Statistical methods for clustered competing risk data when the event types are only available in a training dataset.

Background: Clinical data play an important role in medical sector for binary outcome variables. Various methods can be applied to build predictive models for the clinical data with binary outcome variables.Objective: This research was aimed to explore and compare the process of constructing common predictive models.Methodology: Models based on an artificial neural network (the connectionist approach) and binary logistic regressions were compared in their ability to classifying malnourished subjects and those with over-weighted participants in rural areas of Bangladesh. Subjects were classified according to the indicator of nutritional status measured by body mass index (BMI). This study also investigated the effects of different factors on the BMI level of adults of six Villages in Bangladesh. Demographic, anthropometric and clinical data were collected based on aged over 30 years from six Villages in Bangladesh that were identified as mainly dependent on wells contaminated with arsenic.Result: A total of 460 participants were recruited for this study. Out of 460(140 male and 320 females) participants 186(40.44%) were identified as malnourished (BMK18.5 gm), and the remainder 274(59.56%) were found as over-weighted (BMI>18.5 gm). Among other factors, arsenic exposures were found as significant risk factors for low body mass index (BMI) with a higher value of odds ratio. This study shows that, binary logistic regression correctly classified 72.85% of cases with malnourished in the training datasets, 76.08% in the testing datasets and 75.26% of all subjects. The sensitivities of the neural network architecture for the training and testing datasets and for all subjects were 84.28%, 84.78% and 81 .72% respectively, indicate better performance than binary logistic regression model.Conclusion: This study demonstrates a significant performance of artificial neural network than the binary logistic regression models in classification of malnourished participants from over-weighted ones.J Shaheed Suhrawardy Med Coll, 2014; 6(2):71-75

The process of extracting meaningful rules from big and complex data is called data mining. Data mining has an increasing popularity in every field today. Data units are established in customer-oriented industries such as marketing, finance and telecommunication to work on the customer churn and acquisition, in particular. Among the data mining methods, classification algorithms are used in studies conducted for customer acquisition to predict the potential customers of the company in question in the related industry. In this study, bank marketing data set in UCI Machine Learning Data Set was used by creating models with the same classification algorithms in different data mining programs. Accuracy, precision and f- measure criteria were used to test performances of the classification models. When creating the classification models, the test and training data sets were randomly divided by the holdout method to evaluate the performance of the data set. The data set was divided into training and test data sets with the 60-40%, 75­25% and 80-20% separation ratios. Data mining programs used for these processes are the R, Knime, RapidMiner and WEKA. And, classification algorithms commonly used in these platforms are the k-nearest neighbor (k-nn), Naive Bayes, and C4.5 decision tree.

In breast cancer diagnosis and treatment, non-invasive prediction of axillary lymph node (ALN) metastasis can help avoid complications related to sentinel lymph node biopsy. This study aims to develop and evaluate machine learning models using radiomics features extracted from diffusion-weighted whole-body imaging with background signal suppression (DWIBS) examination for predicting the ALN status. A total of 100 patients with histologically proven, invasive, clinically N0 breast cancer who underwent DWIBS examination consisting of short tau inversion recovery (STIR) and DWIBS sequences before surgery were enrolled. Radiomic features were calculated using segmented primary lesions in DWIBS and STIR sequences and were divided into training (n = 75) and test (n = 25) datasets based on the examination date. Using the training dataset, optimal feature selection was performed using the least absolute shrinkage and selection operator algorithm, and the logistic regression model and support vector machine (SVM) classifier model were constructed with DWIBS, STIR, or a combination of DWIBS and STIR sequences to predict ALN status. Receiver operating characteristic curves were used to assess the prediction performance of radiomics models. For the test dataset, the logistic regression model using DWIBS, STIR, and a combination of both sequences yielded an area under the curve (AUC) of 0.765 (95% confidence interval: 0.548-0.982), 0.801 (0.597-1.000), and 0.779 (0.567-0.992), respectively, whereas the SVM classifier model using DWIBS, STIR, and a combination of both sequences yielded an AUC of 0.765 (0.548-0.982), 0.757 (0.538-0.977), and 0.779 (0.567-0.992), respectively. Use of machine learning models incorporating with the quantitative radiomic features derived from the DWIBS and STIR sequences can potentially predict ALN status.

