Identifying clusters of objective functional impairment in patients with degenerative lumbar spinal disease using unsupervised learning

OBJECTIVES

The five-repetition sit-to-stand (5R-STS) test was designed to capture objective functional impairment (OFI), and thus provides an adjunctive dimension in patient assessment. It is conceivable that there are different subsets of patients with OFI and degenerative lumbar disease. We aim to identify clusters of objectively functionally impaired individuals based on 5R-STS and unsupervised machine learning (ML).

METHODS

Data from two prospective cohort studies on patients with surgery for degenerative lumbar disease and 5R-STS times of ≥ 10.5 s-indicating presence of OFI. K-means clustering-an unsupervised ML algorithm-was applied to identify clusters of OFI. Cluster hallmarks were then identified using descriptive and inferential statistical analyses.

RESULTS

We included 173 patients (mean age [standard deviation]: 46.7 [12.7] years, 45% male) and identified three types of OFI. OFI Type 1 (57 pts., 32.9%), Type 2 (81 pts., 46.8%), and Type 3 (35 pts., 20.2%) exhibited mean 5R-STS test times of 14.0 (3.2), 14.5 (3.3), and 27.1 (4.4) seconds, respectively. The grades of OFI according to the validated baseline severity stratification of the 5R-STS increased significantly with each OFI type, as did extreme anxiety and depression symptoms, issues with mobility and daily activities. Types 1 and 2 are characterized by mild to moderate OFI-with female gender, lower body mass index, and less smokers as Type I hallmarks.

CONCLUSIONS

Unsupervised learning techniques identified three distinct clusters of patients with OFI that may represent a more holistic clinical classification of patients with OFI than test-time stratifications alone, by accounting for individual patient characteristics.

Predictive value of the five-repetition sit-to-stand test for outcomes after surgery for lumbar disc herniation: prospective study

OBJECTIVE

It is unknown whether presence of pre-operative objective functional impairment (OFI) can predict post-operative outcomes in patients with lumbar disc herniation (LDH). We aimed to determine whether pre-operative OFI measured by the five-repetition sit-to-stand test (5R-STS) could predict outcomes at 12-months post-discectomy.

METHODS

Adult patients with LDH scheduled for surgery were prospectively recruited from a Dutch short-stay spinal clinic. The 5R-STS time and patient reported outcome measures (PROMs) including Oswestry Disability Index, Roland-Morris Disability Questionnaire, Visual Analogue Scale (VAS) for back and leg pain, EQ-5D-3L health-related quality of life, EQ5D-VAS and ability to work were recorded pre-operatively and at 12-months. A 5R-STS time cut-off of ≥ 10.5 s was used to determine OFI. Mann-Whitney and Chi-square tests were employed to determine significant differences in post-operative outcomes between groups stratified by presence of pre-operative OFI.

RESULTS

We recruited 134 patients in a prospective study. Twelve-month follow-up was completed by 103 (76.8%) patients. Mean age was 53.2 ± 14.35 years and 50 (48.5%) patients were female. Pre-operatively, 53 (51.5%) patients had OFI and 50 (48.5%) did not. Post-operatively, patients with OFI experienced a significantly greater mean change (p < 0.001) across all PROMs compared to patients without OFI, except leg pain (p = 0.176). There were no significant differences in absolute PROMs between groups at 12-months (all p > 0.05).

CONCLUSIONS

The presence of OFI based on 5R-STS time does not appear to decrease a patient's likelihood of experiencing satisfactory post-operative outcomes. The 5R-STS cannot predict how a patient with LDH will respond to surgery at 12-month follow-up.

Whole Spine Segmentation Using Object Detection and Semantic Segmentation

OBJECTIVE

Virtual and augmented reality have enjoyed increased attention in spine surgery. Preoperative planning, pedicle screw placement, and surgical training are among the most studied use cases. Identifying osseous structures is a key aspect of navigating a 3-dimensional virtual reconstruction. To automate the otherwise time-consuming process of labeling vertebrae on each slice individually, we propose a fully automated pipeline that automates segmentation on computed tomography (CT) and which can form the basis for further virtual or augmented reality application and radiomic analysis.

METHODS

Based on a large public dataset of annotated vertebral CT scans, we first trained a YOLOv8m (You-Only-Look-Once algorithm, Version 8 and size medium) to detect each vertebra individually. On the then cropped images, a 2D-U-Net was developed and externally validated on 2 different public datasets.

RESULTS

Two hundred fourteen CT scans (cervical, thoracic, or lumbar spine) were used for model training, and 40 scans were used for external validation. Vertebra recognition achieved a mAP50 (mean average precision with Jaccard threshold of 0.5) of over 0.84, and the segmentation algorithm attained a mean Dice score of 0.75 ± 0.14 at internal, 0.77 ± 0.12 and 0.82 ± 0.14 at external validation, respectively.

CONCLUSION

We propose a 2-stage approach consisting of single vertebra labeling by an object detection algorithm followed by semantic segmentation. In our externally validated pilot study, we demonstrate robust performance for our object detection network in identifying individual vertebrae, as well as for our segmentation model in precisely delineating the bony structures.

TomoRay: Generating Synthetic Computed Tomography of the Spine From Biplanar Radiographs

OBJECTIVE

Computed tomography (CT) imaging is a cornerstone in the assessment of patients with spinal trauma and in the planning of spinal interventions. However, CT studies are associated with logistical problems, acquisition costs, and radiation exposure. In this proof-of-concept study, the feasibility of generating synthetic spinal CT images using biplanar radiographs was explored. This could expand the potential applications of x-ray machines pre-, post-, and even intraoperatively.

METHODS

A cohort of 209 patients who underwent spinal CT imaging from the VerSe2020 dataset was used to train the algorithm. The model was subsequently evaluated using an internal and external validation set containing 55 from the VerSe2020 dataset and a subset of 56 images from the CTSpine1K dataset, respectively. Digitally reconstructed radiographs served as input for training and evaluation of the 2-dimensional (2D)-to-3-dimentional (3D) generative adversarial model. Model performance was assessed using peak signal to noise ratio (PSNR), structural similarity index (SSIM), and cosine similarity (CS).

RESULTS

At external validation, the developed model achieved a PSNR of 21.139 ± 1.018 dB (mean ± standard deviation). The SSIM and CS amounted to 0.947 ± 0.010 and 0.671 ± 0.691, respectively.

CONCLUSION

Generating an artificial 3D output from 2D imaging is challenging, especially for spinal imaging, where x-rays are known to deliver insufficient information frequently. Although the synthetic CT scans derived from our model do not perfectly match their ground truth CT, our proof-of-concept study warrants further exploration of the potential of this technology.

Safety and Effectiveness of an Enhanced Recovery Protocol in Patients Undergoing Burr Hole Evacuation for Chronic Subdural Hematoma

BACKGROUND AND OBJECTIVES

Enhanced recovery programs may be especially useful in patients with chronic subdural hematoma or hygroma (cSDH), who frequently exhibit frailty and multimorbidity. We aim to evaluate the real-world safety and effectiveness of an enhanced recovery protocol in this population.

