Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertension

Abstract To evaluate the effectiveness of deep learning radiomics nomogram in distinguishing early intracranial hypertension (IH) following primary decompressive craniectomy (DC) in patients with severe traumatic brain injury (TBI) and to demonstrate its potential clinical value as a noninvasive too...

Full description

Saved in:
Bibliographic Details
Main Authors: Zunfeng Fu, Jing Wang, Wenyi Shen, Yanqing Wu, Jiajun Zhang, Yan Liu, Chongqiang Wang, Yanlin Shen, Ye Zhu, Weifu Zhang, Chunju Lv, Lin Peng
Format: Article
Language:English
Published: Nature Portfolio 2025-07-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-09115-7
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849238581266087936
author Zunfeng Fu
Jing Wang
Wenyi Shen
Yanqing Wu
Jiajun Zhang
Yan Liu
Chongqiang Wang
Yanlin Shen
Ye Zhu
Weifu Zhang
Chunju Lv
Lin Peng
author_facet Zunfeng Fu
Jing Wang
Wenyi Shen
Yanqing Wu
Jiajun Zhang
Yan Liu
Chongqiang Wang
Yanlin Shen
Ye Zhu
Weifu Zhang
Chunju Lv
Lin Peng
author_sort Zunfeng Fu
collection DOAJ
description Abstract To evaluate the effectiveness of deep learning radiomics nomogram in distinguishing early intracranial hypertension (IH) following primary decompressive craniectomy (DC) in patients with severe traumatic brain injury (TBI) and to demonstrate its potential clinical value as a noninvasive tool for guiding timely intervention and improving patient outcomes. This study included 238 patients with severe TBI (training cohort: n = 166; testing cohort: n = 72). Postoperative ultrasound images of the optic nerve sheath (ONS) and Spectral doppler imaging of middle cerebral artery (MCASDI) were obtained at 6 and 18 h after DC. Patients were grouped according to threshold values of 15 mmHg and 20 mmHg based on invasive intracranial pressure (ICPi) measurements. Clinical-semantic features were collected, and radiomics features were extracted from ONS images, and Additionally, deep transfer learning (DTL) features were generated using RseNet101. Predictive models were developed using the Light Gradient Boosting Machine (light GBM) machine learning algorithm. Clinical-ultrasound variables were incorporated into the model through univariate and multivariate logistic regression. A combined nomogram was developed by integrating DLR (deep learning radiomics) features with clinical-ultrasound variables, and its diagnostic performance over different thresholds was evaluated using Receiver Operating Characteristic (ROC) curve analysis and decision curve analysis (DCA). The nomogram model demonstrated superior performance over the clinical model at both 15 mmHg and 20 mmHg thresholds. For 15 mmHg, the AUC was 0.974 (95% confidence interval [CI]: 0.953–0.995) in the training cohort and 0.919 (95% CI: 0.845–0.993) in the testing cohort. For 20 mmHg, the AUC was 0.968 (95% CI: 0.944–0.993) in the training cohort and 0.889 (95% CI: 0.806–0.972) in the testing cohort. DCA curves showed net clinical benefit across all models. Among DLR models based on ONS, MCASDI, or their pre-fusion, the ONS-based model performed best in the testing cohorts. The nomogram model, incorporating clinical-semantic features, radiomics, and DTL features, exhibited promising performance in predicting early IH in post-DC patients. It shows promise for enhancing non-invasive ICP monitoring and supporting individualized therapeutic strategies.
format Article
id doaj-art-e35a0e92057c4b99bcee46ada9ade620
institution Kabale University
issn 2045-2322
language English
publishDate 2025-07-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-e35a0e92057c4b99bcee46ada9ade6202025-08-20T04:01:34ZengNature PortfolioScientific Reports2045-23222025-07-0115111310.1038/s41598-025-09115-7Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertensionZunfeng Fu0Jing Wang1Wenyi Shen2Yanqing Wu3Jiajun Zhang4Yan Liu5Chongqiang Wang6Yanlin Shen7Ye Zhu8Weifu Zhang9Chunju Lv10Lin Peng11Department of Ultrasound, The Second Affiliated Hospital of Shandong First Medical UniversityDepartment of Ultrasound, The Second Affiliated Hospital of Shandong First Medical UniversityDepartment of Ultrasound, The Second Affiliated Hospital of Shandong First Medical UniversityDepartment of Ultrasound, The Affiliated Hospital of Qilu Medical UniversityDepartment of Ultrasound, The Second Affiliated Hospital of Shandong First Medical UniversityDepartment of Ultrasound, The Second Affiliated Hospital of Shandong First Medical UniversityDepartment of Ultrasound, The Second Affiliated Hospital of Shandong First Medical UniversityCollege of Medical Imaging and Laboratory, Jining Medical UniversityCollege of Radiology, Shandong First Medical UniversityDepartment of Public Health Management, The Second Affiliated Hospital of Shandong, First Medical