The pedestal-plus-increment stimulus paradigm I described at the Society's San Diego meeting was administered to subjects with known sensory hearing loss. These subjects, who had histories of exposure to high-intensity noise while in military service, showed both marked high-frequency loss and recruitment when tested by conventional audiometric methods. The averaged auditory evoked potential (vertex potential) was measured at a low frequency where the subjects' auditory thresholds were near normal, and at a high frequency where their thresholds were approximately 50 dB above normal. At both of the pedestal levels employed, the functions relating log (N1−P2) to ΔI (the magnitude of the 150-msec increment) differed markedly. At the lower frequency, the functions resembled those obtained from normal subjects; at the higher frequency, increments as small as 1 dB yielded N1−P2 potentials of approximately 10 μV. At this higher frequency, the vertex potential was near its maximum for a 1-dB change in stimulus level, and increases in ΔI up to 12 dB resulted in only slight increases in N1−P2. These findings suggest that this method may prove useful in the detection of recruitment of loudness.

To investigate the effect of changes within the spiral ligament and stria vascularis on hearing in cochlear otosclerosis, we examined spiral ligament hyalinization, stria vascularis atrophy, and sensory hearing loss in cochlear otosclerosis and described changes in ion transport molecule expression. Retrospective. Tertiary referral center. Thirty-two cochleae from 24 temporal bone donors with histologic evidence of cochlear otosclerosis, including spiral ligament hyalinization. Audiography. Measurements of spiral ligament width, stria vascularis, and bone-conduction thresholds were compared by the amount of hyalinization. Expression of the ion transport molecules Na,K-ATPase, connexin 26, and carbonic anhydrase II were assessed by immunohistochemical techniques. Hyalinization most often involved the posterior basal turn (88%) and the posterior middle turn (27%). Spiral ligament hyalinization correlated significantly with stria vascularis atrophy in the posterior middle turn of the cochlea (rho = -0.63, p < 0.01). There was a trend toward a significant association in the posterior basal turn (rho = -0.31, p < 0.08). Bone-conduction thresholds at 2,000 and 4,000 Hz were significantly associated with the amount of stria vascularis atrophy (rho = -0.44, -0.40, p < 0.05). In addition, we observed decreased immunostaining for both carbonic anhydrase II with Type I fibrocytes and Na,K-ATPase with stria vascularis and Type II and Type IV fibrocytes of the spiral ligament in cochlear otosclerosis sections compared with normal cochlea. Na,K-ATPase staining within the stria vascularis was further decreased in the presence of spiral ligament hyalinization. No significant differences were seen with connexin 26 immunostaining. However, immunostaining results were somewhat inconsistent. These data suggest that spiral ligament structure and function are essential for stria vascularis survival. In addition, dampened expression of ion transport molecules within the spiral ligament and stria vascularis may disrupt potassium ion recycling, resulting in loss of endocochlear potential and sensory hearing loss.