METHODS

From a prospective registry, burr hole evacuations for cSDH carried out under the protocol (including early thromboprophylaxis, no flat bed rest, early mobilization without drain clamping, and early resumption of antithrombotic medication) were extracted, along with those procedures carried out within the past year before protocol change. Propensity score-based matching was carried out. A range of clinical and imaging outcomes were analyzed, including modified Rankin Scale as effectiveness and Clavien-Dindo adverse event grading as safety primary end points.

RESULTS

Per group, 91 procedures were analyzed. At discharge, there was no significant difference in the modified Rankin Scale among the standard and enhanced recovery groups (1 [1; 2] vs 1 [1; 3], P = .552), or in Clavien-Dindo adverse event grading classifications of adverse events (P = .282) or occurrence of any adverse events (15.4% vs 20.9%, P = .442). There were no significant differences in time to drain removal (2.00 [2.00; 2.00] vs 2.00 [1.25; 2.00] days, P = .058), time from procedure to discharge (4.0 [3.0; 6.0] vs 4.0 [3.0; 6.0] days, P = .201), or total hospital length of stay (6.0 [5.0; 9.0] vs 5.0 [4.0; 8.0] days, P = .113). All-cause mortality was similar in both groups (8.8% vs 4.4%, P = .289), as was discharge disposition (P = .192). Other clinical and imaging outcomes were similar too (all P > .05).

CONCLUSION

In a matched cohort study comparing perioperative standard of care with a novel enhanced recovery protocol focusing on evidence-based drainage, mobilization, and thromboprophylaxis regimens as well as changes to the standardized reuptake of oral anticoagulants and antiaggregants, no differences in safety or effectiveness were observed after burr hole evacuation of cSDH.

From molecular signatures to radiomics: tailoring neurooncological strategies through forecasting of glioma growth

OBJECTIVE

Contemporary oncological paradigms for adjuvant treatment of low- and intermediate-grade gliomas are often guided by a limited array of parameters, overlooking the dynamic nature of the disease. The authors' aim was to develop a comprehensive multivariate glioma growth model based on multicentric data, to facilitate more individualized therapeutic strategies.

METHODS

Random slope models with subject-specific random intercepts were fitted to a retrospective cohort of grade II and III gliomas from the database at Kepler University Hospital (n = 191) to predict future mean tumor diameters. Deep learning-based radiomics was used together with a comprehensive clinical dataset and evaluated on an external prospectively collected validation cohort from University Hospital Zurich (n = 9). Prediction quality was assessed via mean squared prediction error.

RESULTS

A mean squared prediction error of 0.58 cm for the external validation cohort was achieved, indicating very good prognostic value. The mean ± SD time to adjuvant therapy was 28.7 ± 43.3 months and 16.1 ± 14.6 months for the training and validation cohort, respectively, with a mean of 6.2 ± 5 and 3.6 ± 0.7, respectively, for number of observations. The observed mean tumor diameter per year was 0.38 cm (95% CI 0.25-0.51) for the training cohort, and 1.02 cm (95% CI 0.78-2.82) for the validation cohort. Glioma of the superior frontal gyrus showed a higher rate of tumor growth than insular glioma. Oligodendroglioma showed less pronounced growth, anaplastic astrocytoma-unlike anaplastic oligodendroglioma-was associated with faster tumor growth. Unlike the impact of extent of resection, isocitrate dehydrogenase (IDH) had negligible influence on tumor growth. Inclusion of radiomics variables significantly enhanced the prediction performance of the random slope model used.

CONCLUSIONS

The authors developed an advanced statistical model to predict tumor volumes both pre- and postoperatively, using comprehensive data prior to the initiation of adjuvant therapy. Using radiomics enhanced the precision of the prediction models. Whereas tumor extent of resection and topology emerged as influential factors in tumor growth, the IDH status did not. This study emphasizes the imperative of advanced computational methods in refining personalized low-grade glioma treatment, advocating a move beyond traditional paradigms.

A critical moment in machine learning in medicine: on reproducible and interpretable learning

Over the past two decades, advances in computational power and data availability combined with increased accessibility to pre-trained models have led to an exponential rise in machine learning (ML) publications. While ML may have the potential to transform healthcare, this sharp increase in ML research output without focus on methodological rigor and standard reporting guidelines has fueled a reproducibility crisis. In addition, the rapidly growing complexity of these models compromises their interpretability, which currently impedes their successful and widespread clinical adoption. In medicine, where failure of such models may have severe implications for patients' health, the high requirements for accuracy, robustness, and interpretability confront ML researchers with a unique set of challenges. In this review, we discuss the semantics of reproducibility and interpretability, as well as related issues and challenges, and outline possible solutions to counteracting the "black box". To foster reproducibility, standard reporting guidelines need to be further developed and data or code sharing encouraged. Editors and reviewers may equally play a critical role by establishing high methodological standards and thus preventing the dissemination of low-quality ML publications. To foster interpretable learning, the use of simpler models more suitable for medical data can inform the clinician how results are generated based on input data. Model-agnostic explanation tools, sensitivity analysis, and hidden layer representations constitute further promising approaches to increase interpretability. Balancing model performance and interpretability are important to ensure clinical applicability. We have now reached a critical moment for ML in medicine, where addressing these issues and implementing appropriate solutions will be vital for the future evolution of the field.

Microsurgical management of midbrain gliomas: surgical results and long-term outcome in a large, single-surgeon, consecutive series

OBJECTIVE

The authors report on a large, consecutive, single-surgeon series of patients undergoing microsurgical removal of midbrain gliomas. Emphasis is put on surgical indications, technique, and results as well as long-term oncological follow-up.

METHODS

A retrospective analysis was performed of prospectively collected data from a consecutive series of patients undergoing microneurosurgery for midbrain gliomas from March 2006 through June 2022 at the authors' institution. According to the growth pattern and location of the lesion in the midbrain (tegmentum, central mesencephalic structures, and tectum), one of the following approaches was chosen: transsylvian (TS), extreme anterior interhemispheric transcallosal (eAIT), posterior interhemispheric transtentorial subsplenial (PITS), paramedian supracerebellar transtentorial (PST), perimedian supracerebellar (PeS), perimedian contralateral supracerebellar (PeCS), and transuvulotonsillar fissure (TUTF). Clinical and radiological data were gathered according to a standard protocol and reported according to common descriptive statistics. The main outcomes were rate of gross-total resection; extent of resection; occurrence of any complications; variation in Karnofsky Performance Status score at discharge, 3 months, and last follow-up; progression-free survival (PFS); and overall survival (OS).

