Academic Journal
View in EDS

MS-HIENet: multi-scale hybrid implicit-explicit registration network.

Bibliographic Details
Title: MS-HIENet: multi-scale hybrid implicit-explicit registration network.
Authors: Wang Z; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China., Liu X; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China., Chen X; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China., Wei Y; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China.; Shandong Research Institute of Industrial Technology, Jinan, Shandong Province, People's Republic of China.
Source: Physics in medicine and biology [Phys Med Biol] 2026 Feb 25; Vol. 71 (4). Date of Electronic Publication: 2026 Feb 25.
Publication Type: Journal Article
Language: English
Journal Info: Publisher: IOP Publishing Country of Publication: England NLM ID: 0401220 Publication Model: Electronic Cited Medium: Internet ISSN: 1361-6560 (Electronic) Linking ISSN: 00319155 NLM ISO Abbreviation: Phys Med Biol Subsets: MEDLINE
Imprint Name(s): Original Publication: Bristol : IOP Publishing
MeSH Terms: Image Processing, Computer-Assisted*/methods, Lung/diagnostic imaging ; Convolutional Neural Networks ; Humans ; Tomography, X-Ray Computed
Abstract: Objective.Medical image registration serves as a cornerstone for precision diagnosis and treatment, particularly for dynamic organs like the lungs, where high deformability and complex motion patterns impose significant challenges. Traditional iterative methods are time-consuming, while existing deep learning approaches often struggle to synergistically optimize large deformation modeling and localized fine structure preservation due to limited receptive fields. This study aims to address the challenge of jointly handling large-scale and localized deformations in pulmonary imaging by developing a novel deep learning framework.Approach.We propose a multi-scale hybrid implicit-explicit registration network (MS-HIENet), a mask-free end-to-end framework that integrates implicit neural representation (INR) and convolutional neural networks (CNN). The method employs a two-fold strategy: First, a multi-scale optimization mechanism where low-resolution layers leverage INR to capture global deformations and high-resolution layers utilize CNN to refine local anatomical structures, enabling coarse-to-fine hierarchical registration. Second, an INR-based coordinate-to-displacement implicit mapping framework is used to directly model continuous deformation fields, eliminating the dependency on mask annotations.Main results.Experimental results on the deformable image registration laboratory dataset demonstrate that MS-HIENet achieves a mean target registration error of 1.00 mm, representing an average reduction of 29.5% compared to state-of-the-art deep learning methods. Ablation studies validate the effectiveness of the multi-scale collaboration and hybrid implicit-explicit representation, with the deformation field folding rate reaching an minimal level (mean: 0.00017).Significance.The proposed method effectively bridges the gap between global deformation consistency and local anatomical precision. By combining the continuous modeling capabilities of INR with the local feature refinement of CNNs, MS-HIENet significantly enhances topological consistency and clinical applicability, offering a robust solution for high-precision lung image analysis.
(© 2026 Institute of Physics and Engineering in Medicine. All rights, including for text and data mining, AI training, and similar technologies, are reserved.)
Contributed Indexing: Keywords: deformable registration; implicit neural representation; lung CT; multi-scale
Entry Date(s): Date Created: 20260213 Date Completed: 20260707 Latest Revision: 20260707
Update Code: 20260708
DOI: 10.1088/1361-6560/ae45e7
PMID: 41687255
Database: MEDLINE
FullText Text:
  Availability: 0
Header DbId: cmedm
DbLabel: MEDLINE
An: 41687255
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: MS-HIENet: multi-scale hybrid implicit-explicit registration network.
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AU" term="%22Wang+Z%22">Wang Z</searchLink>; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China.<br /><searchLink fieldCode="AU" term="%22Liu+X%22">Liu X</searchLink>; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China.<br /><searchLink fieldCode="AU" term="%22Chen+X%22">Chen X</searchLink>; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China.<br /><searchLink fieldCode="AU" term="%22Wei+Y%22">Wei Y</searchLink>; School of Control Science and Engineering, Shandong University, Jinan, Shandong Province, People's Republic of China.; Shandong Research Institute of Industrial Technology, Jinan, Shandong Province, People's Republic of China.
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="JN" term="%220401220%22">Physics in medicine and biology</searchLink> [Phys Med Biol] 2026 Feb 25; Vol. 71 (4). <i>Date of Electronic Publication: </i>2026 Feb 25.