Currently, many hospitals use vital signs-based criteria such as modified early warning score (MEWS) and national early warning score (NEWS) to classify high-risk patients for cardiac arrest, but there are limitations in selecting high-risk patients with a possibility of cardiac arrest. The purpose of this study is to develop a cardiac arrest classification model to identify patients at high risk of cardiac arrest based on the patient family and past history, and blood test results after hospitalization, rather than vital signs. This study used electronic medical record (EMR) data from A university hospital, and patients in the high-risk group for cardiac arrest were defined as those who underwent cardio-pulmonary resuscitation (CPR) after cardiac arrest. Considering the use of the rapid response team of A university hospital, patients hospitalized in intensive care units (ICU), emergency medicine departments, psychiatric departments, pediatric departments, cardiology departments, and palliative care wards were excluded. This study included 325,534 patients, of which 3291 low-risk and 382 high-risk patients were selected for study. Data were split into training and validation data sets and univariate analysis was performed for 13 candidate risk factors. Then, multivariate analysis was performed using a bivariate logistic regression model, and an optimal model was selected using simulation analysis. In the training data set, it was calculated as sensitivity 75.25%, precision 21.59%, specificity 66.89%, accuracy 67.79%, F1 score 33.56, area under curve (AUC) 71.1 (95% confidence interval [CI] = 68.9–73.1 P value=<.001). In the validation data set, sensitivity 73.37%, precision 25.81%, specificity 75.03%, accuracy 74.86%, F1 score 38.19, AUC 74.2 (95% CI = 72.1–76.2, P value=<.001) were calculated. A model for classifying the high-risk group of cardiac arrest should be developed from various perspectives. In the future, in order to classify patients with high risk of cardiac arrest, a prospective study on the combined use of the model developed by this study and NEWS or MEWS should be conducted.

Large image collections generated from camera traps offer valuable insights into species richness, occupancy, and activity patterns, significantly aiding biodiversity monitoring. However, the manual processing of these data sets is time‐consuming, hindering analytical processes. To address this, deep neural networks have been widely adopted to automate image labelling, but the impact of classification error on key ecological metrics remains unclear. Here, we analyze data from camera trap collections in an African savannah (82,300 labelled images, 47 species) and an Asian sub‐tropical dry forest (40,308 labelled images, 29 species) to compare ecological metrics derived from expert‐generated species identifications with those generated by deep‐learning classification models. We specifically assess the impact of deep‐ learning model architecture, the proportion of label noise in the training data, and the size of the training data set on three key ecological metrics: species richness, occupancy, and activity patterns. We found that predictions of species richness derived from deep neural networks closely match those calculated from expert labels and remained resilient to up to 10% noise in the training data set (mis‐labelled images) and a 50% reduction in the training data set size. We found that our choice of deep‐learning model architecture (ResNet vs. ConvNext‐T) or depth (ResNet18, 50, 101) did not impact predicted ecological metrics. In contrast, species‐specific metrics were more sensitive; less common and visually similar species were disproportionately affected by a reduction in deep neural network accuracy, with consequences for occupancy and diel activity pattern estimates. To ensure the reliability of their findings, practitioners should prioritize creating large, clean training sets and account for class imbalance across species over exploring numerous deep‐learning model architectures.

A novel approach for accurate detection of the DDoS attacks in SDN-based SCADA systems based on deep recurrent neural networks

In the present era of the medical field, the mortality rate of the infant increased by 1/3 percentage in a year. Even though we have modernized and have great expertise in the medical domain, we failed to control the infantile mortality rate. So continuous monitoring of the fetus during pregnancy is important. If there is any complication in the growth of the fetus, then the patient is put into the appropriate examination and medication proposed by the physician. In this paper, we propose methods to control the infant mortality rate in the early stage of pregnancy. Fetal heart rate of 2126 patients was collected and developed as datasets. Then with these datasets, we developed machine learning classifier models to classify the normal, suspect, and pathologic cases using the Decision tree, Naive Bayes, Random forest, and K-nearest neighbors. We divided the datasets into training dataset and testing dataset. Then the base classifier model was created using training datasets, and then the same models were verified by appending the test datasets. We improved the techniques and efficacy of the base classifiers by ensemble methods such as bagging and boosting. Finally, the datasets were classified as normal, suspect, and pathological cases with an improved accuracy of 96.617% with the help of the random forest classifier.