RESULTS

Fifty-four patients (28 of them pediatric) met the inclusion criteria (6 with high-grade and 48 with low-grade gliomas [LGGs]). Twenty-two tumors were in the tegmentum, 7 in the central mesencephalic structures, and 25 in the tectum. In no instance did the glioma originate in the cerebral peduncle. TS was performed in 2 patients, eAIT in 6, PITS in 23, PST in 16, PeS in 4, PeCS in 1, and TUTF in 2 patients. Gross-total resection was achieved in 39 patients (72%). The average extent of resection was 98.0% (median 100%, range 82%-100%). There were no deaths due to surgery. Nine patients experienced transient and 2 patients experienced permanent new neurological deficits. At a mean follow-up of 72 months (median 62, range 3-193 months), 49 of the 54 patients were still alive. All patients with LGGs (48/54) were alive with no decrease in their KPS score, whereas 42 showed improvement compared with their preoperative status.

CONCLUSIONS

Microneurosurgical removal of midbrain gliomas is feasible with good surgical results and long-term clinical outcomes, particularly in patients with LGGs. As such, microneurosurgery should be considered as the first therapeutic option. Adequate microsurgical technique and anesthesiological management, along with an accurate preoperative understanding of the tumor's exact topographic origin and growth pattern, is crucial for a good surgical outcome.

Automated volumetric assessment of pituitary adenoma

PURPOSE

Assessment of pituitary adenoma (PA) volume and extent of resection (EOR) through manual segmentation is time-consuming and likely suffers from poor interrater agreement, especially postoperatively. Automated tumor segmentation and volumetry by use of deep learning techniques may provide more objective and quick volumetry.

METHODS

We developed an automated volumetry pipeline for pituitary adenoma. Preoperative and three-month postoperative T1-weighted, contrast-enhanced magnetic resonance imaging (MRI) with manual segmentations were used for model training. After adequate preprocessing, an ensemble of convolutional neural networks (CNNs) was trained and validated for preoperative and postoperative automated segmentation of tumor tissue. Generalization was evaluated on a separate holdout set.

RESULTS

In total, 193 image sets were used for training and 20 were held out for validation. At validation using the holdout set, our models (preoperative / postoperative) demonstrated a median Dice score of 0.71 (0.27) / 0 (0), a mean Jaccard score of 0.53 ± 0.21/0.030 ± 0.085 and a mean 95th percentile Hausdorff distance of 3.89 ± 1.96./12.199 ± 6.684. Pearson's correlation coefficient for volume correlation was 0.85 / 0.22 and -0.14 for extent of resection. Gross total resection was detected with a sensitivity of 66.67% and specificity of 36.36%.

CONCLUSIONS

Our volumetry pipeline demonstrated its ability to accurately segment pituitary adenomas. This is highly valuable for lesion detection and evaluation of progression of pituitary incidentalomas. Postoperatively, however, objective and precise detection of residual tumor remains less successful. Larger datasets, more diverse data, and more elaborate modeling could potentially improve performance.

Promotion of a neurosurgical academic journal on social media: a 1-year experience

BACKGROUND

Social media (SoMe) use, in all of its forms, has seen massively increased throughout the past two decades, including academic publishing. Many journals have established a SoMe presence, yet the influence of promotion of scientific publications on their visibility and impact remains poorly studied. The European Journal of Neurosurgery «Acta Neurochirurgica» has established its SoMe presence in form of a Twitter account that regularly promotes its publications. We aim to analyze the impact of this initial SoMe campaign on various alternative metrics (altmetrics).

METHODS

A retrospective analysis of all articles published in the journal Acta Neurochirurgica between May 1st, 2018, and April 30th, 2020, was performed. These articles were divided into a historical control group - containing the articles published between May 1st, 2018, and April 30th, 2019, when the SoMe campaign was not yet established - and into an intervention group. Several altmetrics were analyzed, along with website visits and PDF downloads per month.

RESULTS

In total, 784 articles published during the study period, 128 (16.3%) were promoted via Twitter. During the promotion period, 29.7% of published articles were promoted. Overall, the published articles reached a mean of 31.3 ± 50.5 website visits and 17.5 ± 31.25 PDF downloads per month. Comparing the two study periods, no statistically significant differences in website visits (26.91 ± 32.87 vs. 34.90 ± 61.08, p = 0.189) and PDF downloads (17.52 ± 31.25 vs. 15.33 ± 16.07, p = 0.276) were detected. However, overall compared to non-promoted articles, promoted articles were visited (48.9 ± 95.0 vs. 29.0 ± 37.0, p = 0.005) and downloaded significantly more (25.7 ± 66.7 vs. 16.6 ± 18.0, p = 0.045) when compared to those who were not promoted during the promotion period.

CONCLUSIONS

We report a 1-year initial experience with promotion of a general neurosurgical journal on Twitter. Our data suggest a clear benefit of promotion on article site visits and article downloads, although no single responsible element could be determined in terms of altmetrics. The impact of SoMe promotion on other metrics, including traditional bibliometrics such as citations and journal impact factor, remains to be determined.

How we do it: the Zurich Microsurgery Lab technique for placenta preparation

BACKGROUND

Perfused placentas provide an excellent and accessible model for microvascular dissection, microsuturing and microanastomosis training - particularly in the early microsurgical learning curve. This way, a significant amount of live animals can be spared.

METHOD

We present the Zurich Microsurgery Lab protocol, detailing steps for obtaining, selecting, cleaning, flushing, cannulating, and preserving human placentas - as well as microsurgical training examples - in a tried-and-true, safe, cost-effective, and high-yield fashion.

CONCLUSION

Our technique enables highly realistic microsurgical training (microdissection, microvascular repair, microanastomosis) based on readily available materials. Proper handling, preparation, and preservation of the perfused placenta models is key.

Early versus delayed mobilization after aneurysmal subarachnoid hemorrhage: a systematic review and meta-analysis of efficacy and safety

OBJECTIVE

A central tenet of Enhanced Recovery After Surgery (ERAS) is evidence-based medicine. Survivors of aneurysmal subarachnoid hemorrhage (aSAH) constitute a fragile patient population prone to prolonged hospitalization within neurointensive care units (NICUs), prolonged immobilization, and a range of nosocomial adverse events. Potentially, well-monitored early mobilization (EM) could constitute a beneficial element of ERAS protocols in this population. Therefore, the objective was to summarize the available evidence on EM strategies in patients with aSAH.

METHODS

The authors retrieved prospective and retrospective studies that reported efficacy or safety data on EM (defined as EM in the NICU starting ≤ 7 days after ictus) versus delayed mobilization (DM) (any strategy that comparatively delayed mobilization) after aSAH and were published after January 1, 2000, in PubMed/MEDLINE, Embase, and the Cochrane Library. Random-effects meta-analysis was performed.

RESULTS

Ten studies analyzing 1292 patients were included for quantitative synthesis, including 1 randomized, 1 prospective nonrandomized, and 8 retrospective studies. Modified Rankin Scale scores at discharge were not different between the EM and DM groups (mean difference [MD] [95% CI] -0.86 [-2.93 to 1.20] points, p = 0.41). Hospital length of stay in days was markedly reduced in the EM group (MD [95% CI] -6.56 [-10.64 to -2.47] days, p = 0.002). Although there was a statistically significant reduction in radiological vasospasms (OR [95% CI] 0.65 [0.44-0.97], p = 0.03), the reduction in clinically relevant vasospasms was nonsignificant (OR [95% CI] 0.63 [0.31-1.26], p = 0.19). The odds of shunting were significantly lower in the EM group (OR [95% CI] 0.61 [0.39-0.95], p = 0.03). The rates of mortality, pneumonia, and thrombosis were similar among groups (p > 0.05).

