Deep Learning and Machine Learning with Grid Search to Predict Later Occurrence of Breast Cancer Metastasis Using Clinical Data

Association between sleep duration, depression and breast cancer in the United States: a national health and nutrition examination survey analysis 2009-2018

OBJECTIVE

Breast cancer is the most common cancer in women, threatening both physical and mental health. The epidemiological evidence for association between sleep duration, depression and breast cancer is inconsistent. The aim of this study was to determine the association between them and build machine-learning algorithms to predict breast cancer.

METHODS

A total of 1,789 participants from the National Health and Nutrition Examination Survey (NHANES) were included in the study, and 263 breast cancer patients were identified. Sleep duration was collected using a standardized questionnaire, and the Nine-item Patient Health Questionnaire (PHQ-9) was used to assess depression. Logistic regression yielded multivariable-adjusted breast cancer odds ratios (OR) and 95% confidence intervals (CI) for sleep duration and depression. Then, six machine learning algorithms, including AdaBoost, random forest, Boost tree, artificial neural network, limit gradient enhancement and support vector machine, were used to predict the development of breast cancer and find out the best algorithm.

RESULTS

Body mass index (BMI), race and smoking were statistically different between breast cancer and non-breast cancer groups. Participants with depression were associated with breast cancer (OR = 1.99, 95%CI: 1.55-3.51). Compared with 7-9h of sleep, the ORs for <7 and >9 h of sleep were 1.25 (95% CI: 0.85-1.37) and 1.05 (95% CI: 0.95-1.15), respectively. The AdaBoost model outperformed other machine learning algorithms and predicted well for breast cancer, with an area under curve (AUC) of 0.84 (95%CI: 0.81-0.87).

CONCLUSIONS

No significant association was observed between sleep duration and breast cancer, and participants with depression were associated with an increased risk for breast cancer. This finding provides new clues into the relationship between breast cancer and depression and sleep duration, and provides potential evidence for subsequent studies of pathological mechanisms.

Prediction of teicoplanin plasma concentration in critically ill patients: a combination of machine learning and population pharmacokinetics

BACKGROUND

Teicoplanin has been widely used in patients with infections caused by Staphylococcus aureus, especially for critically ill patients. The pharmacokinetics (PK) of teicoplanin vary between individuals and within the same individual. We aim to establish a prediction model via a combination of machine learning and population PK (PPK) to support personalized medication decisions for critically ill patients.

METHODS

A retrospective study was performed incorporating 33 variables, including PPK parameters (clearance and volume of distribution). Multiple algorithms and Shapley additive explanations were employed for feature selection of variables to determine the strongest driving factors.

RESULTS

The performance of each algorithm with PPK parameters was superior to that without PPK parameters. The composition of support vector regression, categorical boosting and a backpropagation neural network (7:2:1) with the highest R2 (0.809) was determined as the final ensemble model. The model included 15 variables after feature selection, of which the predictive performance was superior to that of models considering all variables or using only PPK. The R2, mean absolute error, mean squared error, absolute accuracy (±5 mg/L) and relative accuracy (±30%) of external validation were 0.649, 3.913, 28.347, 76.12% and 76.12%, respectively.

CONCLUSIONS

Our study offers a non-invasive, fast and cost-effective prediction model of teicoplanin plasma concentration in critically ill patients. The model serves as a fundamental tool for clinicians to determine the effective plasma concentration range of teicoplanin and formulate individualized dosing regimens accordingly.

Joint use of population pharmacokinetics and machine learning for prediction of valproic acid plasma concentration in elderly epileptic patients

BACKGROUND

Valproic acid (VPA) is a commonly used broad-spectrum antiepileptic drug. For elderly epileptic patients, VPA plasma concentrations have a considerable variation. We aim to establish a prediction model via a combination of machine learning and population pharmacokinetics (PPK) for VPA plasma concentration.

METHODS

A retrospective study was performed incorporating 43 variables, including PPK parameters. Recursive Feature Elimination with Cross-Validation was used for feature selection. Multiple algorithms were employed for ensemble model, and the model was interpreted by Shapley Additive exPlanations.

RESULTS

The inclusion of PPK parameters significantly enhances the performance of individual algorithm model. The composition of categorical boosting, light gradient boosting machine, and random forest (7:2:1) with the highest R2 (0.74) was determined as the ensemble model. The model included 11 variables after feature selection, of which the predictive performance was comparable to the model that incorporated all variables.

CONCLUSIONS

Our model was specifically tailored for elderly epileptic patients, providing an efficient and cost-effective approach to predict VPA plasma concentration. The model combined classical PPK with machine learning, and underwent optimization through feature selection and algorithm integration. Our model can serve as a fundamental tool for clinicians in determining VPA plasma concentration and individualized dosing regimens accordingly.

Optimized machine learning model for predicting unplanned reoperation after rectal cancer anterior resection

BACKGROUND

Unplanned reoperation (URO) after surgery adversely affects the quality of life and prognosis of patients undergoing anterior resection for rectal cancer. This study aims to meet the urgent need for reliable predictive tools by developing an optimized machine learning model to estimate the risk of URO following anterior resection in rectal cancer patients.