Low nuclear grade ductal carcinoma in situ (DCIS) patients can adopt proactive management strategies to avoid unnecessary surgical resection. Different personalized treatment modalities may be selected based on the expression status of molecular markers, which is also predictive of different outcomes and risks of recurrence. DCIS ultrasound findings are mostly non mass lesions, making it difficult to determine boundaries. Currently, studies have shown that models based on deep learning radiomics (DLR) have advantages in automatic recognition of tumor contours. Machine learning models based on clinical imaging features can explain the importance of imaging features. The available ultrasound data of 349 patients with pure DCIS confirmed by surgical pathology [54 low nuclear grade, 175 positive estrogen receptor (ER+), 163 positive progesterone receptor (PR+), and 81 positive human epidermal growth factor receptor 2 (HER2+)] were collected. Radiologists extracted ultrasonographic features of DCIS lesions based on the 5th Edition of Breast Imaging Reporting and Data System (BI-RADS). Patient age and BI-RADS characteristics were used to construct clinical machine learning (CML) models. The RadImageNet pretrained network was used for extracting radiomics features and as an input for DLR modeling. For training and validation datasets, 80% and 20% of the data, respectively, were used. Logistic regression (LR), support vector machine (SVM), random forest (RF), and eXtreme Gradient Boosting (XGBoost) algorithms were performed and compared for the final classification modeling. Each task used the area under the receiver operating characteristic curve (AUC) to evaluate the effectiveness of DLR and CML models. In the training dataset, low nuclear grade, ER+, PR+, and HER2+ DCIS lesions accounted for 19.20%, 65.12%, 61.21%, and 30.19%, respectively; the validation set, they consisted of 19.30%, 62.50%, 57.14%, and 30.91%, respectively. In the DLR models we developed, the best AUC values for identifying features were 0.633 for identifying low nuclear grade, completed by the XGBoost Classifier of ResNet50; 0.618 for identifying ER, completed by the RF Classifier of InceptionV3; 0.755 for identifying PR, completed by the XGBoost Classifier of InceptionV3; and 0.713 for identifying HER2, completed by the LR Classifier of ResNet50. The CML models had better performance than DLR in predicting low nuclear grade, ER+, PR+, and HER2+ DCIS lesions. The best AUC values by classification were as follows: for low nuclear grade by RF classification, AUC: 0.719; for ER+ by XGBoost classification, AUC: 0.761; for PR+ by XGBoost classification, AUC: 0.780; and for HER2+ by RF classification, AUC: 0.723. Based on small-scale datasets, our study showed that the DLR models developed using RadImageNet pretrained network and CML models may help predict low nuclear grade, ER+, PR+, and HER2+ DCIS lesions so that patients benefit from hierarchical and personalized treatment.

In elderly individuals suffering from hip fractures, a prolonged hospital length of stay (PLOS) not only heightens the probability of patient complications but also amplifies mortality risks. Yet, most elderly hip fracture patients present compromised baseline health conditions. Additionally, PLOS leads to increased expenses for patient treatment and care, while also diminishing hospital turnover rates. This, in turn, jeopardizes the prompt allocation of beds for urgent cases. A retrospective study was carried out from October 2021 to November 2023 on 360 elderly hip fracture patients who underwent surgical treatment at West China Hospital. The 75th percentile of the total patient cohort's hospital stay duration, which was 12 days, was used to define prolonged hospital length of stay (PLOS). The cohort was divided into training and testing datasets with a 70:30 split. A predictive model was developed using the random forest algorithm, and its performance was validated and compared with the Lasso regression model. Out of 360 patients, 103 (28.61%) experienced PLOS. A Random Forest classification model was developed using the training dataset, identifying 10 essential variables. The Random Forest model achieved perfect performance in the training set, with an area under the curve (AUC), balanced accuracy, Kappa value, and F1 score of 1.000. In the testing set, the model's performance was assessed with an AUC of 0.846, balanced accuracy of 0.7294, Kappa value of 0.4325, and F1 score of 0.6061. This study aims to develop a prognostic model for predicting delayed discharge in elderly patients with hip fractures, thereby improving the accuracy of predicting PLOS in this population. By utilizing machine learning models, clinicians can optimize the allocation of medical resources and devise effective rehabilitation strategies for geriatric hip fracture patients. Additionally, this method can potentially improve hospital bed turnover rates, providing latent benefits for the healthcare system.