CONCLUSIONS

Due to a lack of high-quality studies, vastly varying protocols, and resulting statistical clinical and statistical heterogeneity, the level of evidence for recommendations regarding EM in patients with aSAH remains low. The currently available data indicated that mobilization within the first 5 days after aneurysm repair was feasible and safe without significant excessive adverse events, that neurological outcome with EM was almost certainly not worse than with prolonged immobilization, and that there was likely at least some reduction in length of hospital stay. Radiological and clinical vasospasms were not more frequent-with signals even trending toward a decrease-in patients who mobilized early. Higher-quality studies and implementation of full ERAS protocols are necessary to evaluate efficacy and safety with a higher level of evidence and to guide practical implementation through increased standardization. Clinical trial registration no.: CRD42023432828 (www.crd.york.ac.uk/prospero).

Influence of gender and sexual hormones on outcomes after pituitary surgery: a systematic review and meta-analysis

BACKGROUND

Although there is an increasing body of evidence showing gender differences in various medical domains as well as presentation and biology of pituitary adenoma (PA), gender differences regarding outcome of patients who underwent transsphenoidal resection of PA are poorly understood. The aim of this study was to identify gender differences in PA surgery.

METHODS

The PubMed/MEDLINE database was searched up to April 2023 to identify eligible articles. Quality appraisal and extraction were performed in duplicate.

RESULTS

A total of 40 studies including 4989 patients were included in this systematic review and meta-analysis. Our analysis showed odds ratio of postoperative biochemical remission in males vs. females of 0.83 (95% CI 0.59-1.15, P = 0.26), odds ratio of gross total resection in male vs. female patients of 0.68 (95% CI 0.34-1.39, P = 0.30), odds ratio of postoperative diabetes insipidus in male vs. female patients of 0.40 (95% CI 0.26-0.64, P < 0.0001), and a mean difference of preoperative level of prolactin in male vs. female patients of 11.62 (95% CI - 119.04-142.27, P = 0.86).

CONCLUSIONS

There was a significantly higher rate of postoperative DI in female patients after endoscopic or microscopic transsphenoidal PA surgery, and although there was some data in isolated studies suggesting influence of gender on postoperative biochemical remission, rate of GTR, and preoperative prolactin levels, these findings could not be confirmed in this meta-analysis and demonstrated no statistically significant effect. Further research is needed and future studies concerning PA surgery should report their data by gender or sexual hormones and ideally further assess their impact on PA surgery.

Development and external validation of clinical prediction models for pituitary surgery

Introduction

Gross total resection (GTR), Biochemical Remission (BR) and restitution of a priorly disrupted hypothalamus pituitary axis (new improvement, IMP) are important factors in pituitary adenoma (PA) resection surgery. Prediction of these metrics using simple and preoperatively available data might help improve patient care and contribute to a more personalized medicine.

Research question

This study aims to develop machine learning models predicting GTR, BR, and IMP in PA resection surgery, using preoperatively available data.

Material and methods

With data from patients undergoing endoscopic transsphenoidal surgery for PAs machine learning models for prediction of GTR, BR and IMP were developed and externally validated. Development was carried out on a registry from Bologna, Italy while external validation was conducted using patient data from Zurich, Switzerland.

Results

The model development cohort consisted of 1203 patients. GTR was achieved in 207 (17.2%, 945 (78.6%) missing), BR in 173 (14.4%, 992 (82.5%) missing) and IMP in 208 (17.3%, 167 (13.9%) missing) cases. In the external validation cohort 206 patients were included and GTR was achieved in 121 (58.7%, 32 (15.5%) missing), BR in 46 (22.3%, 145 (70.4%) missing) and IMP in 42 (20.4%, 7 (3.4%) missing) cases. The AUC at external validation amounted to 0.72 (95% CI: 0.63-0.80) for GTR, 0.69 (0.52-0.83) for BR, as well as 0.82 (0.76-0.89) for IMP.

Discussion and conclusion

All models showed adequate generalizability, performing similarly in training and external validation, confirming the possible potentials of machine learning in helping to adapt surgical therapy to the individual patient.

Differences between real-world and score-based decision-making in the microsurgical management of patients with unruptured intracranial aneurysms

BACKGROUND

Management of unruptured intracranial aneurysms (UIAs) is complex, balancing the risk of rupture and risk of treatment. Therefore, prediction scores have been developed to support clinicians in the management of UIAs. We analyzed the discrepancies between interdisciplinary cerebrovascular board decision-making factors and the results of the prediction scores in our cohort of patients who received microsurgical treatment of UIAs.

METHODS

Clinical, radiological, and demographical data of 221 patients presenting with 276 microsurgically treated aneurysms were collected, from January 2013 to June 2020. UIATS, PHASES, and ELAPSS were calculated for each treated aneurysm, resulting in subgroups favoring treatment or conservative management for each score. Cerebrovascular board decision-factors were collected and analyzed.

RESULTS

UIATS, PHASES, and ELAPSS recommended conservative management in 87 (31.5%) respectively in 110 (39.9%) and in 81 (29.3%) aneurysms. The cerebrovascular board decision-factors leading to treatment in these aneurysms (recommended to manage conservatively in the three scores) were: high life expectancy/young age (50.0%), angioanatomical factors (25.0%), multiplicity of aneurysms (16.7%). Analysis of cerebrovascular board decision-making factors in the "conservative management" subgroup of the UIATS showed that angioanatomical factors (P=0.001) led more frequently to surgery. PHASES and ELAPSS subgroups "conservative management" were more frequently treated due to clinical risk factors (P=0.002).

CONCLUSIONS

Our analysis showed more aneurysms were treated based on "real-world" decision-making than recommended by the scores. This is because these scores are models trying to reproduce reality, which is yet not fully understood. Aneurysms, which were recommended to manage conservatively, were treated mainly because of angioanatomy, high life expectancy, clinical risk factors, and patient's treatment wish. The UIATS is suboptimal regarding assessment of angioanatomy, the PHASES regarding clinical risk factors, complexity, and high life expectancy, and the ELAPSS regarding clinical risk factors and multiplicity of aneurysms. These findings support the need to optimize prediction models of UIAs.

Development and External Validation of a Deep Learning Algorithm to Identify and Localize Subarachnoid Hemorrhage on CT Scans

BACKGROUND AND OBJECTIVES

In medical imaging, a limited number of trained deep learning algorithms have been externally validated and released publicly. We hypothesized that a deep learning algorithm can be trained to identify and localize subarachnoid hemorrhage (SAH) on head computed tomography (CT) scans and that the trained model performs satisfactorily when tested using external and real-world data.

