Tee-Tau Eric Nyam

Spontaneous intracerebral hemorrhage (SICH) is a devastating condition that significantly contributes to high mortality rates. This study aims to construct a mortality prediction model for patients with SICH using four various artificial intelligence (AI) machine learning algorithms. A retrospective analysis was conducted on electronic medical records of SICH patients aged 20 and above, admitted to Chi Mei Medical Center’s intensive care unit between January 2016 and December 2021. The study utilized 37 features related to mortality. Predictive models were developed using logistic regression, Random forest, LightGBM, XGBoost, and Multi-layer Perceptron (MLP), with assessments of feature importance, and Area under the curve (AUC). A total of 1451 SICH patients were enrolled. Factors associated with mortality included lower initial GCS scores (p < 0.001), pupillary changes (P < 0.001), kidney disease (p < 0.001), and respiratory failure requiring intubation (p < 0.001). Negative correlations were observed between mortality and pupil light reflexes, as well as GCS components E(r=-0.4602), V (r=-0.4132), M(r=-0.4082). Positive correlations were identified with vasopressors (r = 0.4464), FiO2 (r = 0.3901), and sedative-hypnotic drugs (r = 0.1178). XGBoost demonstrated the best predictive performance (AUC = 0.913), outperforming LR (0.899), RF (0.905), LightGBM (0.909), and MLP (0.892). The XGBoost model, utilizing both 18 and 36 features, continues to outperform both the Acute Physiology and Chronic Health Evaluation (APACHE II) (p < 0.001) and Sequential Organ Failure Assessment (SOFA) scoring systems (p < 0.001). This study successfully developed an AI mortality prediction model for SICH patients, with XGBoost exhibiting superior performance. The model, incorporating 18 key features, has been integrated into clinical practice assisting clinicians in treatment decisions and communication with patients’ families.

Background: Traumatic brain injury (TBI) research often focuses on mortality rates or functional recovery, yet the critical need for long-term care among patients dependent on institutional or Respiratory Care Ward (RCW) support remains underexplored. This study aims to address this gap by employing machine learning techniques to develop and validate predictive models that analyze the prognosis of this patient population. Method: Retrospective data from electronic medical records at Chi Mei Medical Center, encompassing 2020 TBI patients admitted to the ICU between January 2016 and December 2021, were collected. A total of 44 features were included, utilizing four machine learning models and various feature combinations based on clinical significance and Spearman correlation coefficients. Predictive performance was evaluated using the area under the curve (AUC) of the receiver operating characteristic (ROC) curve and validated with the DeLong test and SHAP (SHapley Additive exPlanations) analysis. Result: Notably, 236 patients (11.68%) were transferred to long-term care centers. XGBoost with 27 features achieved the highest AUC (0.823), followed by Random Forest with 11 features (0.817), and LightGBM with 44 features (0.813). The DeLong test revealed no significant differences among the best predictive models under various feature combinations. SHAP analysis illustrated a similar distribution of feature importance for the top 11 features in XGBoost, with 27 features, and Random Forest with 11 features. Conclusions: Random Forest, with an 11-feature combination, provided clinically meaningful predictive capability, offering early insights into long-term care trends for TBI patients. This model supports proactive planning for institutional or RCW resources, addressing a critical yet often overlooked aspect of TBI care.

Background: Numerous mortality prediction tools are currently available to assist patients with moderate to severe traumatic brain injury (TBI). However, an algorithm that utilizes various machine learning methods and employs diverse combinations of features to identify the most suitable predicting outcomes of brain injury patients in the intensive care unit (ICU) has not yet been well-established. Method: Between January 2016 and December 2021, we retrospectively collected data from the electronic medical records of Chi Mei Medical Center, comprising 2260 TBI patients admitted to the ICU. A total of 42 features were incorporated into the analysis using four different machine learning models, which were then segmented into various feature combinations. The predictive performance was assessed using the area under the curve (AUC) of the receiver operating characteristic (ROC) curve and validated using the Delong test. Result: The AUC for each model under different feature combinations ranged from 0.877 (logistic regression with 14 features) to 0.921 (random forest with 22 features). The Delong test indicated that the predictive performance of the machine learning models is better than that of traditional tools such as APACHE II and SOFA scores. Conclusion: Our machine learning training demonstrated that the predictive accuracy of the LightGBM is better than that of APACHE II and SOFA scores. These features are readily available on the first day of patient admission to the ICU. By integrating this model into the clinical platform, we can offer clinicians an immediate prognosis for the patient, thereby establishing a bridge for educating and communicating with family members.

