Validation of machine-learning model for first-trimester prediction of pre-eclampsia using cohort from PREVAL study.

OBJECTIVE: Effective first-trimester screening for pre-eclampsia (PE) can be achieved using a competing-risks model that combines risk factors from the maternal history with multiples of the median (MoM) values of biomarkers. A new model using artificial intelligence through machine-learning methods has been shown to achieve similar screening performance without the need for conversion of raw data of biomarkers into MoM. This study aimed to investigate whether this model can be used across populations without specific adaptations. METHODS: Previously, a machine-learning model derived with the use of a fully connected neural network for first-trimester prediction of early (< 34 weeks), preterm (< 37 weeks) and all PE was developed and tested in a cohort of pregnant women in the UK. The model was based on maternal risk factors and mean arterial blood pressure (MAP), uterine artery pulsatility index (UtA-PI), placental growth factor (PlGF) and pregnancy-associated plasma protein-A (PAPP-A). In this study, the model was applied to a dataset of 10 110 singleton pregnancies examined in Spain who participated in the first-trimester PE validation (PREVAL) study, in which first-trimester screening for PE was carried out using the Fetal Medicine Foundation (FMF) competing-risks model. The performance of screening was assessed by examining the area under the receiver-operating-characteristics curve (AUC) and detection rate (DR) at a 10% screen-positive rate (SPR). These indices were compared with those derived from the application of the FMF competing-risks model. The performance of screening was poor if no adjustment was made for the analyzer used to measure PlGF, which was different in the UK and Spain. Therefore, adjustment for the analyzer used was performed using simple linear regression. RESULTS: The DRs at 10% SPR for early, preterm and all PE with the machine-learning model were 84.4% (95% CI, 67.2-94.7%), 77.8% (95% CI, 66.4-86.7%) and 55.7% (95% CI, 49.0-62.2%), respectively, with the corresponding AUCs of 0.920 (95% CI, 0.864-0.975), 0.913 (95% CI, 0.882-0.944) and 0.846 (95% CI, 0.820-0.872). This performance was achieved with the use of three of the biomarkers (MAP, UtA-PI and PlGF); inclusion of PAPP-A did not provide significant improvement in DR. The machine-learning model had similar performance to that achieved by the FMF competing-risks model (DR at 10% SPR, 82.7% (95% CI, 69.6-95.8%) for early PE, 72.7% (95% CI, 62.9-82.6%) for preterm PE and 55.1% (95% CI, 48.8-61.4%) for all PE) without requiring specific adaptations to the population. CONCLUSIONS: A machine-learning model for first-trimester prediction of PE based on a neural network provides effective screening for PE that can be applied in different populations. However, before doing so, it is essential to make adjustments for the analyzer used for biochemical testing. © 2023 International Society of Ultrasound in Obstetrics and Gynecology.

Machine-learning-based prediction of pre-eclampsia using first-trimester maternal characteristics and biomarkers.