METHODS

We used noncontrast head CT images of patients admitted to Helsinki University Hospital between 2012 and 2017. We manually segmented (i.e., delineated) SAH on 90 head CT scans and used the segmented CT scans together with 22 negative (no SAH) control CT scans in training an open-source convolutional neural network (U-Net) to identify and localize SAH. We then tested the performance of the trained algorithm by using external data sets (137 SAH and 1,242 control cases) collected in 2 foreign countries and also by creating a data set of consecutive emergency head CT scans (8 SAH and 511 control cases) performed during on-call hours in 5 different domestic hospitals in September 2021. We assessed the algorithm's capability to identify SAH by calculating patient- and slice-level performance metrics, such as sensitivity and specificity.

RESULTS

In the external validation set of 1,379 cases, the algorithm identified 136 of 137 SAH cases correctly (sensitivity 99.3% and specificity 63.2%). Of the 49,064 axial head CT slices, the algorithm identified and localized SAH in 1845 of 2,110 slices with SAH (sensitivity 87.4% and specificity 95.3%). Of 519 consecutive emergency head CT scans imaged in September 2021, the algorithm identified all 8 SAH cases correctly (sensitivity 100.0% and specificity 75.3%). The slice-level (27,167 axial slices in total) sensitivity and specificity were 87.3% and 98.8%, respectively, as the algorithm identified and localized SAH in 58 of 77 slices with SAH. The performance of the algorithm can be tested on through a web service.

DISCUSSION

We show that the shared algorithm identifies SAH cases with a high sensitivity and that the slice-level specificity is high. In addition to openly sharing a high-performing deep learning algorithm, our work presents infrequently used approaches in designing, training, testing, and reporting deep learning algorithms developed for medical imaging diagnostics.

CLASSIFICATION OF EVIDENCE

This study provides Class III evidence that a deep learning algorithm correctly identifies the presence of subarachnoid hemorrhage on CT scan.

DeepEOR: automated perioperative volumetric assessment of variable grade gliomas using deep learning

PURPOSE

Volumetric assessments, such as extent of resection (EOR) or residual tumor volume, are essential criterions in glioma resection surgery. Our goal is to develop and validate segmentation machine learning models for pre- and postoperative magnetic resonance imaging scans, allowing us to assess the percentagewise tumor reduction after intracranial surgery for gliomas.

METHODS

For the development of the preoperative segmentation model (U-Net), MRI scans of 1053 patients from the Multimodal Brain Tumor Segmentation Challenge (BraTS) 2021 as well as from patients who underwent surgery at the University Hospital in Zurich were used. Subsequently, the model was evaluated on a holdout set containing 285 images from the same sources. The postoperative model was developed using 72 scans and validated on 45 scans obtained from the BraTS 2015 and Zurich dataset. Performance is evaluated using Dice Similarity score, Jaccard coefficient and Hausdorff 95%.

RESULTS

We were able to achieve an overall mean Dice Similarity Score of 0.59 and 0.29 on the pre- and postoperative holdout sets, respectively. Our algorithm managed to determine correct EOR in 44.1%.

CONCLUSION

Although our models are not suitable for clinical use at this point, the possible applications are vast, going from automated lesion detection to disease progression evaluation. Precise determination of EOR is a challenging task, but we managed to show that deep learning can provide fast and objective estimates.

Predictors of five-repetition sit-to-stand test performance in patients with lumbar degenerative disease

BACKGROUND

The five-repetition sit-to-stand test (5R-STS) has recently been validated as an objective measure of functional impairment in patients with lumbar degenerative disease (LDD). Knowledge of factors influencing 5R-STS performance is useful to correct for confounders, create personalized adjusted test times, and potentially identify prognostic subgroups. We evaluate factors predicting the 5R-STS performance in patients with LDD.

METHODS

Patients with LDD requiring surgery were included. Each participant performed the 5R-STS and completed a questionnaire that included their age, gender, weight, height, body mass index (BMI), smoking status, education level, employment type, ability to work, analgesic drug usage, history of previous spinal surgery, and EQ5D depression and anxiety domain. Surgical indication and index level of the spinal pathology were also recorded. Predictors of 5R-STS were identified through multivariable linear regression.

RESULTS

The cohort consisted of 240 patients, 47.9% being female (mean age, 47.7 ± 13.6 years). In the final multivariable model incorporating confounders, height (regression coefficient (RC), 0.08; 95% confidence interval (CI), 0.003/0.16, p = 0.042) and being an active smoker (RC, 2.44; 95%CI, 0.56/4.32, p = 0.012) were significant predictors of worse 5R-STS performance. Full ability to work (RC, - 2.39; 95%CI, - 4.39/ - 0.39, p = 0.020) was associated with a better 5R-STS performance. Age, height, surgical indication, index level of pathology, history of previous spine surgery, history of pain, analgesic drug use, employment type, and severity of anxiety and depression symptoms demonstrated confounding effect on the 5R-STS time.

CONCLUSIONS

Greater height, being an active smoker, and inability to work are significant predictors of worse 5R-STS performance in patients with LDD.

TRIAL REGISTRATION

ClinicalTrials.gov Identifier: NCT03303300 and NCT03321357.

Meta-topologies define distinct anatomical classes of brain tumours linked to histology and survival

The current World Health Organization classification integrates histological and molecular features of brain tumours. The aim of this study was to identify generalizable topological patterns with the potential to add an anatomical dimension to the classification of brain tumours. We applied non-negative matrix factorization as an unsupervised pattern discovery strategy to the fine-grained topographic tumour profiles of 936 patients with neuroepithelial tumours and brain metastases. From the anatomical features alone, this machine learning algorithm enabled the extraction of latent topological tumour patterns, termed meta-topologies. The optimal part-based representation was automatically determined in 10 000 split-half iterations. We further characterized each meta-topology's unique histopathologic profile and survival probability, thus linking important biological and clinical information to the underlying anatomical patterns. In neuroepithelial tumours, six meta-topologies were extracted, each detailing a transpallial pattern with distinct parenchymal and ventricular compositions. We identified one infratentorial, one allopallial, three neopallial (parieto-occipital, frontal, temporal) and one unisegmental meta-topology. Each meta-topology mapped to distinct histopathologic and molecular profiles. The unisegmental meta-topology showed the strongest anatomical-clinical link demonstrating a survival advantage in histologically identical tumours. Brain metastases separated to an infra- and supratentorial meta-topology with anatomical patterns highlighting their affinity to the cortico-subcortical boundary of arterial watershed areas.Using a novel data-driven approach, we identified generalizable topological patterns in both neuroepithelial tumours and brain metastases. Differences in the histopathologic profiles and prognosis of these anatomical tumour classes provide insights into the heterogeneity of tumour biology and might add to personalized clinical decision-making.

Comparison of Deep Learning and Classical Machine Learning Algorithms to Predict Postoperative Outcomes for Anterior Cervical Discectomy and Fusion Procedures With State-of-the-art Performance

STUDY DESIGN

Retrospective cohort.

