Scholarlink

<p>This article presents a numerical investigation of the heat transfer during interaction of electromagnetic field (EM field) with biological medium in the context of nanoparticle assisted hyperthermia, along with the prediction of thermally induced apoptosis using machine learning technique. The variation in temperature and heat transfer in tissues are critical in the development and upgradation of nanoparticle assisted hyperthermia. The governing equations for electromagnetic wave propagation, transient heat transfer in biological medium induced by the EM field effect under local thermal non-equilibrium consideration, and thermally induced apoptosis have been written along with the appropriate initial and boundary conditions, which are then nondimensionalized. The finite-difference scheme is used to solve the governing heat transfer and apoptosis equations to discuss the effects of pertinent flow parameters on the blood, tissue phase temperatures and apoptosis. Moving a step further from usual numerical investigation, a support vector machine (SVM) is designed for classification of successful and unsuccessful apoptosis. The data required for training the machine has been generated by sweeping the pertinent parameters over appropriate ranges. It has been observed that higher blood volume fraction results in lower peak tissue temperatures. Also, the SVM correctly identifies <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(67.6\%\)</EquationSource> </InlineEquation> of successful treatments whereas it has <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(65.1\%\)</EquationSource> </InlineEquation> specificity for failure cases , achieving an Area Under the Curve (AUC) of 0.716 with standard deviation of 0.039. The SVM classifier is trained and evaluated entirely on simulation-derived labels; the framework is therefore hypothesis-generating and is proposed as a treatment planning exploration tool, not a clinically validated decision system.</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

Numerical investigation of heat transfer in biological medium with machine learning assisted prediction of apoptosis in nanoparticle-mediated hyperthermia

  • 期刊论文

  • 发表日期：

    2026年5月19日

查看全文
  • Abhishek Tiwari,
  • Shubham Tripathi,
  • Rajat Tripathi

摘要

This article presents a numerical investigation of the heat transfer during interaction of electromagnetic field (EM field) with biological medium in the context of nanoparticle assisted hyperthermia, along with the prediction of thermally induced apoptosis using machine learning technique. The variation in temperature and heat transfer in tissues are critical in the development and upgradation of nanoparticle assisted hyperthermia. The governing equations for electromagnetic wave propagation, transient heat transfer in biological medium induced by the EM field effect under local thermal non-equilibrium consideration, and thermally induced apoptosis have been written along with the appropriate initial and boundary conditions, which are then nondimensionalized. The finite-difference scheme is used to solve the governing heat transfer and apoptosis equations to discuss the effects of pertinent flow parameters on the blood, tissue phase temperatures and apoptosis. Moving a step further from usual numerical investigation, a support vector machine (SVM) is designed for classification of successful and unsuccessful apoptosis. The data required for training the machine has been generated by sweeping the pertinent parameters over appropriate ranges. It has been observed that higher blood volume fraction results in lower peak tissue temperatures. Also, the SVM correctly identifies \(67.6\%\) of successful treatments whereas it has \(65.1\%\) specificity for failure cases , achieving an Area Under the Curve (AUC) of 0.716 with standard deviation of 0.039. The SVM classifier is trained and evaluated entirely on simulation-derived labels; the framework is therefore hypothesis-generating and is proposed as a treatment planning exploration tool, not a clinically validated decision system.