Traumatic brain injury (TBI) remains a critical public health challenge. Although studies have found several prognostic factors for TBI, a useful early predictive tool for mortality has yet to be developed in the triage of the emergency room. This study aimed to use machine learning algorithms of artificial intelligence (AI) to develop predictive models for TBI patients in the emergency room triage. We retrospectively enrolled 18,249 adult TBI patients in the electronic medical records of three hospitals of Chi Mei Medical Group from January 2010 to December 2019, and undertook the 12 potentially predictive feature variables for predicting mortality during hospitalization. Six machine learning algorithms including logistical regression (LR) random forest (RF), support vector machines (SVM), LightGBM, XGBoost, and multilayer perceptron (MLP) were used to build the predictive model. The results showed that all six predictive models had high AUC from 0.851 to 0.925. Among these models, the LR-based model was the best model for mortality risk prediction with the highest AUC of 0.925; thus, we integrated the best model into the existed hospital information system for assisting clinical decision-making. These results revealed that the LR-based model was the best model to predict the mortality risk in patients with TBI in the emergency room. Since the developed prediction system can easily obtain the 12 feature variables during the initial triage, it can provide quick and early mortality prediction to clinicians for guiding deciding further treatment as well as helping explain the patient’s condition to family members.

Background: Traumatic brain injury (TBI) has been reported as a risk factor for brain cancer development. However, the magnitude of the impact of TBI on systemic cancer development has not been clarified. Methods: A retrospective longitudinal cohort study was conducted using the Taiwan Longitudinal Health Insurance Database between January 2000 and December 2011. A total of 35,306 patients were initially enrolled, and 14,795 patients with mild TBI and 14,795 patients with moderate/severe TBI were matched using the National Health Insurance Research Database in Taiwan. The Cox proportional hazard regression model was used to estimate the hazard ratio (HR) of TBI adjusted for potential confounding factors. Results: After matching, the results showed that patients with moderate/severe TBI had a high mortality rate (17.7% vs. 10.4%) and shorter time interval from TBI to death (mean 3.6 years vs. 5.8 years). No differences were observed in cancer incidence (4.1% vs. 4.1%) or risk factors for mortality between mild and moderate/severe TBI patients. However, patients aged between 46 and 55 years, female patients, and patients with pre-existing renal disease had a significant higher cancer incidence risk in moderate/severe TBI compared with mild TBI patients. The top 15 most common cancers showed that mild TBI patients had a higher percentage of head and neck cancer. The overall mortality rate in all TBI patients diagnosed with cancer was about 50%, and the cancer-specific mortality is approximately 85% in death of TBI patients with cancer. Conclusions: We concluded that the incidence risk of a new cancer diagnosis and mortality risk of TBI patients with cancer between the mild TBI and moderate/severe TBI patients were not significantly different.

Background: Ceftriaxone is a β-lactam antibiotic used to treat central nervous system infections. Whether the neuroprotective effects of ceftriaxone after TBI are mediated by attenuating neuroinflammation but not its antibacterial actions is not well established. Methods: Anesthetized male Sprague–Dawley rats were divided into sham-operated, TBI + vehicle, and TBI + ceftriaxone groups. Ceftriaxone was intraperitoneally injected at 0, 24, and 48 h with 50 or 250 mg/kg/day after TBI. During the first 120 min after TBI, we continuously measured heart rate, arterial pressure, intracranial pressure (ICP), and cerebral perfusion pressure. The infarct volume was measured by TTC staining. Motor function was measured using the inclined plane. Glutamate transporter 1 (GLT-1), neuronal apoptosis and TNF-α expression in the perilesioned cortex were investigated using an immunofluorescence assay. Bacterial evaluation was performed by Brown and Brenn’s Gram staining. These parameters above were measured at 72 h after TBI. Results: Compared with the TBI + vehicle group, the TBI + ceftriaxone 250 mg/kg group showed significantly lower ICP, improved motor dysfunction, reduced body weight loss, decreased infarct volume and neuronal apoptosis, decreased TBI-induced microglial activation and TNF-α expression in microglia, and increased GLT-1 expression in neurons and microglia. However, the grades of histopathological changes of antibacterial effects are zero. Conclusions: The intraperitoneal injection of ceftriaxone with 250 mg/kg/day for three days may attenuate TBI by increasing GLT-1 expression and reducing neuroinflammation and neuronal apoptosis, thereby resulting in an improvement in functional outcomes, and this neuroprotective effect is not related to its antibacterial effects.

