Scholarlink

This paper introduces a groundbreaking algorithm that drastically reduces the time and cost of spinal MRI procedures while maintaining high diagnostic accuracy. The algorithm reconstructs a detailed 3D spinal model from 2D MRI slices taken along a single axis, improving visualization for more precise diagnosis, treatment planning, and patient education. By utilizing Lagrange interpolation, the method preserves edge fidelity and accelerates the reconstruction process, making spinal imaging faster and more efficient. Implemented on a Raspberry Pi platform that combines VLSI hardware with Python-based software, the system performs real-time denoising, interpolation, and 3D reconstruction. With a processing time reduced to just 10–20 min, this approach minimizes patient discomfort, particularly for elderly or critically ill patients, and offers a cost-effective, time-saving alternative to traditional MRI methods. The algorithm delivers high-quality reconstructions, achieving a Structural Similarity Index (SSIM) of approximately 0.96, Mutual Information (MI) of around 2.01, and a Root Mean Square (RMS) error of about 0.9. This innovative solution not only enhances computational techniques but also leverages affordable hardware, offering a scalable and accessible tool that can streamline clinical workflows and improve healthcare delivery.

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

Three-Dimensional MRI Data Reconstruction Using Edge Computing

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Rikta Pal,
  • Subharthi Ray,
  • Sayan Chatterjee,
  • Swagata Mandal,
  • Amlan Chakrabarti,
  • Somoballi Ghoshal

摘要

This paper introduces a groundbreaking algorithm that drastically reduces the time and cost of spinal MRI procedures while maintaining high diagnostic accuracy. The algorithm reconstructs a detailed 3D spinal model from 2D MRI slices taken along a single axis, improving visualization for more precise diagnosis, treatment planning, and patient education. By utilizing Lagrange interpolation, the method preserves edge fidelity and accelerates the reconstruction process, making spinal imaging faster and more efficient. Implemented on a Raspberry Pi platform that combines VLSI hardware with Python-based software, the system performs real-time denoising, interpolation, and 3D reconstruction. With a processing time reduced to just 10–20 min, this approach minimizes patient discomfort, particularly for elderly or critically ill patients, and offers a cost-effective, time-saving alternative to traditional MRI methods. The algorithm delivers high-quality reconstructions, achieving a Structural Similarity Index (SSIM) of approximately 0.96, Mutual Information (MI) of around 2.01, and a Root Mean Square (RMS) error of about 0.9. This innovative solution not only enhances computational techniques but also leverages affordable hardware, offering a scalable and accessible tool that can streamline clinical workflows and improve healthcare delivery.