UniversityDepartment of Ultrasound, The Second Affiliated Hospital of Shandong First Medical UniversityDepartment of General Practice, The Second Affiliated Hospital of Shandong First Medical UniversityAbstract To evaluate the effectiveness of deep learning radiomics nomogram in distinguishing early intracranial hypertension (IH) following primary decompressive craniectomy (DC) in patients with severe traumatic brain injury (TBI) and to demonstrate its potential clinical value as a noninvasive tool for guiding timely intervention and improving patient outcomes. This study included 238 patients with severe TBI (training cohort: n = 166; testing cohort: n = 72). Postoperative ultrasound images of the optic nerve sheath (ONS) and Spectral doppler imaging of middle cerebral artery (MCASDI) were obtained at 6 and 18 h after DC. Patients were grouped according to threshold values of 15 mmHg and 20 mmHg based on invasive intracranial pressure (ICPi) measurements. Clinical-semantic features were collected, and radiomics features were extracted from ONS images, and Additionally, deep transfer learning (DTL) features were generated using RseNet101. Predictive models were developed using the Light Gradient Boosting Machine (light GBM) machine learning algorithm. Clinical-ultrasound variables were incorporated into the model through univariate and multivariate logistic regression. A combined nomogram was developed by integrating DLR (deep learning radiomics) features with clinical-ultrasound variables, and its diagnostic performance over different thresholds was evaluated using Receiver Operating Characteristic (ROC) curve analysis and decision curve analysis (DCA). The nomogram model demonstrated superior performance over the clinical model at both 15 mmHg and 20 mmHg thresholds. For 15 mmHg, the AUC was 0.974 (95% confidence interval [CI]: 0.953–0.995) in the training cohort and 0.919 (95% CI: 0.845–0.993) in the testing cohort. For 20 mmHg, the AUC was 0.968 (95% CI: 0.944–0.993) in the training cohort and 0.889 (95% CI: 0.806–0.972) in the testing cohort. DCA curves showed net clinical benefit across all models. Among DLR models based on ONS, MCASDI, or their pre-fusion, the ONS-based model performed best in the testing cohorts. The nomogram model, incorporating clinical-semantic features, radiomics, and DTL features, exhibited promising performance in predicting early IH in post-DC patients. It shows promise for enhancing non-invasive ICP monitoring and supporting individualized therapeutic strategies.https://doi.org/10.1038/s41598-025-09115-7Severe traumatic brain injuryIntracranial pressureUltrasound radiomicsMachine learningOptic nerve sheath diameter(ONSD)Transcranial color doppler (TCCD)
spellingShingle Zunfeng Fu
Jing Wang
Wenyi Shen
Yanqing Wu
Jiajun Zhang
Yan Liu
Chongqiang Wang
Yanlin Shen
Ye Zhu
Weifu Zhang
Chunju Lv
Lin Peng
Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertension
Scientific Reports
Severe traumatic brain injury
Intracranial pressure
Ultrasound radiomics
Machine learning
Optic nerve sheath diameter(ONSD)
Transcranial color doppler (TCCD)
title Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertension
title_full Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertension
title_fullStr Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertension
title_full_unstemmed Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertension
title_short Multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post-craniotomy intracranial hypertension
title_sort multimodal nomogram integrating deep learning radiomics and hemodynamic parameters for early prediction of post craniotomy intracranial hypertension
topic Severe traumatic brain injury
Intracranial pressure
Ultrasound radiomics
Machine learning
Optic nerve sheath diameter(ONSD)
Transcranial color doppler (TCCD)
url https://doi.org/10.1038/s41598-025-09115-7
work_keys_str_mv AT zunfengfu multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT jingwang multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT wenyishen multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT yanqingwu multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT jiajunzhang multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT yanliu multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT chongqiangwang multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT yanlinshen multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT yezhu multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT weifuzhang multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT chunjulv multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension
AT linpeng multimodalnomogramintegratingdeeplearningradiomicsandhemodynamicparametersforearlypredictionofpostcraniotomyintracranialhypertension