for clinical.

Dissecting Psychiatric Heterogeneity and Comorbidity with Core Region-Based Machine Learning / 神经科学通报·英文版

Machine learning approaches are increasingly being applied to neuroimaging data from patients with psychiatric disorders to extract brain-based features for diagnosis and prognosis. The goal of this review is to discuss recent practices for evaluating machine learning applications to obsessive-compulsive and related disorders and to advance a novel strategy of building machine learning models based on a set of core brain regions for better performance, interpretability, and generalizability. Specifically, we argue that a core set of co-altered brain regions (namely 'core regions') comprising areas central to the underlying psychopathology enables the efficient construction of a predictive model to identify distinct symptom dimensions/clusters in individual patients. Hypothesis-driven and data-driven approaches are further introduced showing how core regions are identified from the entire brain. We demonstrate a broadly applicable roadmap for leveraging this core set-based strategy to accelerate the pursuit of neuroimaging-based markers for diagnosis and prognosis in a variety of psychiatric disorders.

Development and Application of Medical Imaging Analysis Platform Based on Radiomics and Machine Learning Technologies / 中国医疗器械杂志

OBJECTIVE@#In order to solve the technical problems, clinical researchers face the process of medical imaging analysis such as data labeling, feature extraction and algorithm selection, a medical imaging oriented multi-disease research platform based on radiomics and machine learning technology was designed and constructed.@*METHODS@#Five aspects including data acquisition, data management, data analysis, modeling and data management were considered. This platform provides comprehensive functions such as data retrieve and data annotation, image feature extraction and dimension reduction, machine learning model running, results validation, visual analysis and automatic generation of analysis reports, thus an integrated solution for the whole process of radiomics analysis has been generated.@*RESULTS@#Clinical researchers can use this platform for the whole process of radiomics and machine learning analysis for medical images, and quickly produce research results.@*CONCLUSIONS@#This platform greatly shortens the time for medical image analysis research, decreasing the work difficulty of clinical researchers, as well as significantly promoting their working efficiency.

Recent research on machine learning in the diagnosis and treatment of necrotizing enterocolitis in neonates / 中国当代儿科杂志

Necrotizing enterocolitis (NEC), with the main manifestations of bloody stool, abdominal distension, and vomiting, is one of the leading causes of death in neonates, and early identification and diagnosis are crucial for the prognosis of NEC. The emergence and development of machine learning has provided the potential for early, rapid, and accurate identification of this disease. This article summarizes the algorithms of machine learning recently used in NEC, analyzes the high-risk predictive factors revealed by these algorithms, evaluates the ability and characteristics of machine learning in the etiology, definition, and diagnosis of NEC, and discusses the challenges and prospects for the future application of machine learning in NEC.

A method for photoplethysmography signal quality assessment fusing multi-class features with multi-scale series information / 生物医学工程学杂志

Photoplethysmography (PPG) is often affected by interference, which could lead to incorrect judgment of physiological information. Therefore, performing a quality assessment before extracting physiological information is crucial. This paper proposed a new PPG signal quality assessment by fusing multi-class features with multi-scale series information to address the problems of traditional machine learning methods with low accuracy and deep learning methods requiring a large number of samples for training. The multi-class features were extracted to reduce the dependence on the number of samples, and the multi-scale series information was extracted by a multi-scale convolutional neural network and bidirectional long short-term memory to improve the accuracy. The proposed method obtained the highest accuracy of 94.21%. It showed the best performance in all sensitivity, specificity, precision, and F1-score metrics, compared with 6 quality assessment methods on 14 700 samples from 7 experiments. This paper provides a new method for quality assessment in small samples of PPG signals and quality information mining, which is expected to be used for accurate extraction and monitoring of clinical and daily PPG physiological information.

Advances in heart failure clinical research based on deep learning / 生物医学工程学杂志

Heart failure is a disease that seriously threatens human health and has become a global public health problem. Diagnostic and prognostic analysis of heart failure based on medical imaging and clinical data can reveal the progression of heart failure and reduce the risk of death of patients, which has important research value. The traditional analysis methods based on statistics and machine learning have some problems, such as insufficient model capability, poor accuracy due to prior dependence, and poor model adaptability. In recent years, with the development of artificial intelligence technology, deep learning has been gradually applied to clinical data analysis in the field of heart failure, showing a new perspective. This paper reviews the main progress, application methods and major achievements of deep learning in heart failure diagnosis, heart failure mortality and heart failure readmission, summarizes the existing problems and presents the prospects of related research to promote the clinical application of deep learning in heart failure clinical research.