METHODS

This retrospective study collected multidimensional data from patients who underwent anterior resection for rectal cancer at Tongji Hospital of Huazhong University of Science and Technology from January 2012 to December 2022. Feature selection was conducted using both least absolute shrinkage and selection operator (LASSO) regression and the Boruta algorithm. Multiple machine learning models were developed, with parameter optimization via grid search and cross-validation. Performance metrics included accuracy, specificity, sensitivity, and area under curve (AUC). The optimal model was interpreted using SHapley Additive exPlanations (SHAP), and an online platform was created for real-time risk prediction.

RESULTS

A total of 2384 patients who underwent anterior resection for rectal cancer were included in this study. Following rigorous selection, 14 variables were identified for constructing the machine learning model. The optimized model demonstrated high predictive accuracy, with the random forest (RF) model achieving the best overall performance. The model achieved an AUC of 0.889 and an accuracy of 0.842 on the test dataset. SHAP analysis revealed that the tumor location, previous abdominal surgery, and operative time were the most significant factors influencing the risk of URO.

CONCLUSION

This study developed an optimized machine learning-based online predictive system to assess the risk of URO after anterior resection in rectal cancer patients. Accessible at https://yangsu2023.shinyapps.io/UROrisk/, this system improves prediction accuracy and offers real-time risk assessment, providing a valuable tool that may support clinical decision-making and potentially improve the prognosis of rectal cancer patients.

Sugar detection in adulterated honey using hyper-spectral imaging with stacking generalization method

This paper develops a new hybrid, automated, and non-invasive approach by combining hyper-spectral imaging, Savitzky-Golay (SG) Filter, Principal Components Analysis (PCA), Machine Learning (ML) classifiers/regressors, and stacking generalization methods to detect sugar in honey. First, the 32 different sugar concentration levels in honey were predicted using various ML regressors. Second, the six ranges of sugar were classified using various classifiers. Third, the 11 types of honey and 100% sugar were classified using classifiers. The stacking model (STM) obtained R2: 0.999, RMSE: 0.493 ml (v/v), RPD: 40.2, a 10-fold average R2: 0.996 and RMSE: 1.27 ml (v/v) for predicting 32 sugar concentrations. The STM achieved a Matthews Correlation Coefficient (MCC) of 99.7% and a Kappa score of 99.7%, a 10-fold average MCC of 98.9% and a Kappa score of 98.9% for classifying the six sugar ranges and 12 categories of honey types and a sugar.

Predicting Reduction Mammaplasty Total Resection Weight With Machine Learning

BACKGROUND

Machine learning (ML) is a form of artificial intelligence that has been used to create better predictive models in medicine. Using ML algorithms, we sought to create a predictive model for breast resection weight based on anthropometric measurements.

METHODS

We analyzed 237 patients (474 individual breasts) who underwent reduction mammoplasty at our institution. Anthropometric variables included body surface area (BSA), body mass index, sternal notch-to-nipple (SN-N), and nipple-to-inframammary fold values. Four different ML algorithms (linear regression, ridge regression, support vector regression, and random forest regression) either including or excluding the Schnur Scale prediction for the same data were trained and tested on their ability to recognize the relationship between the anthropometric variables and total resection weights. Resection weight prediction accuracy for each model and the Schnur scale alone were evaluated based on using mean absolute error (MAE).

RESULTS

In our cohort, mean age was 40.36 years. Most patients (71.61%) were African American. Mean BSA was 2.0 m 2 , mean body mass index was 33.045 kg/m 2 , mean SN-N was 35.0 cm, and mean nipple-to-inframammary fold was 16.0 cm. Mean SN-N was found to have the greatest variable importance. All 4 models made resection weight predictions with MAE lower than that of the Schnur Scale alone in both the training and testing datasets. Overall, the random forest regression model without Schnur scale weight had the lowest MAE at 186.20.

CONCLUSION

Our ML resection weight prediction model represents an accurate and promising alternative to the Schnur Scale in the setting of reduction mammaplasty consultations.

Enhancing groundwater quality assessment in coastal area: A hybrid modeling approach

