Multi level incremental influence measure based classification of medical data for improved classification

Ananthajothi, K; Subramaniam, M.

doi:10.1007/s10586-018-2498-z

Cited by 12 publications

(4 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, the use of deep learning should also be considered for future studies. On the other hand, other research to improve the performance of disease predictions uses the development of a classification algorithm based on a multi-level iterative influence measure that would be interesting to use in future research [ 41 ]. Another limitation we had is that it was not possible to access patient demographic data.…”

Section: Discussionmentioning

confidence: 99%

Classification Models of Action Research Arm Test Activities in Post-Stroke Patients Based on Human Hand Motion

Padilla-Magaña

Pitarch

2022

Sensors

View full text Add to dashboard Cite

The Action Research Arm Test (ARAT) presents a ceiling effect that prevents the detection of improvements produced with rehabilitation treatments in stroke patients with mild finger joint impairments. The aim of this study was to develop classification models to predict whether activities with similar ARAT scores were performed by a healthy subject or by a subject post-stroke using the extension and flexion angles of 11 finger joints as features. For this purpose, we used three algorithms: Support Vector Machine (SVM), Random Forest (RF), and K-Nearest Neighbors (KNN). The dataset presented class imbalance, and the classification models presented a low recall, especially in the stroke class. Therefore, we implemented class balance using Borderline-SMOTE. After data balancing the classification models showed significantly higher accuracy, recall, f1-score, and AUC. However, after data balancing, the SVM classifier showed a higher performance with a precision of 98%, a recall of 97.5%, and an AUC of 0.996. The results showed that classification models based on human hand motion features in combination with the oversampling algorithm Borderline-SMOTE achieve higher performance. Furthermore, our study suggests that there are differences in ARAT activities performed between healthy and post-stroke individuals that are not detected by the ARAT scoring process.

show abstract

Section: Discussionmentioning

confidence: 99%

Classification Models of Action Research Arm Test Activities in Post-Stroke Patients Based on Human Hand Motion

Padilla-Magaña

Pitarch

2022

Sensors

View full text Add to dashboard Cite

show abstract

“…The inherent data level that hinders the effectiveness of machine learning, statistical, and other methods [16,17], as well as conventional techniques, and causes feature analysis problems to provide inaccurate predictions [18,19], is questioned by this lack of data. In the study [20,21], the author described a medical data classification method that calculates the impact measure on several levels to determine the target class. The approach is built on various attributes and repeated influence measures.…”

Section: Literature Reviewmentioning

confidence: 99%

Fuzzy Rule Based- Model for Proficient Heart Disease Prediction in Data Lake

Mani,

Munusamy

2023

RIA

View full text Add to dashboard Cite

Cardiovascular diseases remain the leading cause of mortality on a global scale. Currently, healthcare organizations are shifting their operational strategies to enhance efficiency and reduce costs. For this transition, the integration of analytics into IT strategy is imperative. Data lakes, which aggregate data from multiple sources and employ analytical models, provide an innovative approach to information management, reporting, and predictive analytics. These models enable the production of advanced analytical insights, the implementation of evidence-based care plans, and the improvement of patient engagement outcomes, thus setting the stage for IoT-based prognostic systems aimed at reducing mortality rates. The present research proposes a comprehensive data analysis for the prognosis of coronary heart disease, a task that poses considerable challenges due to the volume of data across various disciplines and the complexity involved in analyzing, extracting, managing, and configuring data with massive data technologies and tools. To tackle this challenge, a multi-level fuzzy rule generation is suggested for identifying the features used in heart disease prediction. These features are then trained using an optimized recurrent neural network. The features are classified into labeled classes based on the risk assessment of a medical professional, enabling the prediction of the class based on risk. Early diagnosis and treatment are thus facilitated. When benchmarked against traditional systems, the proposed approach demonstrates superior performance, validating its potential for efficient heart disease prediction.

show abstract

“…While the use of effective regularization methods yielded promising results, no unsupervised learning was required. When faced with data shortages and increased dimensionality, unsupervised learning is considered important in the field of deep learning [19]. This represents the first efficient implementation of a volumetric CNN-based framework on MRI data employing 3D-stacked Convolutional Autoencoders for Alzheimer's disease classifications, but the prototype could only be emulated with an exactness of 80%.…”

Section: Machinelearning and Deep Learning Models For Ad Predictionmentioning

confidence: 99%

An Intensive Learning of Deep Learning in Alzheimer's Disease Prediction

Selvi,

Karthick

2023

Recent Trends in Data Science and Its Applications

View full text Add to dashboard Cite

The senior population is disproportionately affected by Alzheimer's disease (AD), making it the most common form of neurodegenerative disorders. It has an impact on patients' life, causing cognitive capacities such as consciousness, speech, actions, and problem-solving to deteriorate with time.Unfortunately, there is currently no cure for AD, and progress has been slow on finding a cure.However, the focus was on use of current medical information to diagnose the cognitive status of a patient, which in fact shows that a computer can replicate the clinical decision-making process of a physician. These rapid improvements have made it possible to collect enormous amounts of multifunctional neuroimaging data, which has led to the recent deployment of deep learning to the prompt identification and classification of Alzheimer's disease. Computational intelligence study into Alzheimer's disease continues to be in its infancy, but it is expanding rapidly as even more hybrid information sources are being incorporated, and as transparency is increased through the use of explicit strategies that integrate knowledge of specific behaviours and paths associated with the disease. This finding can be further understood by comparing multiple data-driven methodologies based on large-scale organization healthcare information from AD risk stratification, which may one day lead to improved selection of people at risk for AD in clinical laboratories or early identification of AD in clinical studies.

show abstract

Multi level incremental influence measure based classification of medical data for improved classification

Cited by 12 publications

References 10 publications

Classification Models of Action Research Arm Test Activities in Post-Stroke Patients Based on Human Hand Motion

Classification Models of Action Research Arm Test Activities in Post-Stroke Patients Based on Human Hand Motion

Fuzzy Rule Based- Model for Proficient Heart Disease Prediction in Data Lake

An Intensive Learning of Deep Learning in Alzheimer's Disease Prediction

Contact Info

Product

Resources

About