OBJECTIVE

Due to anterior cervical discectomy and fusion (ACDF) popularity, it is important to predict postoperative complications, unfavorable 90-day readmissions, and two-year reoperations to improve surgical decision-making, prognostication, and planning.

SUMMARY OF BACKGROUND DATA

Machine learning has been applied to predict postoperative complications for ACDF; however, studies were limited by sample size and model type. These studies achieved ≤0.70 area under the curve (AUC). Further approaches, not limited to ACDF, focused on specific complication types and resulted in AUC between 0.70 and 0.76.

MATERIALS AND METHODS

The IBM MarketScan Commercial Claims and Encounters Database and Medicare Supplement were queried from 2007 to 2016 to identify adult patients who underwent an ACDF procedure (N=176,816). Traditional machine learning algorithms, logistic regression, and support vector machines, were compared with deep neural networks to predict: 90-day postoperative complications, 90-day readmission, and two-year reoperation. We further generated random deep learning model architectures and trained them on the 90-day complication task to approximate an upper bound. Last, using deep learning, we investigated the importance of each input variable for the prediction of 90-day postoperative complications in ACDF.

RESULTS

For the prediction of 90-day complication, 90-day readmission, and two-year reoperation, the deep neural network-based models achieved AUC of 0.832, 0.713, and 0.671. Logistic regression achieved AUCs of 0.820, 0.712, and 0.671. Support vector machine approaches were significantly lower. The upper bound of deep learning performance was approximated as 0.832. Myelopathy, age, human immunodeficiency virus, previous myocardial infarctions, obesity, and documentary weakness were found to be the strongest variable to predict 90-day postoperative complications.

CONCLUSIONS

The deep neural network may be used to predict complications for clinical applications after multicenter validation. The results suggest limited added knowledge exists in interactions between the input variables used for this task. Future work should identify novel variables to increase predictive power.

Collagen-bound fibrin sealant (TachoSil®) for dural closure in cranial surgery: single-centre comparative cohort study and systematic review of the literature

Cerebrospinal fluid (CSF) leakage is a well-known complication of craniotomies and there are several dural closure techniques. One commonly used commercial product as adjunct for dural closure is the collagen-bound fibrin sealant TachoSil®. We analysed whether the addition of TachoSil has beneficial effects on postoperative complications and outcomes. Our prospective, institutional database was retrospectively queried, and 662 patients undergoing craniotomy were included. Three hundred fifty-two were treated with dural suture alone, and in 310, TachoSil was added after primary suture. Our primary endpoint was the rate of postoperative complications associated with CSF leakage. Secondary endpoints included functional, disability and neurological outcome. Systematic review according to PRISMA guidelines was performed to identify studies comparing primary dural closure with and without additional sealants. Postoperative complications associated with CSF leakage occurred in 24 (7.74%) and 28 (7.95%) procedures with or without TachoSil, respectively (p = 0.960). Multivariate analysis confirmed no significant differences in complication rate between the two groups (aOR 0.97, 95% CI 0.53–1.80, p = 0.930). There were no significant disparities in postoperative functional, disability or neurological scores. The systematic review identified 661 and included 8 studies in the qualitative synthesis. None showed a significant superiority of additional sealants over standard technique regarding complications, rates of revision surgery or outcome. According to our findings, we summarize that routinary use of TachoSil and similar products as adjuncts to primary dural sutures after intracranial surgical procedures is safe but without clear advantage in complication avoidance or outcome. Future studies should investigate whether their use is beneficial in high-risk settings.

Sex-related differences in postoperative complications following elective craniotomy for intracranial lesions: An observational study

INTRODUCTION

The integration of sex-related differences in neurosurgery is crucial for new, possible sex-specific, therapeutic approaches. In neurosurgical emergencies, such as traumatic brain injury and aneurysmal subarachnoid hemorrhage, these differences have been investigated. So far, little is known concerning the impact of sex on frequency of postoperative complications after elective craniotomy. This study investigates whether sex-related differences exist in frequency of postoperative complications in patients who underwent elective craniotomy for intracranial lesion.

MATERIAL AND METHODS

All consecutive patients who underwent an elective intracranial procedure over a 2-year period at our center were eligible for inclusion in this retrospective study. Demographic data, comorbidities, frequency of postoperative complications at 24 hours following surgery and at discharge, and hospital length of stay were compared among females and males.

RESULTS

Overall, 664 patients were considered for the analysis. Of those, 339 (50.2%) were females. Demographic data were comparable among females and males. More females than males suffered from allergic, muscular, and rheumatic disorders. No differences in frequency of postoperative complications at 24 hours after surgery and at discharge were observed among females and males. Similarly, the hospital length of stay was comparable.

CONCLUSIONS

In the present study, no sex-related differences in frequency of early postoperative complications and at discharge following elective craniotomy for intracranial lesions were observed.

Association of Medical Comorbidities With Objective Functional Impairment in Lumbar Degenerative Disc Disease

STUDY DESIGN

Analysis of a prospective 2-center database.

OBJECTIVES

Medical comorbidities co-determine clinical outcome. Objective functional impairment (OFI) provides a supplementary dimension of patient assessment. We set out to study whether comorbidities are associated with the presence and degree of OFI in this patient population.

METHODS

Patients with degenerative diseases of the spine preoperatively performed the timed-up-and-go (TUG) test and a battery of questionnaires. Comorbidities were quantified using the Charlson Comorbidity Index (CCI) and the American Society of Anesthesiology (ASA) grading. Crude and adjusted linear regression models were fitted.

RESULTS

Of 375 included patients, 97 (25.9%) presented at least some degree of medical comorbidity according to the CCI, and 312 (83.2%) according to ASA grading. In the univariate analysis, the CCI was inconsistently associated with OFI. Only patients with low-grade CCI comorbidity displayed significantly higher TUG test times (p = 0.004). In the multivariable analysis, this effect persisted for patients with CCI = 1 (p = 0.030). Regarding ASA grade, patients with ASA = 3 exhibited significantly increased TUG test times (p = 0.003) and t-scores (p = 0.015). This effect disappeared after multivariable adjustment (p = 0.786 and p = 0.969). In addition, subjective functional impairment according to ODI, and EQ5D index was moderately associated with comorbidities according to ASA (all p < 0.05).

CONCLUSION

The degree of medical comorbidities appears only weakly and inconsistently associated with OFI in patients scheduled for degenerative lumbar spine surgery, especially after controlling for potential confounders. TUG testing may be valid even in patients with relatively severe comorbidities who are able to complete the test.

Machine learning-based clinical outcome prediction in surgery for acromegaly

PURPOSE

Biochemical remission (BR), gross total resection (GTR), and intraoperative cerebrospinal fluid (CSF) leaks are important metrics in transsphenoidal surgery for acromegaly, and prediction of their likelihood using machine learning would be clinically advantageous. We aim to develop and externally validate clinical prediction models for outcomes after transsphenoidal surgery for acromegaly.