Monitoring of groundwater (GW) resources in coastal areas is vital for human needs, agriculture, ecosystems, securing water supply, biodiversity, and environmental sustainability. Although the utilization of water quality index (WQI) models has proven effective in monitoring GW resources, it has faced substantial criticism due to its inconsistent outcomes, prompting the need for more reliable assessment methods. Therefore, this study addressed this concern by employing the data-driven root mean squared (RMS) models to evaluate groundwater quality (GWQ) in the coastal Bhola district near the Bay of Bengal, Bangladesh. To enhance the reliability of the RMS-WQI model, the research incorporated the extreme gradient boosting (XGBoost) machine learning (ML) algorithm. For the assessment of GWQ, the study utilized eleven crucial indicators, including turbidity (TURB), electric conductivity (EC), pH, total dissolved solids (TDS), nitrate (NO3 -), ammonium (NH4 +), sodium (Na), potassium (K), magnesium (Mg), calcium (Ca), and iron (Fe). In terms of the GW indicators, concentration of K, Ca and Mg exceeded the guideline limit in the collected GW samples. The computed RMS-WQI scores ranged from 54.3 to 72.1, with an average of 65.2, categorizing all sampling sites' GWQ as "fair." In terms of model reliability, XGBoost demonstrated exceptional sensitivity (R2 = 0.97) in predicting GWQ accurately. Furthermore, the RMS-WQI model exhibited minimal uncertainty (<1 %) in predicting WQI scores. These findings implied the efficacy of the RMS-WQI model in accurately assessing GWQ in coastal areas, that would ultimately assist regional environmental managers and strategic planners for effective monitoring and sustainable management of coastal GW resources.

Analysis of the fatigue status of medical security personnel during the closed-loop period using multiple machine learning methods: a case study of the Beijing 2022 Olympic Winter Games

Using machine learning methods to analyze the fatigue status of medical security personnel and the factors influencing fatigue (such as BMI, gender, and wearing protective clothing working hours), with the goal of identifying the key factors contributing to fatigue. By validating the predicted outcomes, actionable and practical recommendations can be offered to enhance fatigue status, such as reducing wearing protective clothing working hours. A questionnaire was designed to assess the fatigue status of medical security personnel during the closed-loop period, aiming to capture information on fatigue experienced during work and disease recovery. The collected data was then preprocessed and used to determine the structural parameters for each machine learning algorithm. To evaluate the prediction performance of different models, the mean relative error (MRE) and goodness of fit (R2) between the true and predicted values were calculated. Furthermore, the importance rankings of various parameters in relation to fatigue status were determined using the RF feature importance analysis method. The fatigue status of medical security personnel during the closed-loop period was analyzed using multiple machine learning methods. The prediction performance of these methods was ranked from highest to lowest as follows: Gradient Boosting Regression (GBM) > Random Forest (RF) > Adaptive Boosting (AdaBoost) > K-Nearest Neighbors (KNN) > Support Vector Regression (SVR). Among these algorithms, four out of the five achieved good prediction results, with the GBM method performing the best. The five most critical parameters influencing fatigue status were identified as working hours in protective clothing, a customized symptom and disease score (CSDS), physical exercise, body mass index (BMI), and age, all of which had importance scores exceeding 0.06. Notably, working hours in protective clothing obtained the highest importance score of 0.54, making it the most critical factor impacting fatigue status. Fatigue is a prevalent and pressing issue among medical security personnel operating in closed-loop environments. In our investigation, we observed that the GBM method exhibited superior predictive performance in determining the fatigue status of medical security personnel during the closed-loop period, surpassing other machine learning techniques. Notably, our analysis identified several critical factors influencing the fatigue status of medical security personnel, including the duration of working hours in protective clothing, CSDS, and engagement in physical exercise. These findings shed light on the multifaceted nature of fatigue among healthcare workers and emphasize the importance of considering various contributing factors. To effectively alleviate fatigue, prudent management of working hours for security personnel, along with minimizing the duration of wearing protective clothing, proves to be promising strategies. Furthermore, promoting regular physical exercise among medical security personnel can significantly impact fatigue reduction. Additionally, the exploration of medication interventions and the adoption of innovative protective clothing options present potential avenues for mitigating fatigue. The insights derived from this study offer valuable guidance to management personnel involved in organizing large-scale events, enabling them to make informed decisions and implement targeted interventions to address fatigue among medical security personnel. In our upcoming research, we will further expand the fatigue dataset while considering higher precisionprediction algorithms, such as XGBoost model, ensemble model, etc., and explore their potential contributions to our research.

Prediction model for spinal cord injury in spinal tuberculosis patients using multiple machine learning algorithms: a multicentric study

Spinal cord injury (SCI) is a prevalent and serious complication among patients with spinal tuberculosis (STB) that can lead to motor and sensory impairment and potentially paraplegia. This research aims to identify factors associated with SCI in STB patients and to develop a clinically significant predictive model. Clinical data from STB patients at a single hospital were collected and divided into training and validation sets. Univariate analysis was employed to screen clinical indicators in the training set. Multiple machine learning (ML) algorithms were utilized to establish predictive models. Model performance was evaluated and compared using receiver operating characteristic (ROC) curves, area under the curve (AUC), calibration curve analysis, decision curve analysis (DCA), and precision-recall (PR) curves. The optimal model was determined, and a prospective cohort from two other hospitals served as a testing set to assess its accuracy. Model interpretation and variable importance ranking were conducted using the DALEX R package. The model was deployed on the web by using the Shiny app. Ten clinical characteristics were utilized for the model. The random forest (RF) model emerged as the optimal choice based on the AUC, PRs, calibration curve analysis, and DCA, achieving a test set AUC of 0.816. Additionally, MONO was identified as the primary predictor of SCI in STB patients through variable importance ranking. The RF predictive model provides an efficient and swift approach for predicting SCI in STB patients.