OBJECTIVE: To evaluate the accuracy of predicting the risk of developing pre-eclampsia (PE) according to first-trimester maternal demographic characteristics, medical history and biomarkers using artificial-intelligence and machine-learning methods. METHODS: The data were derived from prospective non-interventional screening for PE at 11-13 weeks' gestation at two maternity hospitals in the UK. The data were divided into three subsets. The first set, including 30 437 subjects, was used to develop the training process, the second set of 10 000 subjects was utilized to optimize the machine-learning hyperparameters and the third set of 20 352 subjects was coded and used for model validation. An artificial neural network was used to predict from the demographic characteristics and medical history the prior risk that was then combined with biomarker values to determine the risk of PE and preterm PE with delivery at < 37 weeks' gestation. An additional network was trained without including race as input. Biomarkers included uterine artery pulsatility index (UtA-PI), mean arterial blood pressure (MAP), placental growth factor (PlGF) and pregnancy-associated plasma protein-A. All markers were entered using raw values without conversion into standardized multiples of the median. The prediction accuracy was estimated using the area under the receiver-operating-characteristics curve (AUC). We further computed the detection rate at 10%, 20% and 40% false-positive rates (FPR). The impact of taking aspirin was also added. Shapley values were calculated to evaluate the contribution of each parameter to the prediction of risk. We used a non-parametric test to compare the expected AUC with the one obtained when we randomly scrambled the labels and kept the predictions. For the general prediction, we performed 10 000 permutations of the labels. When the AUC was higher than the one obtained in all 10 000 permutations, we reported a P-value of < 0.0001. For the race-specific analysis, we performed 1000 permutations. When the AUC was higher than the AUC in permutations, we reported a P-value of < 0.001. RESULTS: The detection rate for preterm PE vs no PE, at a 10% FPR, was 53.3% when screening by maternal factors only, and the corresponding AUC was 0.816; these increased to 75.3% and 0.909, respectively, with the addition of biomarkers into the model. Information on race was important for the prediction accuracy; when race was not used to train the model, at a 10% FPR, the detection rate of preterm PE vs no PE decreased to 34.5-45.5% (for different races) when screening by maternal factors only and to 55.0-62.1% when biomarkers were added. The major predictors of PE were high MAP and UtA-PI, and low PlGF. The accuracy of prediction of all PE cases was lower than that for preterm PE. Aspirin use was recommended for cases who were at high risk of preterm PE. The AUC of all PE vs no PE was 0.770 when screening by maternal factors and 0.817 when the biomarkers were added; the respective detection rates, at a 10% FPR, were 41.3% and 52.9%. CONCLUSIONS: Screening for PE using a non-linear machine-learning-based approach does not require a population-based normalization, and its performance is similar to that of logistic regression. Removing race information from the model reduces its prediction accuracy, especially for the non-white populations when only maternal factors are considered. © 2022 International Society of Ultrasound in Obstetrics and Gynecology.

A snapshot of European neurosurgery December 2019 vs. March 2020: just before and during the Covid-19 pandemic.

BACKGROUND: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2 or Covid-19), which began as an epidemic in China and spread globally as a pandemic, has necessitated resource management to meet emergency needs of Covid-19 patients and other emergent cases. We have conducted a survey to analyze caseload and measures to adapt indications for a perception of crisis. METHODS: We constructed a questionnaire to survey a snapshot of neurosurgical activity, resources, and indications during 1 week with usual activity in December 2019 and 1 week during SARS-CoV-2 pandemic in March 2020. The questionnaire was sent to 34 neurosurgical departments in Europe; 25 departments returned responses within 5 days. RESULTS: We found unexpectedly large differences in resources and indications already before the pandemic. Differences were also large in how much practice and resources changed during the pandemic. Neurosurgical beds and neuro-intensive care beds were significantly decreased from December 2019 to March 2020. The utilization of resources decreased via less demand for care of brain injuries and subarachnoid hemorrhage, postponing surgery and changed surgical indications as a method of rationing resources. Twenty departments (80%) reduced activity extensively, and the same proportion stated that they were no longer able to provide care according to legitimate medical needs. CONCLUSION: Neurosurgical centers responded swiftly and effectively to a sudden decrease of neurosurgical capacity due to relocation of resources to pandemic care. The pandemic led to rationing of neurosurgical care in 80% of responding centers. We saw a relation between resources before the pandemic and ability to uphold neurosurgical services. The observation of extensive differences of available beds provided an opportunity to show how resources that had been restricted already under normal conditions translated to rationing of care that may not be acceptable to the public of seemingly affluent European countries.

Langerhans' cell histiocytosis presenting as intracranial hypertension.

Benign intracranial hypertension is known to be associated with obesity, endocrine abnormalities, various medications, and cerebral venous sinus thrombosis. We report a patient presenting with headaches and vomiting attributed to benign intracranial hypertension. The diagnostic work-up revealed Langerhans' cell histiocytosis of the occipital bone. There was no evidence for cerebral vein thrombosis by cranial computed tomography scan, Doppler ultrasonography, planar and single photon emission computed tomography technetium 99m-labelled red blood cell scintigraphy, and magnetic resonance angiography. Excision of the occipital bone lesion and a short course of acetazolamide and prednisone were curative. We hypothesize that cytokines secreted by the tumor were responsible for the development of intracranial hypertension.