In low- and middle-income countries, cervical cancer remains a leading cause of death and morbidity for women. Early detection and treatment of precancerous lesions are critical in cervical cancer prevention, and colposcopy is a primary diagnostic tool for identifying cervical lesions and guiding biopsies. The transformation zone (TZ) is where a stratified squamous epithelium develops from the metaplasia of simple columnar epithelium and is the most common site of precancerous lesions. However, inexperienced colposcopists may find it challenging to accurately identify the type and location of the TZ during a colposcopy examination. This study aims to present an artificial intelligence (AI) method for identifying the TZ to enhance colposcopy examination and evaluate its potential clinical application. The study retrospectively collected data from 3616 women who underwent colposcopy at 6 tertiary hospitals in China between 2019 and 2021. A dataset from 4 hospitals was collected for model conduction. An independent dataset was collected from the other 2 geographic hospitals to validate model performance. There is no overlap between the training and validation datasets. Anonymized digital records, including each colposcopy image, baseline clinical characteristics, colposcopic findings, and pathological outcomes, were collected. The classification model was proposed as a lightweight neural network with multiscale feature enhancement capabilities and designed to classify the 3 types of TZ. The pretrained FastSAM model was first implemented to identify the location of the new squamocolumnar junction for segmenting the TZ. Overall accuracy, average precision, and recall were evaluated for the classification and segmentation models. The classification performance on the external validation was assessed by sensitivity and specificity. The optimal TZ classification model performed with 83.97% classification accuracy on the test set, which achieved average precision of 91.84%, 89.06%, and 95.62% for types 1, 2, and 3, respectively. The recall and mean average precision of the TZ segmentation model were 0.78 and 0.75, respectively. The proposed model demonstrated outstanding performance in predicting 3 types of the TZ, achieving the sensitivity with 95% CIs for TZ1, TZ2, and TZ3 of 0.78 (0.74-0.81), 0.81 (0.78-0.82), and 0.8 (0.74-0.87), respectively, with specificity with 95% CIs of 0.94 (0.92-0.96), 0.83 (0.81-0.86), and 0.91 (0.89-0.92), based on a comprehensive external dataset of 1335 cases from 2 of the 6 hospitals. Our proposed AI-based identification system classified the type of cervical TZs and delineated their location on multicenter, colposcopic, high-resolution images. The findings of this study have shown its potential to predict TZ types and specific regions accurately. It was developed as a valuable assistant to encourage precise colposcopic examination in clinical practice.

Permanent congenital hearing loss in children does not only influence speech acquisition but also has a severe impact on the entire intellectual and psychological development, and thus social integration is handicapped. However, at least partial compensation can be achieved by complex rehabilitative measures. One of the first medical therapies in children with congenital hearing loss is to equip them with hearing aids. As part of quality management the purpose of this study is to discuss pedaudiological aspects and results of hearing - impaired children equipped with hearing aids against the background of guidelines from Germany and the US. Data of 107 children equipped with hearing aids from 1997 to 1999 were analysed referring to the relevant German and US guideline. In 53 girls and 54 boys, 45% were suffering from at least severe, 45% from moderate and 10% from mild hearing loss. 93 children presented sensory hearing loss, 8 combined and 2 conductive hearing loss. 2 children had unilateral hearing loss, one a combined on one ear and a sensory hearing loss on the other and one child had a central auditory processing disorder. The age when the hearing loss was diagnosed depended on its grade and correlated with adverse environmental aspects. 34 children suffered from a syndrome or multiple handicaps, in these cases, the hearing loss was diagnosed at a later date than in children suffering from hearing loss only. In cases with near deafness audiological results with hearing aids were not sufficient; in these cases, cochlea implantation had to be discussed as an alternative measure. The data obtained from this study underline the need for universal hearing screening. Discussing the data against the relevant German and US guidelines clearly shows that German benchmarks are not as strict as the American ones. This has to be taken into account for further guideline development in Germany. As permanent hearing loss in children with syndromes and multiple handicaps and children with adverse environment is diagnosed at a later date than in other children public, measures have to be taken to improve this situation.

