Scholarlink

<p>In healthcare, breast cancer remains most common and aggressive health condition impacting women globally. Timely recurrence prediction is pivotal for effective prognosis and treatment planning. In our research, a multifactor prediction algorithm is proposed for identifying the low and high- risk recurrence and second primary breast cancers. Data on women for breast recurrence prediction is derived from Breast Cancer Coimbra Dataset, METABRIC Dataset and Gene Expression datasets- GSE2034 and GSE2990. These datasets comprehensively leverage demographic, clinical and genomic information. The proposed algorithm considers broad multiple influential variables like subject age, BMI, size of tumor, status of node and hormone receptor, stage of cancer, type of treatment, daily activity, demographics and dietary intakes to accurately distinguish between recurrence and second primary cancer cases. The superior prediction power of prediction algorithm is yielded by the incorporation of Cox Proportional Hazards model based survival analysis and Gradient Boosting based classification. This combined setup developed as an innovative solution, enhancing patient outcomes and supports ongoing initiatives to address breast cancer effectively. The multifactor model demonstrated exceptional results, attaining a peak accuracy of 98.75% and a precision rate of 97.15% in identifying the most critical cases. Beyond its high predictive accuracy, the algorithm offers practical clinical value by supporting healthcare professionals in making informed decisions about patient care and long-term management.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Signal processing techniques for multifactor breast cancer recurrence prediction: integrating clinical and genomic data

  • 期刊论文

  • 发表日期：

    2025年9月9日

查看全文
  • Vidhya Anbalagan,
  • Vanathi Balasubramanian

摘要

In healthcare, breast cancer remains most common and aggressive health condition impacting women globally. Timely recurrence prediction is pivotal for effective prognosis and treatment planning. In our research, a multifactor prediction algorithm is proposed for identifying the low and high- risk recurrence and second primary breast cancers. Data on women for breast recurrence prediction is derived from Breast Cancer Coimbra Dataset, METABRIC Dataset and Gene Expression datasets- GSE2034 and GSE2990. These datasets comprehensively leverage demographic, clinical and genomic information. The proposed algorithm considers broad multiple influential variables like subject age, BMI, size of tumor, status of node and hormone receptor, stage of cancer, type of treatment, daily activity, demographics and dietary intakes to accurately distinguish between recurrence and second primary cancer cases. The superior prediction power of prediction algorithm is yielded by the incorporation of Cox Proportional Hazards model based survival analysis and Gradient Boosting based classification. This combined setup developed as an innovative solution, enhancing patient outcomes and supports ongoing initiatives to address breast cancer effectively. The multifactor model demonstrated exceptional results, attaining a peak accuracy of 98.75% and a precision rate of 97.15% in identifying the most critical cases. Beyond its high predictive accuracy, the algorithm offers practical clinical value by supporting healthcare professionals in making informed decisions about patient care and long-term management.