Single-cell analysis reveals landscape of endometrial cancer response to estrogen and identification of early diagnostic markers

BACKGROUND

The development of endometrial cancer (EC) is closely related to the abnormal activation of the estrogen signaling pathway. Effective diagnostic markers are important for the early detection and treatment of EC.

METHOD

We downloaded single-cell RNA sequencing (scRNA-seq) and spatial transcriptome (ST) data of EC from public databases. Enrichment scores were calculated for EC cell subpopulations using the "AddModuleScore" function and the AUCell package, respectively. Six predictive models were constructed, including logistic regression (LR), Gaussian naive Bayes (GaussianNB), k-nearest neighbor (KNN), support vector machine (SVM), extreme gradient boosting (XGB), and neural network (NK). Subsequently, receiver-operating characteristics with areas under the curves (AUCs) were used to assess the robustness of the predictive model.

RESULT

We classified EC cell coaggregation into six cell clusters, of which the epithelial, fibroblast and endothelial cell clusters had higher estrogen signaling pathway activity. We founded the epithelial cell subtype Epi cluster1, the fibroblast cell subtype Fib cluster3, and the endothelial cell subtype Endo cluster3 all showed early activation levels of estrogen response. Based on EC cell subtypes, estrogen-responsive early genes, and genes encoding Stage I and para-cancer differentially expressed proteins in EC patients, a total of 24 early diagnostic markers were identified. The AUCs values of all six classifiers were higher than 0.95, which indicates that the early diagnostic markers we screened have superior robustness across different classification algorithms.

CONCLUSION

Our study elucidates the potential biological mechanism of EC response to estrogen at single-cell resolution, which provides a new direction for early diagnosis of EC.

The impact of Bayesian optimization on feature selection

Feature selection is an indispensable step for the analysis of high-dimensional molecular data. Despite its importance, consensus is lacking on how to choose the most appropriate feature selection methods, especially when the performance of the feature selection methods itself depends on hyper-parameters. Bayesian optimization has demonstrated its advantages in automatically configuring the settings of hyper-parameters for various models. However, it remains unclear whether Bayesian optimization can benefit feature selection methods. In this research, we conducted extensive simulation studies to compare the performance of various feature selection methods, with a particular focus on the impact of Bayesian optimization on those where hyper-parameters tuning is needed. We further utilized the gene expression data obtained from the Alzheimer's Disease Neuroimaging Initiative to predict various brain imaging-related phenotypes, where various feature selection methods were employed to mine the data. We found through simulation studies that feature selection methods with hyper-parameters tuned using Bayesian optimization often yield better recall rates, and the analysis of transcriptomic data further revealed that Bayesian optimization-guided feature selection can improve the accuracy of disease risk prediction models. In conclusion, Bayesian optimization can facilitate feature selection methods when hyper-parameter tuning is needed and has the potential to substantially benefit downstream tasks.

Role of Optimization in RNA-Protein-Binding Prediction

RNA-binding proteins (RBPs) play an important role in regulating biological processes, such as gene regulation. Understanding their behaviors, for example, their binding site, can be helpful in understanding RBP-related diseases. Studies have focused on predicting RNA binding by means of machine learning algorithms including deep convolutional neural network models. One of the integral parts of modeling deep learning is achieving optimal hyperparameter tuning and minimizing a loss function using optimization algorithms. In this paper, we investigate the role of optimization in the RBP classification problem using the CLIP-Seq 21 dataset. Three optimization methods are employed on the RNA-protein binding CNN prediction model; namely, grid search, random search, and Bayesian optimizer. The empirical results show an AUC of 94.42%, 93.78%, 93.23% and 92.68% on the ELAVL1C, ELAVL1B, ELAVL1A, and HNRNPC datasets, respectively, and a mean AUC of 85.30 on 24 datasets. This paper's findings provide evidence on the role of optimizers in improving the performance of RNA-protein binding prediction.

Assessment of human health risk from potentially toxic elements and predicting groundwater contamination using machine learning approaches