Background: The aim of this study was to investigate whether AMN082 exerts its neuroprotective effect by attenuating glutamate receptor-associated neuronal apoptosis and improving functional outcomes after traumatic brain injury (TBI). Methods: Anesthetized male Sprague–Dawley rats were divided into the sham-operated, TBI + vehicle, and TBI + AMN082 groups. AMN082 (10 mg/kg) was intraperitoneally injected 0, 24, or 48 h after TBI. In the 120 min after TBI, heart rate, mean arterial pressure, intracranial pressure (ICP), and cerebral perfusion pressure (CPP) were continuously measured. Motor function, the infarct volume, neuronal nitrosative stress-associated apoptosis, and N-methyl-d-aspartate receptor 2A (NR2A) and NR2B expression in the pericontusional cortex were measured on the 3rd day after TBI. Results: The results showed that the AMN082-treated group had a lower ICP and higher CPP after TBI. TBI-induced motor deficits, the increase in infarct volume, neuronal apoptosis, and 3-nitrotyrosine and inducible nitric oxide synthase expression in the pericontusional cortex were significantly improved by AMN082 therapy. Simultaneously, AMN082 increased NR2A and NR2B expression in neuronal cells. Conclusions: We concluded that intraperitoneal injection of AMN082 for 3 days may ameliorate TBI by attenuating glutamate receptor-associated nitrosative stress and neuronal apoptosis in the pericontusional cortex. We suggest that AMN082 administration in the acute stage may be a promising strategy for TBI.

Background To date, cardiac dysfunction after traumatic brain injury (TBI) has not been consistent. In this study, we hypothesized that TBI may play a role in the development of new-onset cardiac dysfunction in healthy experimental rats. Materials and Methods Anesthetized healthy male Sprague–Dawley rats were divided into two groups: a sham-operated control group and a TBI group. The brain was injured with 2.4 atm percussion via a fluid percussion injury model. During the 120 min after TBI, we continuously measured brain parameters, including intracranial pressure (ICP) and cerebral perfusion pressure (CPP), and cardiac parameters, such as heart rate (HR), inter-ventricular septum dimension (IVSD), left ventricular internal dimension diastole (LVIDd), end-diastolic volume (EDV), ejection fraction (EF), fractional shortening (FS), and LV mass diastole (LVd mass) by cardiac echo. On days 1, 3, 7, and 14 after TBI, the brain damage volume was evaluated with triphenyltetrazolium chloride; the physiological parameters of the heart, including HR, IVSd, LVIDd, EDV, EF, FS, and LVd mass, were evaluated with cardiac echo; the morphology of cardiomyocytes was examined by hematoxylin and eosin (HE) and Masson trichrome staining; and the biomarkers of cardiac injury troponin I and B-type natriuretic peptide (BNP) were also examined. Results Compared to sham-operated controls, the TBI groups had higher ICP, lower CPP, and higher brain neuronal apoptosis and infarction contusion volume. The impact of TBI on heart function showed hyperdynamic response trends in IVSd, LVIDd, EDV, EF, FS, and LVd mass within 30 min after TBI; however, EF and FS exhibited eventual decreasing trends. Simultaneously, the values of the biomarkers troponin I and BNP were within normal limits, and HE and Mass trichrome staining revealed no significant differences between the sham-operated control group and the TBI group. Conclusions Our results suggest that TBI due to 2.4 atm fluid percussion injury in healthy experimental rats may cause significant damage to the brain and affect the heart function as investigated by cardiac echo but not as investigated by HE and Masson trichrome stainings or troponin I and BNP evaluation.