Unimodal and Multimodal Deep Learning for Pressure Injury Identification: Scoping Review
Unimodales und multimodales Deep Learning zur Identifikation von Pressure Injuries: Scoping Review
JMIR Medical Informatics · 10.08.2026 · 14 · e91048-e91048
DOI: 10.2196/91048 ↗Auf einen Blick
Die Scoping-Review ordnet Deep-Learning-Anwendungen zur Identifikation von Pressure Injuries ein. Von 15 Artikeln nutzten 6 multimodale Integration, 9 unimodale Bildgebung; die Genauigkeit lag zwischen 54,8 % und 93,7 %.
Fragestellung
Wie lässt sich die Literatur zu Deep-Learning-Anwendungen für die Identifikation von Pressure Injuries nach unimodaler Bildgebung und multimodaler Integration gliedern?
Methode
Scoping Review nach PRISMA-ScR; Suche in PubMed, Cochrane Library, IEEE Xplore und Web of Science.
Ergebnisse
15 Artikel wurden eingeschlossen: 6 zu multimodaler Integration und 9 zu unimodaler Bildgebung. Insgesamt waren 26 Modelle beteiligt; die Genauigkeit zur Stadieneinteilung lag zwischen 54,8 % und 93,7 %.
Einschränkungen
Die Landschaft ist stark zugunsten unimodaler Bildgebung verschoben; große Unterschiede der Erkennungsgenauigkeit zwischen Studien.
Schlussfolgerung der Autor:innen
Die aktuelle Landschaft des Deep Learning bei der Identifikation von Pressure Injuries ist stark auf unimodale Bildgebung ausgerichtet und erschwert die klinische Translation.
Kernaussagen
- 15 Artikel, 26 Modelle
- Genauigkeit zur Stadieneinteilung: 54,8 % bis 93,7 %
- Multimodale Integration erschien relevant für die klinische Validierung
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
Kurz gefasst
Scoping Review in der Zeitschrift JMIR Medical Informatics, online erschienen am 10.08.2026. Verfasst von Yingxue Sun und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Critical Care Medicine (Ward 2), Shandong Provincial Hospital Affiliated to Shandong First Medical University, No. 324, Jingwu-Weiqu Ro…. Hauptthemen laut MEDLINE: Deep Learning, Pressure Ulcer. Die Arbeit nennt 43 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.
Themen: Dekubitus
Abstract (Originalsprache)
Background
Pressure injury (PI) can cause severe infections or death, and has a high global prevalence. Accurate staging is vital for effective intervention. Deep learning streamlines PI assessment; enhances efficiency; and yields practical, accurate results.
Objective
The primary objective of this scoping review was to map the landscape of deep learning applications for PI identification and categorize the existing body of evidence into distinct technological trajectories: unimodal imaging and multimodal integration.
Methods
We searched the following databases and sources: PubMed, the Cochrane Library, IEEE Xplore, and Web of Science. This scoping review was conducted in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines.
Results
A total of 15 articles were included: 40% (n=6) studies applied multimodal integration, whereas 60% (n=9) applied unimodal imaging. In total, 26 models were involved. Different models exhibited varying accuracy rates in staging PI, with overall accuracy fluctuating between 54.8% and 93.7%. The same model demonstrated significant variations in recognition accuracy across different studies. Multimodal integration appeared to be a distinguishing characteristic of the studies that will progress toward clinical validation.
Conclusions
The current landscape of deep learning in PI identification is heavily skewed toward unimodal imaging, creating a modality gap that hinders clinical translation. While unimodal models demonstrate high accuracy in controlled settings, future research must prioritize bridging the gap from unimodal imaging to true multimodal integration to achieve clinical-grade reliability.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Critical Care Medicine (Ward 2), Shandong Provincial Hospital Affiliated to Shandong First Medical University, No. 324, Jingwu-Weiqu Road, Huaiyin District, Jinan, Shandong Province, 250021, China, 86 1586…
ORCID-Profile: Yingxue Sun, Congcong Liu, Ji-Cheng Zhang, Zhengang Wei, Zhenfeng Zhou, Chunmei Fan, Fangyan Yue
Zitieren
APA
Sun, Y., Liu, C., Zhang, J. C., Wei, Z., Zhou, Z., Fan, C., & Yue, F. (2026). Unimodal and Multimodal Deep Learning for Pressure Injury Identification: Scoping Review. JMIR Medical Informatics, 14, e91048-e91048. https://doi.org/10.2196/91048
Vancouver
Sun Y, Liu C, Zhang JC, Wei Z, Zhou Z, Fan C, et al. Unimodal and Multimodal Deep Learning for Pressure Injury Identification: Scoping Review. JMIR Medical Informatics. 2026;14:e91048-e91048. doi:10.2196/91048
BibTeX
@article{Sun2026,
author = {Sun, Yingxue and Liu, Congcong and Zhang, Ji-Cheng and Wei, Zhengang and Zhou, Zhenfeng and Fan, Chunmei and Yue, Fangyan},
title = {Unimodal and Multimodal Deep Learning for Pressure Injury Identification: Scoping Review},
journal = {JMIR Medical Informatics},
year = {2026},
volume = {14},
pages = {e91048--e91048},
doi = {10.2196/91048}
}
EUVIMED