The Rooppur Nuclear Power Plant (RNPP) at Ishwardi, Bangladesh is planning to go into operation within 2024 and therefore, adjacent areas of RNPP is gaining adequate attention from the scientific community for environmental monitoring purposes especially for water resources management. However, there is a substantial lack of literature as well as environmental datasets for earlier years since very little was done at the beginning of the RNPP's construction phase. Therefore, this study was conducted to assess the potential toxic elements (PTEs) contamination in the groundwater and its associated health risk for residents at the adjacent part of the RNPP during the year of 2014-2015. For the purposes of achieving the aim of the study, groundwater samples were collected seasonally (dry and wet season) from nine sampling sites and afterwards analyzed for water quality indicators such as temperature (Temp.), pH, electrical conductivity (EC), total dissolved solid (TDS), total hardness (TH) and for PTEs including Iron (Fe), Manganese (Mn), Copper (Cu), Lead (Pb), Chromium (Cr), Cadmium (Cd) and Arsenic (As). This study adopted the newly developed Root Mean Square water quality index (RMS-WQI) model to assess the scenario of contamination from PTEs in groundwater whereas the human health risk assessment model was utilized to quantify the risk of toxicity from PTEs. In most of the sampling sites, PTEs concentration was found higher during the wet season than the dry season and Fe, Mn, Cd and As exceeded the guideline limit for drinking water. The RMS score mostly classified the groundwater in terms of PTEs contamination into "Fair" condition. The non-carcinogenic risks (expressed as Hazard Index-HI) revealed that around 44% and 89% of samples for adults and 67% and 100% of samples for children exceeded the threshold limit set by USEPA (HI > 1) and possessed risks through the oral pathway during dry and wet season, respectively. Furthermore, the calculated cumulative HI score was found higher for children than the adults throughout the study period. In terms of carcinogenic risk (CR) from PTEs, the magnitude of risk decreased following the pattern of Cr > As > Cd. Although the current study is based on old dataset, the findings might serve as a baseline for monitoring purposes to reduce future hazardous impact from the power plant.

The Rapid Non-Destructive Differentiation of Different Varieties of Rice by Fluorescence Hyperspectral Technology Combined with Machine Learning

A rice classification method for the fast and non-destructive differentiation of different varieties is significant in research at present. In this study, fluorescence hyperspectral technology combined with machine learning techniques was used to distinguish five rice varieties by analyzing the fluorescence hyperspectral features of Thai jasmine rice and four rice varieties with a similar appearance to Thai jasmine rice in the wavelength range of 475-1000 nm. The fluorescence hyperspectral data were preprocessed by a first-order derivative (FD) to reduce the background and baseline drift effects of the rice samples. Then, a principal component analysis (PCA) and t-distributed stochastic neighborhood embedding (t-SNE) were used for feature reduction and 3D visualization display. A partial least squares discriminant analysis (PLS-DA), BP neural network (BP), and random forest (RF) were used to build the rice classification models. The RF classification model parameters were optimized using the gray wolf algorithm (GWO). The results show that FD-t-SNE-GWO-RF is the best model for rice classification, with accuracy values of 99.8% and 95.3% for the training and test sets, respectively. The fluorescence hyperspectral technique combined with machine learning is feasible for classifying rice varieties.

Data-driven modelling for assessing trophic status in marine ecosystems using machine learning approaches

Assessing eutrophication in coastal and transitional waters is of utmost importance, yet existing Trophic Status Index (TSI) models face challenges like multicollinearity, data redundancy, inappropriate aggregation methods, and complex classification schemes. To tackle these issues, we developed a novel tool that harnesses machine learning (ML) and artificial intelligence (AI), enhancing the reliability and accuracy of trophic status assessments. Our research introduces an improved data-driven methodology specifically tailored for transitional and coastal (TrC) waters, with a focus on Cork Harbour, Ireland, as a case study. Our innovative approach, named the Assessment Trophic Status Index (ATSI) model, comprises three main components: the selection of pertinent water quality indicators, the computation of ATSI scores, and the implementation of a new classification scheme. To optimize input data and minimize redundancy, we employed ML techniques, including advanced deep learning methods. Specifically, we developed a CHL prediction model utilizing ten algorithms, among which XGBoost demonstrated exceptional performance, showcasing minimal errors during both training (RMSE = 0.0, MSE = 0.0, MAE = 0.01) and testing (RMSE = 0.0, MSE = 0.0, MAE = 0.01) phases. Utilizing a novel linear rescaling interpolation function, we calculated ATSI scores and evaluated the model's sensitivity and efficiency across diverse application domains, employing metrics such as R2, the Nash-Sutcliffe efficiency (NSE), and the model efficiency factor (MEF). The results consistently revealed heightened sensitivity and efficiency across all application domains. Additionally, we introduced a brand new classification scheme for ranking the trophic status of transitional and coastal waters. To assess spatial sensitivity, we applied the ATSI model to four distinct waterbodies in Ireland, comparing trophic assessment outcomes with the Assessment of Trophic Status of Estuaries and Bays in Ireland (ATSEBI) System. Remarkably, significant disparities between the ATSI and ATSEBI System were evident in all domains, except for Mulroy Bay. Overall, our research significantly enhances the accuracy of trophic status assessments in marine ecosystems. The ATSI model, combined with cutting-edge ML techniques and our new classification scheme, represents a promising avenue for evaluating and monitoring trophic conditions in TrC waters. The study also demonstrated the effectiveness of ATSI in assessing trophic status across various waterbodies, including lakes, rivers, and more. These findings make substantial contributions to the field of marine ecosystem management and conservation.