Novel, compact, intraoperative magnetic resonance imaging-guided system for conventional neurosurgical operating rooms.

OBJECTIVE: Preliminary clinical experience with a novel, compact, intraoperative magnetic resonance imaging (MRI)-guided system that can be used in an ordinary operating room is presented. DESCRIPTION OF INSTRUMENTATION: The system features an MRI scanner integrated with an optical and MRI tracking system. Scanning and navigation, which are operated by the surgeon, are controlled by an in-room computer workstation with a liquid crystal display screen. The scanner includes a 0.12-T permanent magnet with a 25-cm vertical gap, accommodating the patient's head. The field of view is 11 x 16 cm, encompassing the surgical area of interest. The magnet is mounted on a transportable gantry that can be positioned under the surgical table when not in use for scanning, thus rendering the surgical environment unmodified and allowing the use of standard instruments. The features of the integrated navigation system allow flap planning and intraoperative tracking based on updated images acquired during surgery. OPERATIVE TECHNIQUE: Twenty patients with brain tumors were surgically treated using craniotomy or trans-sphenoidal approaches. One patient underwent conscious craniotomy with cortical mapping, and two underwent electrocorticography. EXPERIENCE AND RESULTS: Planning was accurate. Resection control images were obtained for all patients during surgery, with precise localization of residual tumor tissue. There were no surgical complications related to the use of the system. CONCLUSION: This intraoperative MRI system can function in a normal operating room modified only to eliminate radiofrequency interference. The operative environment is normal, and standard instruments can be used. The scanning and navigation capabilities of the system eliminate the inaccuracies that may result from brain shift. This novel type of intraoperative MRI system represents another step toward the introduction of the modality as a standard method in neurosurgery.

Anesthesia for magnetic resonance guided neurosurgery: initial experience with a new open magnetic resonance imaging system.

The authors present their initial experience with a compact open magnetic resonance (MR) image-guided system, (PoleStar N-10, Odin Medical Technologies, Yokneam, Israel) used in a standard operating room, modified for radio frequency (RF) shielding. The low intensity of the magnetic field (0.12T), and the ability to lower the magnet from the operative field during surgery allows for an almost routine surgical procedure, in addition to the benefits of using intraoperative MR imaging. Although an MR compatible anesthesia machine and monitoring system are used, the system offers anesthesiologists access to the patient at all times during the procedure, and the ability to use conventional surgical equipment, syringe pumps, and warming devices. Propofol and remifentanil, used for maintaining anesthesia, allow early extubation and neurological evaluation at the end of surgery. Electrocorticographic monitoring can be used during surgery for epilepsy, and awake craniotomy can be performed. More experience with this new imaging system is required to assess its influence on clinical decision making and outcome.

Metabolic and clinical markers of prognosis in the era of CT imaging in children with acute epidural hematomas.

Acute epidural hematoma (AEH), a relatively common complication of head injury in children, persists in bearing high morbidity and mortality. Early establishment of prognosis could guide optimal patient allocation, and early identification of predictive signs could assist in choosing appropriate therapeutic interventions. This study aimed to delineate expeditiously obtainable prognostic markers for determining outcome in a subset of children with AEH. We reviewed our 11-year experience with 61 consecutive children <16 years old with head trauma and isolated AEH. Treatment followed a standard advanced trauma life support protocol. A medical history was obtained, and all patients underwent neurosurgical and physical evaluations. CT scans were performed, as were laboratory tests which included arterial blood gases, glucose, electrolytes (K(+), Na(+)), hemoglobin and coagulation studies. Evaluation of the data collected on cause of injury, interval between trauma occurrence and presentation, clinical symptoms, Glasgow Coma Scale (GCS) scores, vital signs, laboratory test results, physical findings and surgical versus conservative management revealed that the best single predictors of outcome following AEH were the GCS and focal neurological deficits. Of all laboratory data obtained on admission, the blood potassium, pH and glucose test results correlated significantly with prognosis. Prognosis can be adequately and expeditiously estimated by selected markers within a comprehensive evaluation of children with AEH.