METHODS

Using data from two registries, we develop and externally validate machine learning models for GTR, BR, and CSF leaks after endoscopic transsphenoidal surgery in acromegalic patients. For the model development a registry from Bologna, Italy was used. External validation was then performed using data from Zurich, Switzerland. Gender, age, prior surgery, as well as Hardy and Knosp classification were used as input features. Discrimination and calibration metrics were assessed.

RESULTS

The derivation cohort consisted of 307 patients (43.3% male; mean [SD] age, 47.2 [12.7] years). GTR was achieved in 226 (73.6%) and BR in 245 (79.8%) patients. In the external validation cohort with 46 patients, 31 (75.6%) achieved GTR and 31 (77.5%) achieved BR. Area under the curve (AUC) at external validation was 0.75 (95% confidence interval: 0.59-0.88) for GTR, 0.63 (0.40-0.82) for BR, as well as 0.77 (0.62-0.91) for intraoperative CSF leaks. While prior surgery was the most important variable for prediction of GTR, age, and Hardy grading contributed most to the predictions of BR and CSF leaks, respectively.

CONCLUSIONS

Gross total resection, biochemical remission, and CSF leaks remain hard to predict, but machine learning offers potential in helping to tailor surgical therapy. We demonstrate the feasibility of developing and externally validating clinical prediction models for these outcomes after surgery for acromegaly and lay the groundwork for development of a multicenter model with more robust generalization.

Five-Repetition Sit-to-Stand Test Performance in Healthy Individuals: Reference Values and Predictors From 2 Prospective Cohorts

Objective

The 5-repetition-sit-to-stand (5R-STS) test is an objective test of functional impairment- commonly used in various diseases, including lumbar degenerative disc diseases. It is used to measure the severity of disease and to monitor recovery. We aimed to evaluate reference values for the test, as well as factors predicting 5R-STS performance in healthy adults.

Methods

Healthy adults (> 18 years of age) were recruited, and their 5R-STS time was measured. Their age, sex, weight, height, body mass index (BMI), smoking status, education level, work situation and EuroQOL-5D Healthy & Anxiety category were recorded. Linear regression analysis was employed to identify predictors of 5R-STS performance.

Results

We included 172 individuals with mean age of 39.4±14.1 years and mean BMI of 24.0 ±4.0 kg/m². Females constituted 57%. Average 5R-STS time was 6.21 ±1.92 seconds, with an upper limit of normal of 12.39 seconds. In a multivariable model, age (regression coefficient [RC], 0.07; 95% confidence interval [CI], 0.05/0.09; p<0.001), male sex (RC, -0.87; 95% CI, -1.50 to -0.23; p=0.008), BMI (RC, 0.40; 95% CI, 0.10–0.71; p=0.010), height (RC, 0.13; 95% CI, 0.04–0.22; p=0.006), and houseworker status (RC, -1.62; 95% CI, -2.93 to -0.32; p=0.016) were significantly associated with 5R-STS time. Anxiety and depression did not influence performance significantly (RC, 0.82; 95% CI, -0.14 to 1.77; p=0.097).

Conclusion

The presented reference values can be applied as normative data for 5R-STS in healthy adults, and are necessary to judge what constitutes abnormal performance. We identified several significant factors associated with 5R-STS performance that may be used to calculate individualized expected test times.

Neurosurgery outcomes and complications in a monocentric 7-year patient registry

Introduction

Capturing adverse events reliably is paramount for clinical practice and research alike. In the era of “big data”, prospective registries form the basis of clinical research and quality improvement.

Research question

To present results of long-term implementation of a prospective patient registry, and evaluate the validity of the Clavien-Dindo grade (CDG) to classify complications in neurosurgery.

Materials and methods

A prospective registry for cranial and spinal neurosurgical procedures was implemented in 2013. The CDG – a complication grading focused on need for unplanned therapeutic intervention – was used to grade complications. We assess construct validity of the CDG.

Results

Data acquisition integrated into our hospital workflow permitted to include all eligible patients into the registry. We have registered 8226 patients that were treated in 11994 surgeries and 32494 consultations up until December 2020. Similarly, we have captured 1245 complications on 6308 patient discharge forms (20%) since full operational status of the registry. The majority of complications (819/6308 ​= ​13%) were treated without invasive treatment (CDG 1 or CDG 2). At discharge, there was a clear correlation of CDG and the Karnofsky Performance Status (KPS, rho ​= ​-0.29, slope -7 KPS percentage points per increment of CDG) and the length of stay (rho ​= ​0.43, slope 3.2 days per increment of CDG).

Discussion and conclusion

Patient registries with high completeness and objective capturing of complications are central to the process of quality improvement. The CDG demonstrates construct validity as a measure of complication classification in a neurosurgical patient population.

-

A prospective registry for cranial and spinal neurosurgical procedures was implemented in 2013.

-

We have registered 8226 patients that were treated in 11994 surgeries and 32494 consultations up until December 2020.

-

There was a clear correlation of CDG with the Karnofsky Performance Status and with length of hospital stay.

-

The Clavien-Dindo grading (CDG) demonstrates construct validity as a measure of complication severity in a neurosurgical patient population.

Foundations of Machine Learning-Based Clinical Prediction Modeling: Part III-Model Evaluation and Other Points of Significance

Various available metrics to describe model performance in terms of discrimination (area under the curve (AUC), accuracy, sensitivity, specificity, positive predictive value, negative predictive value, F1 Score) and calibration (slope, intercept, Brier score, expected/observed ratio, Estimated Calibration Index, Hosmer-Lemeshow goodness-of-fit) are presented. Recalibration is introduced, with Platt scaling and Isotonic regression as proposed methods. We also discuss considerations regarding the sample size required for optimal training of clinical prediction models-explaining why low sample sizes lead to unstable models, and offering the common rule of thumb of at least ten patients per class per input feature, as well as some more nuanced approaches. Missing data treatment and model-based imputation instead of mean, mode, or median imputation is also discussed. We explain how data standardization is important in pre-processing, and how it can be achieved using, e.g. centering and scaling. One-hot encoding is discussed-categorical features with more than two levels must be encoded as multiple features to avoid wrong assumptions. Regarding binary classification models, we discuss how to select a sensible predicted probability cutoff for binary classification using the closest-to-(0,1)-criterion based on AUC or based on the clinical question (rule-in or rule-out). Extrapolation is also discussed.

Foundations of Machine Learning-Based Clinical Prediction Modeling: Part II-Generalization and Overfitting

We review the concept of overfitting, which is a well-known concern within the machine learning community, but less established in the clinical community. Overfitted models may lead to inadequate conclusions that may wrongly or even harmfully shape clinical decision-making. Overfitting can be defined as the difference among discriminatory training and testing performance, while it is normal that out-of-sample performance is equal to or ever so slightly worse than training performance for any adequately fitted model, a massively worse out-of-sample performance suggests relevant overfitting. We delve into resampling methods, specifically recommending k-fold cross-validation and bootstrapping to arrive at realistic estimates of out-of-sample error during training. Also, we encourage the use of regularization techniques such as L1 or L2 regularization, and to choose an appropriate level of algorithm complexity for the type of dataset used. Data leakage is addressed, and the importance of external validation to assess true out-of-sample performance and to-upon successful external validation-release the model into clinical practice is discussed. Finally, for highly dimensional datasets, the concepts of feature reduction using principal component analysis (PCA) as well as feature elimination using recursive feature elimination (RFE) are elucidated.