Automatic sleep staging algorithm for stochastic depth residual networks based on transfer learning / 生物医学工程学杂志

The existing automatic sleep staging algorithms have the problems of too many model parameters and long training time, which in turn results in poor sleep staging efficiency. Using a single channel electroencephalogram (EEG) signal, this paper proposed an automatic sleep staging algorithm for stochastic depth residual networks based on transfer learning (TL-SDResNet). Firstly, a total of 30 single-channel (Fpz-Cz) EEG signals from 16 individuals were selected, and after preserving the effective sleep segments, the raw EEG signals were pre-processed using Butterworth filter and continuous wavelet transform to obtain two-dimensional images containing its time-frequency joint features as the input data for the staging model. Then, a ResNet50 pre-trained model trained on a publicly available dataset, the sleep database extension stored in European data format (Sleep-EDFx) was constructed, using a stochastic depth strategy and modifying the output layer to optimize the model structure. Finally, transfer learning was applied to the human sleep process throughout the night. The algorithm in this paper achieved a model staging accuracy of 87.95% after conducting several experiments. Experiments show that TL-SDResNet50 can accomplish fast training of a small amount of EEG data, and the overall effect is better than other staging algorithms and classical algorithms in recent years, which has certain practical value.

Machine learning model predicts the occurrence of acute kidney injury after open surgery for abdominal aortic aneurysm repair / 中南大学学报(医学版)

OBJECTIVES@#Abdominal aortic aneurysm is a pathological condition in which the abdominal aorta is dilated beyond 3.0 cm. The surgical options include open surgical repair (OSR) and endovascular aneurysm repair (EVAR). Prediction of acute kidney injury (AKI) after OSR is helpful for decision-making during the postoperative phase. To find a more efficient method for making a prediction, this study aims to perform tests on the efficacy of different machine learning models.@*METHODS@#Perioperative data of 80 OSR patients were retrospectively collected from January 2009 to December 2021 at Xiangya Hospital, Central South University. The vascular surgeon performed the surgical operation. Four commonly used machine learning classification models (logistic regression, linear kernel support vector machine, Gaussian kernel support vector machine, and random forest) were chosen to predict AKI. The efficacy of the models was validated by five-fold cross-validation.@*RESULTS@#AKI was identified in 33 patients. Five-fold cross-validation showed that among the 4 classification models, random forest was the most precise model for predicting AKI, with an area under the curve of 0.90±0.12.@*CONCLUSIONS@#Machine learning models can precisely predict AKI during early stages after surgery, which allows vascular surgeons to address complications earlier and may help improve the clinical outcomes of OSR.

Using machine learning algorithm to predict the risk of post-traumatic stress disorder among firefighters in Changsha / 中南大学学报(医学版)

OBJECTIVES@#Firefighters are prone to suffer from psychological trauma and post-traumatic stress disorder (PTSD) in the workplace, and have a poor prognosis after PTSD. Reliable models for predicting PTSD allow for effective identification and intervention for patients with early PTSD. By collecting the psychological traits, psychological states and work situations of firefighters, this study aims to develop a machine learning algorithm with the aim of effectively and accurately identifying the onset of PTSD in firefighters, as well as detecting some important predictors of PTSD onset.@*METHODS@#This study conducted a cross-sectional survey through convenient sampling of firefighters from 20 fire brigades in Changsha, which were evenly distributed across 6 districts and Changsha County, with a total of 628 firefighters. We used the synthetic minority oversampling technique (SMOTE) to process data sets and used grid search to finish the parameter tuning. The predictive capability of several commonly used machine learning models was compared by 5-fold cross-validation and using the area under the receiver operating characteristic curve (ROC-AUC), accuracy, precision, recall, and F1 score.@*RESULTS@#The random forest model achieved good performance in predicting PTSD with an average AUC score at 0.790. The mean accuracy of the model was 90.1%, with an F1 score of 0.945. The three most important predictors were perseverance, forced thinking, and reflective deep thinking, with weights of 0.165, 0.158, and 0.152, respectively. The next most important predictors were employment time, psychological power, and optimism.@*CONCLUSIONS@#PTSD onset prediction model for Changsha firefighters constructed by random forest has strong predictive ability, and both psychological characteristics and work situation can be used as predictors of PTSD onset risk for firefighters. In the next step of the study, validation using other large datasets is needed to ensure that the predictive models can be used in clinical setting.