Acute and Critical Care

Skip Navigation: Skip to contents

SEARCH: Search

Advanced Search

mobile menu button

Search

Page Path: HOME > Search

Original Articles

Neurosurgery
Deep learning-based model for detection of intracranial waveforms with poor brain compliance in southern Thailand: Thara Tunthanathip, Avika Trakulpanitkit; Acute Crit Care. 2025;40(3):473-481. Published online August 29, 2025; DOI: https://doi.org/10.4266/acc.001425

3,020 View
45 Download
1 Web of Science
5 Crossref

Abstract PDF

Background
Intracranial pressure (ICP) waveform analysis provides critical insights into brain compliance and can aid in the early detection of neurological deterioration. Deep learning (DL) has recently emerged as an effective approach for analyzing complex medical signals and imaging data. The aim of the present research was to develop a DL-based model for detecting ICP waveforms indicative of poor brain compliance. Methods: A retrospective cohort study was conducted using ICP wave images collected from postoperative hydrocephalus (HCP) patients who underwent ventriculostomy. The images were categorized into normal and poor compliance waveforms. Precision, recall, mean average precision at the 0.5 intersection over union (mAP_0.5), and the area under the receiver operating characteristic curve (AUC) were used to test. Results: The dataset consisted of 2,744 ICP wave images from 21 HCP patients. The best-performing model achieved a precision of 0.97, a recall of 0.96, and a mAP_0.5 of 0.989. The confusion matrix for poor brain compliance waveform detection using the test dataset also demonstrated a high classification accuracy, with true positive and true negative rates of 48.5% and 47.8%, respectively. Additionally, the model demonstrated high accuracy, achieving a mAP_0.5 of 0.994, sensitivity of 0.956, specificity of 0.970, and an AUC of 0.96 in the detection of poor compliance waveforms. Conclusions: The DL-based model successfully detected pathological ICP waveforms, thereby enhancing clinical decision-making. As DL advances, its significance in neurocritical care will help to pave the way for more individualized and data-driven approaches to brain monitoring and management

Citations

Citations to this article as recorded by

Perioperative Anesthetic Strategies in Emergent Neurosurgery During Severe Traumatic Brain Injury
Denise Baloi, Clayton Rawson, Deondra Montgomery, Michael Karsy, Mehrdad Pahlevani
Trauma Care.2026; 6(1): 5. CrossRef
Long-Term Outcome and Prognostic Factors of Mortality in Patients With Surgical Pulmonary Embolectomy
Supakorn Soontrapornchai, Voravit Chittithavorn, Thara Tunthanathip
Journal of Health and Allied Sciences NU.2026; 0: 1. CrossRef
Transfusion Indexes and Machine Learning Prediction of Red Cell Transfusion in Brain Tumor Surgery
Thara Tunthanathip, Natthanee Pisitthaworakul
Indian Journal of Neurosurgery.2026;[Epub] CrossRef
Quality Improvement in Neurosurgical Practice: Future Directions
Thara Tunthanathip
PSU Medical Journal.2026; 6(2): 53. CrossRef
Patient safety events and outcomes in brain tumor surgery : A retrospective cohort study
Kanisorn Sungkaro, Pannawit Chumchue, Thara Tunthanathip, Supaporn Sainamsai
Journal of Neurosciences in Rural Practice.2026; 0: 1. CrossRef

Pulmonary
An algorithm to predict the need for invasive mechanical ventilation in hospitalized COVID-19 patients: the experience in Sao Paulo: Eduardo Atsushi Osawa, Alexandre Toledo Maciel; Acute Crit Care. 2022;37(4):580-591. Published online September 8, 2022; DOI: https://doi.org/10.4266/acc.2022.00283

7,512 View
132 Download
2 Web of Science
2 Crossref

Abstract PDF

Background
We aimed to characterize patients hospitalized for coronavirus disease 2019 (COVID-19) and identify predictors of invasive mechanical ventilation (IMV).
Methods
We performed a retrospective cohort study in patients with COVID-19 admitted to a private network in Sao Paulo, Brazil from March to October 2020. Patients were compared in three subgroups: non-intensive care unit (ICU) admission (group A), ICU admission without receiving IMV (group B) and IMV requirement (group C). We developed logistic regression algorithm to identify predictors of IMV.
Results
We analyzed 1,650 patients, the median age was 53 years (42–65) and 986 patients (59.8%) were male. The median duration from symptom onset to hospital admission was 7 days (5–9) and the main comorbidities were hypertension (42.4%), diabetes (24.2%) and obesity (15.8%). We found differences among subgroups in laboratory values obtained at hospital admission. The predictors of IMV (odds ratio and 95% confidence interval [CI]) were male (1.81 [1.11– 2.94], P=0.018), age (1.03 [1.02–1.05], P<0.001), obesity (2.56 [1.57–4.15], P<0.001), duration from symptom onset to admission (0.91 [0.85–0.98], P=0.011), arterial oxygen saturation (0.95 [0.92– 0.99], P=0.012), C-reactive protein (1.005 [1.002–1.008], P<0.001), neutrophil-to-lymphocyte ratio (1.046 [1.005–1.089], P=0.029) and lactate dehydrogenase (1.005 [1.003–1.007], P<0.001). The area under the curve values were 0.860 (95% CI, 0.829–0.892) in the development cohort and 0.801 (95% CI, 0.733–0.870) in the validation cohort.
Conclusions
Patients had distinct clinical and laboratory parameters early in hospital admission. Our prediction model may enable focused care in patients at high risk of IMV.

Citations

Citations to this article as recorded by

Predictive Models of Patient Severity in Intensive Care Units Based on Serum Cytokine Profiles: Advancing Rapid Analysis
Cristiana P. Von Rekowski, Tiago A. H. Fonseca, Rúben Araújo, Ana Martins, Iola Pinto, M. Conceição Oliveira, Gonçalo C. Justino, Luís Bento, Cecília R. C. Calado
Applied Sciences.2025; 15(9): 4823. CrossRef
Intelligent alert system for predicting invasive mechanical ventilation needs via noninvasive parameters: employing an integrated machine learning method with integration of multicenter databases
Guang Zhang, Qingyan Xie, Chengyi Wang, Jiameng Xu, Guanjun Liu, Chen Su
Medical & Biological Engineering & Computing.2024; 62(11): 3445. CrossRef

First
Prev
Page of 1
Next
Last