Hyperspectral imaging techniques (HSI) do not require contact with patients and are non-ionizing as well as non-invasive. As a consequence, they have been extensively applied in the medical field. HSI is being combined with machine learning (ML) processes to obtain models to assist in diagnosis. In particular, the combination of these techniques has proven to be a reliable aid in the differentiation of healthy and tumor tissue during brain tumor surgery. ML algorithms such as support vector machine (SVM), random forest (RF) and convolutional neural networks (CNN) are used to make predictions and provide in-vivo visualizations that may assist neurosurgeons in being more precise, hence reducing damages to healthy tissue. In this work, thirteen in-vivo hyperspectral images from twelve different patients with high-grade gliomas (grade III and IV) have been selected to train SVM, RF and CNN classifiers. Five different classes have been defined during the experiments: healthy tissue, tumor, venous blood vessel, arterial blood vessel and dura mater. Overall accuracy ($OACC$) results vary from 60% to 95% depending on the training conditions. Finally, as far as the contribution of each band to the $OACC$ is concerned, the results obtained in this work are 3.81 times greater than those reported in the literature.

Object: The aim of this study was to determine the incidence of posttraumatic hydrocephalus in severely head-injured patients who required decompressive craniectomy (DC). Additional objectives were to determine the relationship between hydrocephalus and several clinical and radiological features, with special attention to subdural hygromas as a sign of distortion of the CSF circulation. Methods: The authors conducted a retrospective study of 73 patients with severe head injury who required DC. The patients were admitted to the authors' department between January 2000 and January 2006. Posttraumatic hydrocephalus was defined as: 1) modified frontal horn index greater than 33%, and 2) the presence of Gudeman CT criteria. Hygromas were diagnosed based on subdural fluid collection and classified according to location of the craniectomy. Results: Hydrocephalus was diagnosed in 20 patients (27.4%). After uni- and multivariate analysis, the presence of interhemispheric hygromas (IHHs) was the only independent prognostic factor for development of posttraumatic hydrocephalus (p < 0.0001). More than 80% of patients with IHHs developed hydrocephalus within the first 50 days of undergoing DC. In all cases the presence of hygromas preceded the diagnosis of hydrocephalus. The IHH predicts the development of hydrocephalus after DC with 94% sensitivity and 96% specificity. The presence of an IHH showed an area under the receiver-operator characteristic of 0.951 (95% CI 0.87–1.00; p < 0.0001). Conclusions: Hydrocephalus was observed in 27.4% of the patients with severe traumatic brain injury who required DC. The presence of IHHs was a predictive radiological sign of hydrocephalus development within the first 6 months of DC in patients with severe head injury.

Object: There were two main purposes to this study: first, to assess the feasibility and reliability of 2 quantitative methods to assess bleeding volume in patients who suffered spontaneous subarachnoid hemorrhage (SAH), and second, to compare these methods to other qualitative and semiquantitative scales in terms of reliability and accuracy in predicting delayed cerebral ischemia (DCI) and outcome. Methods: A prospective series of 150 patients consecutively admitted to the Hospital 12 de Octubre over a 4-year period were included in the study. All of these patients had a diagnosis of SAH, and diagnostic CT was able to be performed in the first 24 hours after the onset of the symptoms. All CT scans were evaluated by 2 independent observers in a blinded fashion, using 2 different quantitative methods to estimate the aneurysmal bleeding volume: region of interest (ROI) volume and the Cavalieri method. The images were also graded using the Fisher scale, modified Fisher scale, Claasen scale, and the semiquantitative Hijdra scale. Weighted κ coefficients were calculated for assessing the interobserver reliability of qualitative scales and the Hijdra scores. For assessing the intermethod and interrater reliability of volumetric measurements, intraclass correlation coefficients (ICCs) were used as well as the methodology proposed by Bland and Altman. Finally, weighted κ coefficients were calculated for the different quartiles of the volumetric measurements to make comparison with qualitative scales easier. Patients surviving more than 48 hours were included in the analysis of DCI predisposing factors and analyzed using the chi-square or the Mann-Whitney U-tests. Logistic regression analysis was used for predicting DCI and outcome in the different quartiles of bleeding volume to obtain adjusted ORs. The diagnostic accuracy of each scale was obtained by calculating the area under the receiver operating characteristic curve (AUC). Results: Qualitative scores showed a moderate interobserver reproducibility (weighted κ indexes were always < 0.65), whereas the semiquantitative and quantitative scores had a very strong interobserver reproducibility. Reliability was very high for all quantitative measures as expressed by the ICCs for intermethod and interobserver agreement. Poor outcome and DCI occurred in 49% and 31% of patients, respectively. Larger bleeding volumes were related to a poorer outcome and a higher risk of developing DCI, and the proportion of patients suffering DCI or a poor outcome increased with each quartile, maintaining this relationship after adjusting for the main clinical factors related to outcome. Quantitative analysis of total bleeding volume achieved the highest AUC, and had a greater discriminative ability than the qualitative scales for predicting the development of DCI and outcome. Conclusions: The use of quantitative measures may reduce interobserver variability in comparison with categorical scales. These measures are feasible using dedicated software and show a better prognostic capability in relation to outcome and DCI than conventional categorical scales.