Validation of a controlled cortical impact model of head injury in mice.

A controlled cortical impact model of head injury was validated with mice. Mice were randomly assigned to moderate head injury, mild head injury, and sham injury groups. Beam balancing, open field activity, slant board inclination, grasp strength, and motor coordination were assessed prior to the injury and on days 1-5 postinjury. Morris water maze performance was evaluated on days 11-15 postinjury. Moderately head-injured mice took a significantly longer time to complete the motor coordination task and to find the hidden platform on the Morris water maze and had significantly fewer successful trials on both tasks than the mildly head-injured and sham-injured mice. Mildly head-injured and sham-injured mice performed similarly on both tasks. Contusion volume at the site of impact varied with severity of injury. Moderately head-injured mice had significantly larger contusions than mice with a mild head injury, and these mice in turn had significantly larger contusions than the sham-injured mice. Both moderately and mildly head injured mice had significantly fewer surviving cells in CA1 than the sham-injured mice but did not differ from each other in this regard. Although there was a group effect, only the mildly head-injured mice had significantly fewer surviving cells in CA3.

Gunshot wounds in brains of children: prognostic variables in mortality, course, and outcome.

A retrospective study of 51 children presenting with craniocerebral gunshot lesions was carried out to identify predictors of outcome. The patients ranged in age from 2 months to 17 years, with a mean of 14.5 years. The outcome was good in 20 patients, and seven and four were moderately and severely disabled, respectively. Twenty patients died. Statistical analysis showed prognostic significance of the admission Glasgow Coma Score (GCS), computerized tomographic findings of intraventricular hemorrhage and midline shift, and metabolic abnormalities, including hypokalemia and hyperglycemia. These prognostic factors may have implications regarding counseling of families, utilization of resources, and organ transplantation.

Neurologic outcome with hemorrhagic hypotension after closed head trauma in rats: effect of early versus delayed conservative fluid therapy.

OBJECTIVE: This study examined (1) whether two previously reported, well-established models in rats, one a model of hemorrhagic hypotension and the other a model of closed head trauma, could be combined to evaluate neurologic outcome when hemorrhage occurs subsequent to head injury, and (2) the ability of the traditional, conservative approach to fluid therapy (3 mL of intravenous fluid for 1 mL of blood loss) to reverse the detrimental effects of hemorrhagic hypotension after closed head trauma. In addition, two strategies of fluid therapy (early and delayed) were examined. METHODS: Fifty-six Sprague-Dawley male rats were divided into five groups with head injury at time 0 in groups 3 to 5, hemorrhage at 1 hour in groups 1, 2, 4, and 5, and intravenous fluid at 15 minutes (groups 2 and 5) or 60 minutes (groups 1 and 4) after hemorrhage. Head injury was delivered using a weight-drop impact of 0.5 J onto the closed cranium. Neurologic Severity Score (NSS) was determined at 1 hour (just before hemorrhage) and at 4 hours. RESULTS: NSS at 1 hour did not differ between groups 3 to 5 (15.5 (9-24) to 16 (2-21), median (range)). The amount of bleeding did not differ between groups during the first 15 minutes of hemorrhage (2.8 +/- 0.8 to 3.7 +/- 2.0 mL, mean +/- SD). After 60 minutes, cumulative blood loss in the delayed fluid therapy groups was less (3.1 +/- 1.13 mL in group 1 and 4.25 +/- 2.39 mL in group 4) than in the early fluid therapy groups (7.73 +/- 4.41 mL in group 2 and 6.85 +/- 2.36 mL in group 5) (analysis of variance, p < 0.01). The NSS of group 3 (head injury only) improved at 4 hours after injury (12 (5-20)), whereas the NSS of groups 4 and 5 (head injury followed by hemorrhage) deteriorated (24 (17-25) and 19.5 (9-25), respectively) (Kruskal-Wallis test,p < 0.05). In all the hemorrhage groups, fluid therapy failed to restore blood pressure to prehemorrhage levels. CONCLUSION: It is concluded that the two individual models of hemorrhagic hypotension and closed head trauma in rats can be combined to evaluate outcome when hemorrhage occurs subsequent to head injury. Furthermore, traditional, conservative fluid therapy, whether early or delayed, failed to restore blood pressure or to improve NSS when hemorrhage occurred after head injury. Blood loss was greater with early fluid therapy whether or not head injury was present.

