Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs

Introduction Neonatal gastrointestinal perforation is a life-threatening condition that requires timely and accurate diagnosis. However, interpreting abdominal radiographs in this population is often challenging. In this study, we aimed to develop a deep convolutional neural network (DCNN) model to...

Descrición completa

Gardado en:
Detalles Bibliográficos
Autor Principal: figshare admin karger (2628495) (author)
Outros autores: Sanmoto Y. (19800489) (author), Zhang R. (4121311) (author), Peng B. (8640963) (author), Hosokawa T. (4127467) (author), Kondo Y. (4274551) (author), Inoue M. (3395735) (author), Miyazono Y. (22687679) (author), Zhu X. (3872488) (author), Masumoto K. (22687682) (author)
Publicado: 2025
Subjects:
Tags: Engadir etiqueta
Sen Etiquetas, Sexa o primeiro en etiquetar este rexistro!
_version_ 1849927622000115712
author figshare admin karger (2628495)
author2 Sanmoto Y. (19800489)
Zhang R. (4121311)
Peng B. (8640963)
Hosokawa T. (4127467)
Kondo Y. (4274551)
Inoue M. (3395735)
Miyazono Y. (22687679)
Zhu X. (3872488)
Masumoto K. (22687682)
author2_role author
author
author
author
author
author
author
author
author
author_facet figshare admin karger (2628495)
Sanmoto Y. (19800489)
Zhang R. (4121311)
Peng B. (8640963)
Hosokawa T. (4127467)
Kondo Y. (4274551)
Inoue M. (3395735)
Miyazono Y. (22687679)
Zhu X. (3872488)
Masumoto K. (22687682)
author_role author
dc.creator.none.fl_str_mv figshare admin karger (2628495)
Sanmoto Y. (19800489)
Zhang R. (4121311)
Peng B. (8640963)
Hosokawa T. (4127467)
Kondo Y. (4274551)
Inoue M. (3395735)
Miyazono Y. (22687679)
Zhu X. (3872488)
Masumoto K. (22687682)
dc.date.none.fl_str_mv 2025-11-26T06:55:21Z
dc.identifier.none.fl_str_mv 10.6084/m9.figshare.30718997.v1
dc.relation.none.fl_str_mv https://figshare.com/articles/dataset/Supplementary_Material_for_Computer-aided_Diagnosis_of_Pneumoperitoneum_on_Neonatal_Abdominal_Radiographs/30718997
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Medicine
Medicine
dc.title.none.fl_str_mv Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
dc.type.none.fl_str_mv Dataset
info:eu-repo/semantics/publishedVersion
dataset
description Introduction Neonatal gastrointestinal perforation is a life-threatening condition that requires timely and accurate diagnosis. However, interpreting abdominal radiographs in this population is often challenging. In this study, we aimed to develop a deep convolutional neural network (DCNN) model to segment pneumoperitoneum on neonatal abdominal radiographs and to evaluate its potential to assist in detecting neonatal gastrointestinal perforation. Methods This multicenter retrospective study included 1,187 abdominal radiographs (181 perforation and 1,006 control images) from neonates with gastrointestinal perforation and controls. Pneumoperitoneum regions were annotated by experienced clinicians. The dataset was randomly divided into training (n = 830), validation (n = 118), and test (n = 239) sets. A DeepLabV3+ model with ResNet50 backbone was finetuned for pixel-level segmentation. A single pixel-based threshold, derived from ROC analysis, was used to classify gastrointestinal perforation, with diagnostic performance subsequently compared to that of clinicians. Results The DCNN model achieved a median Dice similarity coefficient of 0.81 on the test dataset, indicating strong overlap between predicted and actual pneumoperitoneum regions. Furthermore, segmentation performance was positively correlated with pneumoperitoneum volume (Spearman ρ = 0.83, P < 0.001). Classification using the pixel-based cut-off demonstrated excellent diagnostic accuracy (AUC, 0.999; sensitivity, 100%; specificity, 98.5%), comparable to experienced clinicians. Conclusion The DCNN model demonstrated robust segmentation and classification performance, highlighting its potential as a clinical decision support tool for early detection of gastrointestinal perforation in neonates. Future studies should validate the model’s generalizability and assess its integration into clinical practice.
eu_rights_str_mv openAccess
id Manara_c135dcd0a1ada64b19c1061f1bede4d5
identifier_str_mv 10.6084/m9.figshare.30718997.v1
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/30718997
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographsfigshare admin karger (2628495)Sanmoto Y. (19800489)Zhang R. (4121311)Peng B. (8640963)Hosokawa T. (4127467)Kondo Y. (4274551)Inoue M. (3395735)Miyazono Y. (22687679)Zhu X. (3872488)Masumoto K. (22687682)MedicineMedicineIntroduction Neonatal gastrointestinal perforation is a life-threatening condition that requires timely and accurate diagnosis. However, interpreting abdominal radiographs in this population is often challenging. In this study, we aimed to develop a deep convolutional neural network (DCNN) model to segment pneumoperitoneum on neonatal abdominal radiographs and to evaluate its potential to assist in detecting neonatal gastrointestinal perforation. Methods This multicenter retrospective study included 1,187 abdominal radiographs (181 perforation and 1,006 control images) from neonates with gastrointestinal perforation and controls. Pneumoperitoneum regions were annotated by experienced clinicians. The dataset was randomly divided into training (n = 830), validation (n = 118), and test (n = 239) sets. A DeepLabV3+ model with ResNet50 backbone was finetuned for pixel-level segmentation. A single pixel-based threshold, derived from ROC analysis, was used to classify gastrointestinal perforation, with diagnostic performance subsequently compared to that of clinicians. Results The DCNN model achieved a median Dice similarity coefficient of 0.81 on the test dataset, indicating strong overlap between predicted and actual pneumoperitoneum regions. Furthermore, segmentation performance was positively correlated with pneumoperitoneum volume (Spearman ρ = 0.83, P < 0.001). Classification using the pixel-based cut-off demonstrated excellent diagnostic accuracy (AUC, 0.999; sensitivity, 100%; specificity, 98.5%), comparable to experienced clinicians. Conclusion The DCNN model demonstrated robust segmentation and classification performance, highlighting its potential as a clinical decision support tool for early detection of gastrointestinal perforation in neonates. Future studies should validate the model’s generalizability and assess its integration into clinical practice.2025-11-26T06:55:21ZDatasetinfo:eu-repo/semantics/publishedVersiondataset10.6084/m9.figshare.30718997.v1https://figshare.com/articles/dataset/Supplementary_Material_for_Computer-aided_Diagnosis_of_Pneumoperitoneum_on_Neonatal_Abdominal_Radiographs/30718997CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/307189972025-11-26T06:55:21Z
spellingShingle Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
figshare admin karger (2628495)
Medicine
Medicine
status_str publishedVersion
title Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
title_full Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
title_fullStr Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
title_full_unstemmed Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
title_short Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
title_sort Supplementary Material for: Computer-aided Diagnosis of Pneumoperitoneum on Neonatal Abdominal Radiographs
topic Medicine
Medicine