An Improved Nested U-Net Network for Fluorescence In Situ Hybridization Cell Image Segmentation

Fluorescence in situ hybridization (FISH) is a powerful cytogenetic method used to precisely detect and localize nucleic acid sequences. This technique is proving to be an invaluable tool in medical diagnostics and has made significant contributions to biology and the life sciences. However, the number of cells is large and the nucleic acid sequences are disorganized in the FISH images taken using the microscope. Processing and analyzing images is a time-consuming and laborious task for researchers, as it can easily tire the human eyes and lead to errors in judgment. In recent years, deep learning has made significant progress in the field of medical imaging, especially the successful application of introducing the attention mechanism. The attention mechanism, as a key component of deep learning, improves the understanding and interpretation of medical images by giving different weights to different regions of the image, enabling the model to focus more on important features. To address the challenges in FISH image analysis, we combined medical imaging with deep learning to develop the SEAM-Unet++ automated cell contour segmentation algorithm with integrated attention mechanism. The significant advantage of this algorithm is that it improves the accuracy of cell contours in FISH images. Experiments have demonstrated that by introducing the attention mechanism, our method is able to segment cells that are adherent to each other more efficiently.

Parametric optimization and comparative study of machine learning and deep learning algorithms for breast cancer diagnosis

Breast Cancer is the leading form of cancer found in women and a major cause of increased mortality rates among them. However, manual diagnosis of the disease is time-consuming and often limited by the availability of screening systems. Thus, there is a pressing need for an automatic diagnosis system that can quickly detect cancer in its early stages. Data mining and machine learning techniques have emerged as valuable tools in developing such a system. In this study we investigated the performance of several machine learning models on the Wisconsin Breast Cancer (original) dataset with a particular emphasis on finding which models perform the best for breast cancer diagnosis. The study also explores the contrast between the proposed ANN methodology and conventional machine learning techniques. The comparison between the methods employed in the current study and those utilized in earlier research on the Wisconsin Breast Cancer dataset is also compared. The findings of this study are in line with those of previous studies which also highlighted the efficacy of SVM, Decision Tree, CART, ANN, and ELM ANN for breast cancer detection. Several classifiers achieved high accuracy, precision and F1 scores for benign and malignant tumours, respectively. It is also found that models with hyperparameter adjustment performed better than those without and boosting methods like as XGBoost, Adaboost, and Gradient Boost consistently performed well across benign and malignant tumours. The study emphasizes the significance of hyperparameter tuning and the efficacy of boosting algorithms in addressing the complexity and nonlinearity of data. Using the Wisconsin Breast Cancer (original) dataset, a detailed summary of the current status of research on breast cancer diagnosis is provided.

Optimizing age-related hearing risk predictions: an advanced machine learning integration with HHIE-S

OBJECTIVES

The elderly are disproportionately affected by age-related hearing loss (ARHL). Despite being a well-known tool for ARHL evaluation, the Hearing Handicap Inventory for the Elderly Screening version (HHIE-S) has only traditionally been used for direct screening using self-reported outcomes. This work uses a novel integration of machine learning approaches to improve the predicted accuracy of the HHIE-S tool for ARHL in older adults.

METHODS

We employed a dataset that was gathered between 2016 and 2018 and included 1,526 senior citizens from several Taipei City Hospital branches. 80% of the data were used for training (n = 1220) and 20% were used for testing (n = 356). XGBoost, Gradient Boosting, and LightGBM were among the machine learning models that were only used and assessed on the training set. In order to prevent data leakage and overfitting, the Light Gradient Boosting Machine (LGBM) model-which had the greatest AUC of 0.83 (95% CI 0.81-0.85)-was then only used on the holdout testing data.

RESULTS

On the testing set, the LGBM model showed a strong AUC of 0.82 (95% CI 0.79-0.86), far outperforming conventional techniques. Notably, several HHIE-S items and age were found to be significant characteristics. In contrast to traditional HHIE research, which concentrates on the psychological effects of hearing loss, this study combines cutting-edge machine learning techniques-specifically, the LGBM classifier-with the HHIE-S tool. The incorporation of SHAP values enhances the interpretability of the model's predictions and provides a more comprehensive comprehension of the significance of various aspects.

CONCLUSIONS

Our methodology highlights the great potential that arises from combining machine learning with validated hearing evaluation instruments such as the HHIE-S. Healthcare practitioners can anticipate ARHL more accurately thanks to this integration, which makes it easier to intervene quickly and precisely.

Using artificial intelligence algorithms to predict the overall survival of hemodialysis patients during the COVID-19 pandemic: A prospective cohort study

BACKGROUND

Hemodialysis (HD) patients are a vulnerable population at high risk for severe complications from COVID-19. The impact of partial COVID-19 vaccination on the survival of HD patients remains uncertain. This prospective cohort study was designed to use artificial intelligence algorithms to predict the survival impact of partial COVID-19 vaccination in HD patients.

METHODS

A cohort of 433 HD patients was used to develop machine-learning models based on a subset of clinical features assessed between July 1, 2021, and April 29, 2022. The patient cohort was randomly split into training (80%) and testing (20%) sets for model development and evaluation. Machine-learning models, including categorical boosting (CatBoost), light gradient boosting machines (LightGBM), RandomForest, and extreme gradient boosting models (XGBoost), were applied to evaluate their discriminative performance using the patient cohorts.

