ASHISHA G R
@karunya.edu
Assistant Professor, Biomedical Engineering
Karunya Institute of Technology and Sciences
RESEARCH INTERESTS
Real Time Embedded Systems, Biomedical Signal Processing, Wearable Devices
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
G. R. Ashisha, X. Anitha Mary, E. Grace Mary Kanaga, J. Andrew, and R. Jennifer Eunice
Springer Science and Business Media LLC
AbstractDiabetes mellitus is considered one of the main causes of death worldwide. If diabetes fails to be treated and diagnosed earlier, it can cause several other health problems, such as kidney disease, nerve disease, vision problems, and brain issues. Early detection of diabetes reduces healthcare costs and minimizes the chance of serious complications. In this work, we propose an e-diagnostic model for diabetes classification via a machine learning algorithm that can be executed on the Internet of Medical Things (IoMT). The study uses and analyses two benchmarking datasets, the PIMA Indian Diabetes Dataset (PIDD) and the Behavioral Risk Factor Surveillance System (BRFSS) diabetes dataset, to classify diabetes. The proposed model consists of the random oversampling method to balance the range of classes, the interquartile range technique-based outlier detection to eliminate outlier data, and the Boruta algorithm for selecting the optimal features from the datasets. The proposed approach considers ML algorithms such as random forest, gradient boosting models, light gradient boosting classifiers, and decision trees, as they are widely used classification algorithms for diabetes prediction. We evaluated all four ML algorithms via performance indicators such as accuracy, F1 score, recall, precision, and AUC-ROC. Comparative analysis of this model suggests that the random forest algorithm outperforms all the remaining classifiers, with the greatest accuracy of 92% on the BRFSS diabetes dataset and 94% accuracy on the PIDD dataset, which is greater than the 3% accuracy reported in existing research. This research is helpful for assisting diabetologists in developing accurate treatment regimens for patients who are diabetic.
Ashisha GR, Anitha Mary X, and Mahimai Raja J
Scalable Computing: Practice and Experience
Diabetes is a widespread chronic condition that impacts people all over the globe and requires a clear and timely diagnosis. Untreated diabetes leads to retinopathy, nephropathy, and damage to the nervous system. In this context, Machine Learning (ML) might be used to detect health problems early, diagnose them, and track their progress. Ensemble techniques are a promising approach that combines many classifiers to improve forecast accuracy and resilience. This study investigates the categorization of diabetes using an ensemble machine learning technique known as a voting classifier. Using a variety of classifiers, including Light Gradient Boosting Machine (LightGBM), Gradient Boost classifier (GBC), and Random Forest (RF). The predictions are aggregated using voting methods to get a final classification result. The research is carried out using two benchmarking datasets: the Pima Indian Diabetes Dataset (PIDD) and the German Dataset. The Boruta technique is used to choose the best attributes from the datasets, while the Random Over Sampling approach balances the range of classes and eliminates abnormal data using the interquartile range approach. The findings showed that the combination of the Boruta feature selection algorithm and ensemble Voting Classifier performed better for both PIDD and German datasets with an accuracy of 93% and 90% respectively. These algorithms are evaluated and the maximum accuracy is produced using the combination of the Boruta feature selection algorithm and ensemble Voting Classifier. This research helps medical professionals in the early prediction of diabetes, reducing physician’s time.
G. R. Ashisha, X. Anitha Mary, Subrata Chowdhury, C. Karthik, Tanupriya Choudhury, and Ketan Kotecha
Springer Nature Switzerland
G. R. Ashisha and X. Anitha Mary
Springer Nature Singapore
G. R. Ashisha, X. Anitha Mary, H. Mohamed Ashif, I. Karthikeyan, and J. Roshan
IEEE
The notable developments in healthcare sciences and biotechnology have led to a significant increase in information technology, for instance, clinical data and high efficiency genetic information, produced from intensive Electronic Medical data. The use of data mining and machine learning techniques are essential in the biomedical sciences to distinctly transform all accessible health records into useful knowledge. Early detection of diseases like diabetes is crucial as the growing number of people with diabetes is rising rapidly. According to an International Diabetes Federation report, the number of diabetic cases globally is expected to reach 642 million by 2040. This growing disease requires a lot of attention. Machine learning has quickly advanced, and many facets of medical health have benefited from its use. Due to their capacity for prediction, machine learning algorithms are presently significant in the healthcare industry. The major goal of this work is to design a model that can more accurately predict patients’ diabetes. The research work analyzed the hospital clinical examination data obtained from Sree Guru Hospital, Kanyakumari district, TamilNadu. To improve the accuracy of the system, imbalanced data in the real time dataset were eliminated using oversampling technique. Extra Trees classifier model, and Random Forest (RF) was used to predict diabetes. The result of this work shows that prediction of diabetes with the Extra Trees classifier had the highest accuracy of 98%. Finally, this work effectively determines the pervasiveness and detection of diabetes.
M Smirthy, M Dhanushree, and G R Ashisha
IEEE
Nowadays, stress has become a major cause of many diseases. According to the World Health organization, stress affects about 280 million people globally (i.e.) 3.8% of the population. Early identification of stress is important to prevent its negative impacts on people and is therefore an important step in the service of humanity and health care. These days, smart devices have become an important part of our lives and it achieved a huge amount of usage. This led to the query of whether smartphones and wearable sensors can identify and prevent stress. Although there are numerous works on mental stress detection that exist in controlled laboratory settings, the number of methods for stress detection in everyday life is inadequate. A variety of studies have been conducted to establish an association between stressful events and human responses using diverse psychological, physiological, physical, and behavioral measures. This paper conducts an integrated review of human stress detection and it provides an analysis of strategies and techniques that have been proposed for detecting human stress. This study also focuses on the prospective strategies, advantages, and challenges of mental stress detection using machine learning (ML), and wearable sensors.
G. R. Ashisha and X. Anitha Mary
Springer Singapore
G. R. Ashisha, X. Anitha Mary, and Lina Rose
Informa UK Limited
Abstract The monitoring of a patient can be carried out by continually measuring the vital parameters by using a pulse oximeter. SpO2 (Oxygen saturation) and the heart rate are the certain parameters for the critical patients to be monitored continuously. In this monitoring system, parameters of the patient are monitored by using pulse oximeter having two Light Emitting Diodes (LED’s) with dissimilar wavelengths. Continuous Pain free monitoring is achieved by a non-invasive monitoring procedure. In this paper, the measured values of SpO2 and the heart rate are wirelessly sent to the doctor through Wi-Fi. These measurements provide early caution to the person to avoid the possible problems in the body.
G. R. Ashisha, X. Anitha Mary, K. Rajasekaran, and R. Jegan
Springer Singapore