Shetty Nisha P, Shetty Jayashree, Hegde Veeraj, Dharne Sneha Dattatray, Kv Mamtha
Department of Information and Communication Technology, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal, India.
Department of Information and Communication Technology, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal, India.
J Ayurveda Integr Med. 2024 Nov-Dec;15(6):101051. doi: 10.1016/j.jaim.2024.101051. Epub 2024 Dec 10.
Gestational Diabetes Mellitus (GDM) is a metabolic condition that develops in course of pregnancy. The World Health Organization describes it as carbohydrate intolerance that causes hyperglycemia of varying severity and manifests itself or is first noticed during pregnancy. Early prediction is now possible, owing to the application of cutting-edge methods like machine learning.
In the proposed empirical study, different machine-learning algorithms are applied to predict the prospective risk factors influencing the progression of GDM in gestating mothers.
The performance of these algorithms is evaluated through accuracy, precision, f1-score, etc. The lifestyle interventions and medications listed in Ayurveda literature are discussed for effective management of the disease.
Most of the proposed classifiers achieved a reasonable accuracy range of 75-82 %. Appropriate lifestyle changes, herbal remedies, decoctions, and churnas have all been shown to be useful in lowering the risk of GDM. Early detection using machine learning models can significantly reduce disease severity by facilitating timely Ayurvedic interventions.
The proposed work is more focused on the identification of factors impacting GDM in expectant women. A balanced diet with physical exercise, proper medication, and better lifestyle management (through Garbini Paricharya) can control the perils of GDM if diagnosed prematurely.
妊娠期糖尿病（GDM）是一种在妊娠期间发生的代谢性疾病。世界卫生组织将其描述为导致不同程度高血糖的碳水化合物不耐受，在妊娠期间出现或首次被发现。由于机器学习等前沿方法的应用，现在可以进行早期预测。
在本实证研究中，应用不同的机器学习算法来预测影响妊娠期母亲GDM进展的潜在风险因素。
通过准确率、精确率、F1分数等来评估这些算法的性能。讨论了阿育吠陀文献中列出的生活方式干预措施和药物，以有效管理该疾病。
大多数提出的分类器达到了75%-82%的合理准确率范围。适当的生活方式改变、草药、汤剂和 churnas 都已被证明有助于降低GDM的风险。使用机器学习模型进行早期检测可以通过促进及时的阿育吠陀干预显著降低疾病严重程度。
本研究更侧重于识别影响孕妇GDM的因素。如果早产诊断，均衡饮食、体育锻炼、适当用药和更好的生活方式管理（通过孕期保健）可以控制GDM的危害。