Positive end expiratory pressure reduces intracranial compliance in the rabbit.

Acute respiratory distress syndrome is commonly encountered in head-injured patients. Positive and expiratory pressure (PEEP) is useful in improving oxygenation. However, PEEP, by increasing intrathoracic pressure, decreases venous return, mean arterial pressure, and cardiac output and increases jugular vein pressure. There is conflicting evidence in the literature as to the potential effect of PEEP on intracranial pressure (ICP). The present study was undertaken to examine the effect of PEEP on ICP and intracranial compliance. Twelve male rabbits weighing 3.5-4.5 kg were used. The following parameters were monitored: arterial blood pressure, ICP (intraparenchymal Camino device), PaCO2, and PaO2. A space-occupying lesion was produced by inflation of a double lumen pediatric Swan-Ganz catheter placed over the right parietal dura. The amount of fluid required to reach the point of exponential increase of ICP was recorded at PEEP of 0 and 10 cm H2O. The mean volume needed to reach the deflection point of ICP was significantly lower when PEEP was 10 cm H2O compared to the value when PEEP was 0 cm H2O (685 +/- 48 vs. 883.3 +/- 46 microliters, respectively; p < 0.01). The results of the present study indicated that PEEP of 10 cm H2O decreases intracranial compensatory reserves for maintaining ICP at normal levels in the presence of an expanding intracranial mass.

Influence of closed head injury on isoflurane MAC in the rat.

We designed the present study to determine whether the minimum alveolar concentration (MAC) for isoflurane is decreased after closed head trauma (CHT) in rats and, if so, whether the decrease of MAC is related to the severity of neurological impairment following CHT. Isoflurane MAC was determined in 36 Sprague-Dawley rats. Then, at time = 0 h, animals were grouped. Group 1 (n = 8) received no CHT, group 2 (n = 14) received moderate CHT, and group 3 (n = 14) received severe CHT. Neurological severity score (NSS, 0 = no deficit and 25 = maximal impairment) and MAC were determined at 1, 4, 24, and 48 h. In groups 1 and 2, isoflurane MAC at 1, 2, 24, and 48 h (1.0-1.1 +/- 0.8-1.2%, median +/- range) was not significantly different from baseline (1.0-1.1 +/- 1.0-1.1%). In group 3, isoflurane MAC at 1, 2, 24, and 48 h (0.4 +/- 0.2-0.5%) was decreased as compared to baseline (1.1 +/- 1.0-1.1%). In group 2, NSS at 1 h was 18 +/- 11-21 and improved by 48 h to 9 +/- 4-15. In group 3, NSS at 1 h was 24 +/- 22-25 and was not significantly different from NSS at 48 h (24 +/- 24-25). Thus, moderate CHT does not significantly alter isoflurane MAC, whereas severe CHT equivalent to a Glasgow Coma Scale score of 3 to 6 significantly decreases isoflurane MAC.

Monitoring of cerebral hemodynamics with jugular bulb catheters.

Jugular venous oxygen saturation (SjvO2) monitoring is useful for detecting episodes of cerebral hypoxia/ischemia in patients with head injury, patients undergoing neurosurgical procedures, and patients undergoing cardiopulmonary bypass. The use of SjvO2 monitoring can direct the treatment of ischemic episodes and identify the optimal level of cerebral perfusion pressure and PCO2 for the individual patient.

Effect of magnesium given 1 hour after head trauma on brain edema and neurological outcome.