RESULTS

Among these models, LightGBM achieved the highest F1 score of 0.95, followed by CatBoost, RandomForest, and XGBoost, with area under the receiver operating characteristic curve values of 0.94 on the testing dataset. The SHapley Additive explanation summary plot derived from the XGBoost model indicated that key features such as age, albumin, and vaccination details had a significant impact on survival. Furthermore, the fully vaccinated group exhibited higher levels of anti-spike (S) receptor-binding domain antibodies.

CONCLUSION

This prospective cohort study involved using artificial intelligence algorithms to predict overall survival in HD patients during the COVID-19 pandemic. These predictive models assisted in identifying high-risk individuals and guiding vaccination strategies for HD patients, ultimately improving overall prognosis. Further research is warranted to validate and refine these predictive models in larger and more diverse populations of HD patients.

Marine waters assessment using improved water quality model incorporating machine learning approaches

In marine ecosystems, both living and non-living organisms depend on "good" water quality. It depends on a number of factors, and one of the most important is the quality of the water. The water quality index (WQI) model is widely used to assess water quality, but existing models have uncertainty issues. To address this, the authors introduced two new WQI models: the weight based weighted quadratic mean (WQM) and unweighted based root mean squared (RMS) models. These models were used to assess water quality in the Bay of Bengal, using seven water quality indicators including salinity (SAL), temperature (TEMP), pH, transparency (TRAN), dissolved oxygen (DOX), total oxidized nitrogen (TON), and molybdate reactive phosphorus (MRP). Both models ranked water quality between "good" and "fair" categories, with no significant difference between the weighted and unweighted models' results. The models showed considerable variation in the computed WQI scores, ranging from 68 to 88 with an average of 75 for WQM and 70 to 76 with an average of 72 for RMS. The models did not have any issues with sub-index or aggregation functions, and both had a high level of sensitivity (R2 = 1) in terms of the spatio-temporal resolution of waterbodies. The study demonstrated that both WQI approaches effectively assessed marine waters, reducing uncertainty and improving the accuracy of the WQI score.

PPG Signals-Based Blood-Pressure Estimation Using Grid Search in Hyperparameter Optimization of CNN-LSTM

Researchers commonly use continuous noninvasive blood-pressure measurement (cNIBP) based on photoplethysmography (PPG) signals to monitor blood pressure conveniently. However, the performance of the system still needs to be improved. Accuracy and precision in blood-pressure measurements are critical factors in diagnosing and managing patients' health conditions. Therefore, we propose a convolutional long short-term memory neural network (CNN-LSTM) with grid search ability, which provides a robust blood-pressure estimation system by extracting meaningful information from PPG signals and reducing the complexity of hyperparameter optimization in the proposed model. The multiparameter intelligent monitoring for intensive care III (MIMIC III) dataset obtained PPG and arterial-blood-pressure (ABP) signals. We obtained 75,226 signal segments, with 60,180 signals allocated for training data, 12,030 signals allocated for the validation set, and 15,045 signals allocated for the test data. During training, we applied five-fold cross-validation with a grid-search method to select the best model and determine the optimal hyperparameter settings. The optimized configuration of the CNN-LSTM layers consisted of five convolutional layers, one long short-term memory (LSTM) layer, and two fully connected layers for blood-pressure estimation. This study successfully achieved good accuracy in assessing both systolic blood pressure (SBP) and diastolic blood pressure (DBP) by calculating the standard deviation (SD) and the mean absolute error (MAE), resulting in values of 7.89 ± 3.79 and 5.34 ± 2.89 mmHg, respectively. The optimal configuration of the CNN-LSTM provided satisfactory performance according to the standards set by the British Hypertension Society (BHS), the Association for the Advancement of Medical Instrumentation (AAMI), and the Institute of Electrical and Electronics Engineers (IEEE) for blood-pressure monitoring devices.

Empirical Study of Overfitting in Deep Learning for Predicting Breast Cancer Metastasis

Overfitting may affect the accuracy of predicting future data because of weakened generalization. In this research, we used an electronic health records (EHR) dataset concerning breast cancer metastasis to study the overfitting of deep feedforward neural networks (FNNs) prediction models. We studied how each hyperparameter and some of the interesting pairs of hyperparameters were interacting to influence the model performance and overfitting. The 11 hyperparameters we studied were activate function, weight initializer, number of hidden layers, learning rate, momentum, decay, dropout rate, batch size, epochs, L1, and L2. Our results show that most of the single hyperparameters are either negatively or positively corrected with model prediction performance and overfitting. In particular, we found that overfitting overall tends to negatively correlate with learning rate, decay, batch size, and L2, but tends to positively correlate with momentum, epochs, and L1. According to our results, learning rate, decay, and batch size may have a more significant impact on both overfitting and prediction performance than most of the other hyperparameters, including L1, L2, and dropout rate, which were designed for minimizing overfitting. We also find some interesting interacting pairs of hyperparameters such as learning rate and momentum, learning rate and decay, and batch size and epochs.