– Name: TypePub
  Label: Publication Type
  Group: TypPub
  Data: Journal Article
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: TitleSource
  Label: Journal Info
  Group: Src
  Data: <i>Publisher: </i><searchLink fieldCode="PB" term="%22IOP+Publishing%22">IOP Publishing </searchLink><i>Country of Publication: </i>England <i>NLM ID: </i>0401220 <i>Publication Model: </i>Electronic <i>Cited Medium: </i>Internet <i>ISSN: </i>1361-6560 (Electronic) <i>Linking ISSN: </i><searchLink fieldCode="IS" term="%2200319155%22">00319155 </searchLink><i>NLM ISO Abbreviation: </i>Phys Med Biol <i>Subsets: </i>MEDLINE
– Name: PublisherInfo
  Label: Imprint Name(s)
  Group: PubInfo
  Data: <i>Original Publication</i>: Bristol : IOP Publishing
– Name: SubjectMESH
  Label: MeSH Terms
  Group: Su
  Data: <searchLink fieldCode="MM" term="%22Image+Processing%2C+Computer-Assisted%22">Image Processing, Computer-Assisted*</searchLink>/<searchLink fieldCode="MM" term="%22Image+Processing%2C+Computer-Assisted+methods%22">methods</searchLink><br /><searchLink fieldCode="MH" term="%22Lung%22">Lung</searchLink>/<searchLink fieldCode="MH" term="%22Lung+diagnostic+imaging%22">diagnostic imaging</searchLink> ; <searchLink fieldCode="MH" term="%22Convolutional+Neural+Networks%22">Convolutional Neural Networks</searchLink> ; <searchLink fieldCode="MH" term="%22Humans%22">Humans</searchLink> ; <searchLink fieldCode="MH" term="%22Tomography%2C+X-Ray+Computed%22">Tomography, X-Ray Computed</searchLink>
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Objective.Medical image registration serves as a cornerstone for precision diagnosis and treatment, particularly for dynamic organs like the lungs, where high deformability and complex motion patterns impose significant challenges. Traditional iterative methods are time-consuming, while existing deep learning approaches often struggle to synergistically optimize large deformation modeling and localized fine structure preservation due to limited receptive fields. This study aims to address the challenge of jointly handling large-scale and localized deformations in pulmonary imaging by developing a novel deep learning framework.Approach.We propose a multi-scale hybrid implicit-explicit registration network (MS-HIENet), a mask-free end-to-end framework that integrates implicit neural representation (INR) and convolutional neural networks (CNN). The method employs a two-fold strategy: First, a multi-scale optimization mechanism where low-resolution layers leverage INR to capture global deformations and high-resolution layers utilize CNN to refine local anatomical structures, enabling coarse-to-fine hierarchical registration. Second, an INR-based coordinate-to-displacement implicit mapping framework is used to directly model continuous deformation fields, eliminating the dependency on mask annotations.Main results.Experimental results on the deformable image registration laboratory dataset demonstrate that MS-HIENet achieves a mean target registration error of 1.00 mm, representing an average reduction of 29.5% compared to state-of-the-art deep learning methods. Ablation studies validate the effectiveness of the multi-scale collaboration and hybrid implicit-explicit representation, with the deformation field folding rate reaching an minimal level (mean: 0.00017).Significance.The proposed method effectively bridges the gap between global deformation consistency and local anatomical precision. By combining the continuous modeling capabilities of INR with the local feature refinement of CNNs, MS-HIENet significantly enhances topological consistency and clinical applicability, offering a robust solution for high-precision lung image analysis.<br /> (© 2026 Institute of Physics and Engineering in Medicine. All rights, including for text and data mining, AI training, and similar technologies, are reserved.)
– Name: SubjectMinor
  Label: Contributed Indexing
  Group:
  Data: <i>Keywords: </i>deformable registration; implicit neural representation; lung CT; multi-scale
– Name: DateEntry
  Label: Entry Date(s)
  Group: Date
  Data: <i>Date Created: </i>20260213 <i>Date Completed: </i>20260707 <i>Latest Revision: </i>20260707
– Name: DateUpdate
  Label: Update Code
  Group: Date
  Data: 20260708
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1088/1361-6560/ae45e7
– Name: AN
  Label: PMID
  Group: ID
  Data: 41687255
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=cmedm&AN=41687255
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1088/1361-6560/ae45e7
    Languages:
      – Code: eng
        Text: English
    Subjects:
      – SubjectFull: Lung diagnostic imaging
        Type: general
      – SubjectFull: Convolutional Neural Networks
        Type: general
      – SubjectFull: Humans
        Type: general
      – SubjectFull: Tomography, X-Ray Computed
        Type: general
      – SubjectFull: Image Processing, Computer-Assisted methods
        Type: general
    Titles:
      – TitleFull: MS-HIENet: multi-scale hybrid implicit-explicit registration network.
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Wang Z
      – PersonEntity:
          Name:
            NameFull: Liu X
      – PersonEntity:
          Name:
            NameFull: Chen X
      – PersonEntity:
          Name:
            NameFull: Wei Y
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 25
              M: 02
              Text: 2026 Feb 25
              Type: published
              Y: 2026
          Identifiers:
            – Type: issn-electronic
              Value: 1361-6560
          Numbering:
            – Type: volume
              Value: 71
            – Type: issue
              Value: 4
          Titles:
            – TitleFull: Physics in medicine and biology
              Type: main
ResultId 1