Objectives/HypothesesChildren with unilateral sensory hearing loss (UHL) struggle to understand speech in noise and locate the origin of sound and have reduced quality of hearing. This clinical trial will determine the benefits of cochlear implantation in children with UHL.Study DesignProspective clinical trial.MethodsTwenty children with at least moderate to profound sensory hearing loss and poor speech perception (word score <30%) in one ear and normal hearing in the contralateral ear participated in a Food and Drug Administration‐approved clinical trial. Subjects were evaluated for speech perception in quiet, speech perception in noise, sound localization, and subjective benefits after implantation.ResultsCNC word score perception in quiet significantly improved (1% to 50%, P < .0001) by 12 months after activation. Speech perception in noise by BKB‐SIN significantly improved in all three noise configurations; there was a 3.6 dB advantage in head shadow (P < .0001), a 1.6 dB advantage in summation (P = .003), and a 2.5 dB advantage in squelch (P = .0001). Localization improved by 26° at 9 months (P < .0001). Speech, Spatial, and Qualities (SSQ) demonstrated significant improvements in speech (5.2 to 7.4, P = .0012), qualities of hearing (5.9 to 7.5, P = .0056), and spatial hearing (2.7 to 6.6, P < .0001). SSQ subscales associated with binaural hearing were significantly improved, as was listening effort (P = .0082). Subjects demonstrated a non‐significant improvement in fatigue.ConclusionsThis study demonstrates that children with UHL significantly benefit from cochlear implantation.Level of EvidenceLevel 3 Laryngoscope, 132:S1–S18, 2022

ObjectiveThis study aims to develop and evaluate a radiomics-based machine learning model using T1-enhanced magnetic resonance imaging (MRI) features to differentiate between lung squamous cell carcinoma (SCC) and adenocarcinoma (AC) in patients with brain metastases (BMs). While prior studies have largely focused on primary lung tumors, our work uniquely targets metastatic brain lesions, which pose distinct diagnostic and therapeutic challenges.MethodsIn this retrospective study, 173 patients with BMs from lung cancer were included, consisting of 88 with AC and 85 with SCC. MRI images were acquired using a standardized protocol, and 833 radiomic features were identified from the segmented lesions utilizing the PyRadiomics package. Feature selection was performed using a combination of univariate analysis, correlation analysis, and the least absolute shrinkage and selection operator (LASSO) regression. Ten machine learning classifiers were trained and validated utilizing the selected features. The performance of the classifier models was assessed through receiver operating characteristic (ROC) curves, and the area under the curve (AUC) was examined for analysis.ResultsTen classifier models were built on the basis of features derived from MRI. Among the ten classifier models, the LightGBM model performed the best. In the training dataset, the LightGBM classifier achieved an accuracy of 0.814, with a sensitivity of 0.726 and specificity of 0.896. The classifier’s efficiency was validated on an independent testing dataset, where it maintained an accuracy of 0.779, with a sensitivity of 0.725 and specificity of 0.857. The AUC was 0.858 for the training dataset and 0.857 for the testing dataset. The model effectively distinguished between SCC and AC based on radiomic features, highlighting its potential for noninvasive non-small cell lung cancer (NSCLC) subtype classification.ConclusionThis research demonstrates the efficacy of a radiomics-based machine learning model in accurately classifying NSCLC subtypes from BMs, providing a valuable noninvasive tool for guiding personalized treatment strategies. Further validation on larger, multi-center datasets is crucial to verify these findings.