Prediction of the risk of developing end-stage renal diseases in newly diagnosed type 2 diabetes mellitus using artificial intelligence algorithms

OBJECTIVES

Type 2 diabetes mellitus (T2DM) imposes a great burden on healthcare systems, and these patients experience higher long-term risks for developing end-stage renal disease (ESRD). Managing diabetic nephropathy becomes more challenging when kidney function starts declining. Therefore, developing predictive models for the risk of developing ESRD in newly diagnosed T2DM patients may be helpful in clinical settings.

METHODS

We established machine learning models constructed from a subset of clinical features collected from 53,477 newly diagnosed T2DM patients from January 2008 to December 2018 and then selected the best model. The cohort was divided, with 70% and 30% of patients randomly assigned to the training and testing sets, respectively.

RESULTS

The discriminative ability of our machine learning models, including logistic regression, extra tree classifier, random forest, gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), and light gradient boosting machine were evaluated across the cohort. XGBoost yielded the highest area under the receiver operating characteristic curve (AUC) of 0.953, followed by extra tree and GBDT, with AUC values of 0.952 and 0.938 on the testing dataset. The SHapley Additive explanation summary plot in the XGBoost model illustrated that the top five important features included baseline serum creatinine, mean serum creatine within 1 year before the diagnosis of T2DM, high-sensitivity C-reactive protein, spot urine protein-to-creatinine ratio and female gender.

CONCLUSIONS

Because our machine learning prediction models were based on routinely collected clinical features, they can be used as risk assessment tools for developing ESRD. By identifying high-risk patients, intervention strategies may be provided at an early stage.

The innovative model based on artificial intelligence algorithms to predict recurrence risk of patients with postoperative breast cancer

Purpose

This study aimed to develop a machine learning model to retrospectively study and predict the recurrence risk of breast cancer patients after surgery by extracting the clinicopathological features of tumors from unstructured clinical electronic health record (EHR) data.

Methods

This retrospective cohort included 1,841 breast cancer patients who underwent surgical treatment. To extract the principal features associated with recurrence risk, the clinical notes and histopathology reports of patients were collected and feature engineering was used. Predictive models were next conducted based on this important information. All algorithms were implemented using Python software. The accuracy of prediction models was further verified in the test cohort. The area under the curve (AUC), precision, recall, and F1 score were adopted to evaluate the performance of each model.

Results

A training cohort with 1,289 patients and a test cohort with 552 patients were recruited. From 2011 to 2019, a total of 1,841 textual reports were included. For the prediction of recurrence risk, both LSTM, XGBoost, and SVM had favorable accuracies of 0.89, 0.86, and 0.78. The AUC values of the micro-average ROC curve corresponding to LSTM, XGBoost, and SVM were 0.98 ± 0.01, 0.97 ± 0.03, and 0.92 ± 0.06. Especially the LSTM model achieved superior execution than other models. The accuracy, F1 score, macro-avg F1 score (0.87), and weighted-avg F1 score (0.89) of the LSTM model produced higher values. All P values were statistically significant. Patients in the high-risk group predicted by our model performed more resistant to DNA damage and microtubule targeting drugs than those in the intermediate-risk group. The predicted low-risk patients were not statistically significant compared with intermediate- or high-risk patients due to the small sample size (188 low-risk patients were predicted via our model, and only two of them were administered chemotherapy alone after surgery). The prognosis of patients predicted by our model was consistent with the actual follow-up records.

Conclusions

The constructed model accurately predicted the recurrence risk of breast cancer patients from EHR data and certainly evaluated the chemoresistance and prognosis of patients. Therefore, our model can help clinicians to formulate the individualized management of breast cancer patients.

Evaluation of machine learning algorithms for the prognosis of breast cancer from the Surveillance, Epidemiology, and End Results database

INTRODUCTION

Many researchers used machine learning (ML) to predict the prognosis of breast cancer (BC) patients and noticed that the ML model had good individualized prediction performance.

OBJECTIVE

The cohort study was intended to establish a reliable data analysis model by comparing the performance of 10 common ML algorithms and the the traditional American Joint Committee on Cancer (AJCC) stage, and used this model in Web application development to provide a good individualized prediction for others.

METHODS

This study included 63145 BC patients from the Surveillance, Epidemiology, and End Results database.

RESULTS

Through the performance of the 10 ML algorithms and 7th AJCC stage in the optimal test set, we found that in terms of 5-year overall survival, multivariate adaptive regression splines (MARS) had the highest area under the curve (AUC) value (0.831) and F1-score (0.608), and both sensitivity (0.737) and specificity (0.772) were relatively high. Besides, MARS showed a highest AUC value (0.831, 95%confidence interval: 0.820-0.842) in comparison to the other ML algorithms and 7th AJCC stage (all P < 0.05). MARS, the best performing model, was selected for web application development (https://w12251393.shinyapps.io/app2/).

CONCLUSIONS

The comparative study of multiple forecasting models utilizing a large data noted that MARS based model achieved a much better performance compared to other ML algorithms and 7th AJCC stage in individualized estimation of survival of BC patients, which was very likely to be the next step towards precision medicine.