Quantitative estimation of the hemorrhage volume associated with aneurysm rupture is a new tool of assessing prognosis. To determine the prognostic value of the quantitative estimation of the amount of bleeding after aneurysmal subarachnoid hemorrhage, as well the relative importance of this factor related to other prognostic indicators, and to establish a possible cut-off value of volume of bleeding related to poor outcome. A prospective cohort of 206 patients consecutively admitted with the diagnosis of aneurysmal subarachnoid hemorrhage to Hospital 12 de Octubre were included in the study. Subarachnoid, intraventricular, intracerebral, and total bleeding volumes were calculated using analytic software. For assessing factors related to prognosis, univariate and multivariate analysis (logistic regression) were performed. The relative importance of factors in determining prognosis was established by calculating their proportion of explained variation. Maximum Youden index was calculated to determine the optimal cut point for subarachnoid and total bleeding volume. Variables independently related to prognosis were clinical grade at admission, age, and the different bleeding volumes. The proportion of variance explained is higher for subarachnoid bleeding. The optimal cut point related to poor prognosis is a volume of 20 mL both for subarachnoid and total bleeding. Volumetric measurement of subarachnoid or total bleeding volume are both independent prognostic factors in patients with aneurysmal subarachnoid hemorrhage. A volume of more than 20 mL of blood in the initial noncontrast computed tomography is related to a clear increase in poor outcome risk. : aSAH, aneurysmal subarachnoid hemorrhage.

Object: Traumatic brain injury (TBI) represents a large health and economic burden. Because of the inability of previous randomized controlled trials (RCTs) on TBI to demonstrate the expected benefit of reducing unfavorable outcomes, the IMPACT (International Mission on Prognosis and Analysis of Clinical Trials in TBI) and CRASH (Corticosteroid Randomisation After Significant Head Injury) studies provided new methods for performing prognostic studies of TBI. This study aimed to develop and externally validate a prognostic model for early death (within 48 hours). The secondary aim was to identify patients who were more likely to succumb to an early death to limit their inclusion in RCTs and to improve the efficiency of RCTs. Methods: The derivation cohort was recruited at 1 center, Hospital 12 de Octubre, Madrid (1990–2003, 925 patients). The validation cohort was recruited in 2004–2006 from 7 study centers (374 patients). The eligible patients had suffered closed severe TBIs. The study outcome was early death (within 48 hours post-TBI). The predictors were selected using logistic regression modeling with bootstrapping techniques, and a penalized reduction was used. A risk score was developed based on the regression coefficients of the variables included in the final model. Results: In the validation set, the final model showed a predictive ability of 50% (Nagelkerke R²), with an area under the receiver operating characteristic curve of 89% and an acceptable calibration (goodness-of-fit test, p = 0.32). The final model included 7 variables, and it was used to develop a risk score with a range from 0 to 20 points. Age provided 0, 1, 2, or 3 points depending on the age group; motor score provided 0 points, 2 (untestable), or 3 (no response); pupillary reactivity, 0, 2 (1 pupil reacted), or 6 (no pupil reacted); shock, 0 (no) or 2 (yes); subarachnoid hemorrhage, 0 or 1 (severe deposit); cisternal status, 0 or 3 (compressed/absent); and epidural hematoma, 0 (yes) or 2 (no). Based on the risk of early death estimated with the model, 4 risk of early death groups were established: low risk, sum score 0–3 (< 1% predicted mortality); moderate risk, sum score 4–8 (predicted mortality between 1% and 10%); high risk, sum score 9–12 (probability of early death between 10% and 50%); and very high risk, sum score 13–20 (early mortality probability > 50%). This score could be used for selecting patients for clinical studies. For example, if patients with very high risk scores were excluded from our study sample, the patients included (eligibility score < 13) would represent 80% of the original sample and only 23% of the patients who died early. Conclusions: The combination of Glasgow Coma Scale score, CT scanning results, and secondary insult data into a prognostic score improved the prediction of early death and the classification of TBI patients.