Scholarlink

<p>Accurate and efficient medical image segmentation is a critical yet challenging task due to issues like intensity inhomogeneity, poor contrast, noise, and blur. In this paper, we introduce a novel framework that addresses these challenges by leveraging adaptive level set evolution, enhanced with a unique edge indication function. Unlike prior edge-based algorithms, which frequently fail with noisy images and have large computing costs, our method incorporates an improved edge indicator term into the level set architecture, considerably improving performance on degraded images. The efficiency of proposed model depends on the optimization and implementation of proximal alternating direction technique of multipliers (<InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="41598_2025_97789_Article_IEq1.gif" Format="GIF" Height="14" Rendition="HTML" Resolution="72" Type="Linedraw" Width="79" /> </InlineMediaObject> <EquationSource Format="TEX">\(PADMM\)</EquationSource> </InlineEquation>). Our findings were validated using qualitative and quantitative methods such as dice coefficient assessment, sensitivity, accuracy, and mean absolute distance (<i>MAD</i>). Experimental findings show that the model successfully detects boundaries of objects within noisy and blurred visual data. The algorithm showed exceptional precision through its average dice coefficient of 0.96 which matched the ground truth data measurement standards. The system runs efficiently for only 0.90 seconds on average as a performance result. The framework achieved standout performance metrics that included 0.9552 accuracy together with 0.8854 sensitivity and 0.0796 <i>MAD</i>. The framework demonstrates robust capabilities in medical image evaluation which makes it an optimistic instrument for advancing the field.</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

Enhanced medical image segmentation using novel level set evolution and efficient optimization

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年5月14日

查看全文
  • Samad Wali,
  • Adil Jhangeer,
  • Ariana Abdul Rahimzai,
  • Samina Samina,
  • Mudassar Imran

摘要

Accurate and efficient medical image segmentation is a critical yet challenging task due to issues like intensity inhomogeneity, poor contrast, noise, and blur. In this paper, we introduce a novel framework that addresses these challenges by leveraging adaptive level set evolution, enhanced with a unique edge indication function. Unlike prior edge-based algorithms, which frequently fail with noisy images and have large computing costs, our method incorporates an improved edge indicator term into the level set architecture, considerably improving performance on degraded images. The efficiency of proposed model depends on the optimization and implementation of proximal alternating direction technique of multipliers ( \(PADMM\) ). Our findings were validated using qualitative and quantitative methods such as dice coefficient assessment, sensitivity, accuracy, and mean absolute distance (MAD). Experimental findings show that the model successfully detects boundaries of objects within noisy and blurred visual data. The algorithm showed exceptional precision through its average dice coefficient of 0.96 which matched the ground truth data measurement standards. The system runs efficiently for only 0.90 seconds on average as a performance result. The framework achieved standout performance metrics that included 0.9552 accuracy together with 0.8854 sensitivity and 0.0796 MAD. The framework demonstrates robust capabilities in medical image evaluation which makes it an optimistic instrument for advancing the field.