Excitatory amino acids (EAA), mainly glutamate and aspartate, are released in excessive amounts from terminals of ischemic or traumatically injured neurons. These excessive levels of EAAs initiate a cascade of events believed to lead to secondary delayed damage to the surrounding brain. The N-methyl-D-aspartate receptor antagonists MK-801 and ketamine are reported to suppress excessive EAA release and to attenuate the development of focal brain edema following neuronal injury. Magnesium is also reported to work at the postsynaptic receptor to reduce the neurotoxic effect of glutamate. The present study was undertaken to examine the effect of postinjury treatment with Mg++ on brain edema and neurological outcome after traumatic brain injury. Sixty-nine rats that survived halothane anesthesia and closed head trauma (CHT) were randomly assigned to one of seven experimental groups: sham, CHT, and CHT with administration of Mg++ 1 hour postinjury. At 48 hours, brain tissue Mg++ concentration (calculated from optical density using a standard curve) was significantly increased compared to baseline levels (10.06 +/- 2.44 mg/g vs. 6.83 +/- 0.81 mg/g, p < 0.01 calculated by one-way analysis of variance). Also at 48 hours postinjury, brain tissue specific gravity in the contused hemisphere of Mg(++)-treated rats was significantly greater than that in the contused hemisphere of untreated rats, indicating attenuation of brain edema formation by Mg++. The neurological severity score (NSS) of rats treated with Mg++ improved significantly at both 18 and 48 hours, compared to baseline values obtained 1 hour after CHT but prior to administration of Mg++ (11.2 +/- 2.5 vs. 15.2 +/- 4.1, p = 0.03; and 12.3 +/- 6.1 vs. 17.3 +/- 3.6, p = 0.004, respectively). In the untreated groups, the NSS at 18 and 48 hours was not significantly different from baseline values (that is, no neurological improvement). The present study indicates that postinjury treatment with Mg++ attenuates brain edema formation and improves neurological outcome after experimental CHT.

Brain edema and neurological status with rapid infusion of lactated Ringer's or 5% dextrose solution following head trauma.

Rapid infusion of 0.25 ml/g of 0.9% saline over 30 minutes has been shown to have no effect on electrolyte balance, neurological severity score (NSS), or brain edema, following closed head trauma (CHT). Rapid infusion of the same volume of 5% dextrose solution decreased blood sodium concentration, increased edema, and decreased NSS following CHT. In the present study the authors examined the effect of rapid infusion (30 minutes) of smaller volumes of 5% dextrose (0.08 ml/g and 0.16 ml/g) and of 0.25 ml/g lactated Ringer's solution on blood electrolyte concentrations, plasma osmolality, brain edema, and NSS. The purposes of this study were to determine whether rapid infusion of a large volume of lactated Ringer's solution could be given after CHT without increasing mortality or brain edema or producing electrolyte disturbances, and whether small volumes of 5% dextrose could be infused with few or none of the deleterious effects expected from large volumes of 5% dextrose. One hundred eighteen rats, which survived halothane anesthesia and CHT, were randomly assigned to one of 15 experimental groups. Fluids were administered beginning 1 hour after scalp incision or CHT. The NSS, extent of edema, blood electrolyte concentrations, and plasma osmolality in the groups treated with lactated Ringer's solution were not significantly different from those values in the nontreated groups. In addition, the mortality rate after CHT was not increased by administration of lactated Ringer's solution. The groups treated with 5% dextrose solution showed a significantly higher mortality rate, but the NSSs of the surviving rats were not different from controls. None of the groups treated with 0.16 ml/g 5% dextrose solution survived 24 hours. Although blood glucose concentration increased to 1126 +/- 102 g% (mean +/- standard deviation) and 1568 +/- 283 g% and blood sodium concentration decreased to 110.4 +/- 4.6 mEq/L and 92.0 +/- 5.2 mEq/L in the groups treated with 0.08 ml/g and 0.16 ml/g of 5% dextrose solution, respectively, plasma osmolality was normal and no significant difference could be found between the brain tissue specific gravity of animals in the nontreated and 5% dextrose treatment groups. It is concluded that in the CHT model used in this study, the large volume of lactated Ringer's solution did not affect blood electrolyte concentration, neurological outcome, or formation of brain edema, whereas smaller volumes of 5% dextrose solution increased blood glucose and decreased blood sodium concentrations, did not affect plasma osmolality, and had a deleterious effect on neurological outcome.