Foundations of Machine Learning-Based Clinical Prediction Modeling: Part I-Introduction and General Principles

We provide explanations on the general principles of machine learning, as well as analytical steps required for successful machine learning-based predictive modeling, which is the focus of this series. In particular, we define the terms machine learning, artificial intelligence, as well as supervised and unsupervised learning, continuing by introducing optimization, thus, the minimization of an objective error function as the central dogma of machine learning. In addition, we discuss why it is important to separate predictive and explanatory modeling, and most importantly state that a prediction model should not be used to make inferences. Lastly, we broadly describe a classical workflow for training a machine learning model, starting with data pre-processing and feature engineering and selection, continuing on with a training structure consisting of a resampling method, hyperparameter tuning, and model selection, and ending with evaluation of model discrimination and calibration as well as robust internal or external validation of the fully developed model. Methodological rigor and clarity as well as understanding of the underlying reasoning of the internal workings of a machine learning approach are required, otherwise predictive applications despite being strong analytical tools are not well accepted into the clinical routine.

Machine Intelligence in Clinical Neuroscience: Taming the Unchained Prometheus

The democratization of machine learning (ML) through availability of open-source learning libraries, the availability of datasets in the "big data" era, increasing computing power even on mobile devices, and online training resources have both led to an explosion in applications and publications of ML in the clinical neurosciences, but has also enabled a dangerous amount of flawed analyses and cardinal methodological errors committed by benevolent authors. While powerful ML methods are nowadays available to almost anyone and can be applied after just few minutes of familiarizing oneself with these methods, that does not imply that one has mastered these techniques. This textbook for clinicians aims to demystify ML by illustrating its methodological foundations, as well as some specific applications throughout clinical neuroscience, and its limitations. While our mind can recognize, abstract, and deal with the many uncertainties in clinical practice, algorithms cannot. Algorithms must remain tools of our own mind, tools that we should be able to master, control, and apply to our advantage in an adjunctive manner. Our hope is that this book inspires and instructs physician-scientists to continue to develop the seeds that have been planted for machine intelligence in clinical neuroscience, not forgetting their inherent limitations.

Machine Learning in Pituitary Surgery

Machine learning applications in neurosurgery are increasingly reported for diverse tasks such as faster and more accurate preoperative diagnosis, enhanced lesion characterization, as well as surgical outcome, complications and healthcare cost prediction. Even though the pertinent literature in pituitary surgery is less extensive with respect to other neurosurgical diseases, past research attempted to answer clinically relevant questions to better assist surgeons and clinicians. In the present chapter we review reported ML applications in pituitary surgery including differential diagnosis, preoperative lesion characterization (immunohistochemistry, cavernous sinus invasion, tumor consistency), surgical outcome and complication predictions (gross total resection, tumor recurrence, and endocrinological remission, cerebrospinal fluid leak, postoperative hyponatremia). Moreover, we briefly discuss from a practical standpoint the current barriers to clinical translation of machine learning research. On the topic of pituitary surgery, published reports can be considered mostly preliminary, requiring larger training populations and strong external validation. Thoughtful selection of clinically relevant outcomes of interest and transversal application of model development pipeline-together with accurate methodological planning and multicenter collaborations-have the potential to overcome current limitations and ultimately provide additional tools for more informed patient management.

Machine Learning Algorithms in Neuroimaging: An Overview

Machine learning (ML) and artificial intelligence (AI) applications in the field of neuroimaging have been on the rise in recent years, and their clinical adoption is increasing worldwide. Deep learning (DL) is a field of ML that can be defined as a set of algorithms enabling a computer to be fed with raw data and progressively discover-through multiple layers of representation-more complex and abstract patterns in large data sets. The combination of ML and radiomics, namely the extraction of features from medical images, has proven valuable, too: Radiomic information can be used for enhanced image characterization and prognosis or outcome prediction. This chapter summarizes the basic concepts underlying ML application for neuroimaging and discusses technical aspects of the most promising algorithms, with a specific focus on Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs), in order to provide the readership with the fundamental theoretical tools to better understand ML in neuroimaging. Applications are highlighted from a practical standpoint in the last section of the chapter, including: image reconstruction and restoration, image synthesis and super-resolution, registration, segmentation, classification, and outcome prediction.

Machine learning-augmented objective functional testing in the degenerative spine: quantifying impairment using patient-specific five-repetition sit-to-stand assessment

OBJECTIVE

What is considered "abnormal" in clinical testing is typically defined by simple thresholds derived from normative data. For instance, when testing using the five-repetition sit-to-stand (5R-STS) test, the upper limit of normal (ULN) from a population of spine-healthy volunteers (10.5 seconds) is used to identify objective functional impairment (OFI), but this fails to consider different properties of individuals (e.g., taller and shorter, older and younger). Therefore, the authors developed a personalized testing strategy to quantify patient-specific OFI using machine learning.

METHODS

Patients with disc herniation, spinal stenosis, spondylolisthesis, or discogenic chronic low-back pain and a population of spine-healthy volunteers, from two prospective studies, were included. A machine learning model was trained on normative data to predict personalized "expected" test times and their confidence intervals and ULNs (99th percentiles) based on simple demographics. OFI was defined as a test time greater than the personalized ULN. OFI was categorized into types 1 to 3 based on a clustering algorithm. A web app was developed to deploy the model clinically.

RESULTS

Overall, 288 patients and 129 spine-healthy individuals were included. The model predicted "expected" test times with a mean absolute error of 1.18 (95% CI 1.13-1.21) seconds and R2 of 0.37 (95% CI 0.34-0.41). Based on the implemented personalized testing strategy, 191 patients (66.3%) exhibited OFI. Type 1, 2, and 3 impairments were seen in 64 (33.5%), 91 (47.6%), and 36 (18.8%) patients, respectively. Increasing detected levels of OFI were associated with statistically significant increases in subjective functional impairment, extreme anxiety and depression symptoms, being bedridden, extreme pain or discomfort, inability to carry out activities of daily living, and a limited ability to work.

CONCLUSIONS

In the era of "precision medicine," simple population-based thresholds may eventually not be adequate to monitor quality and safety in neurosurgery. Individualized assessment integrating machine learning techniques provides more detailed and objective clinical assessment. The personalized testing strategy demonstrated concurrent validity with quality-of-life measures, and the freely accessible web app (https://neurosurgery.shinyapps.io/5RSTS/) enabled clinical application.

Fostering reproducibility and generalizability in machine learning for clinical prediction modeling in spine surgery

As the use of machine learning algorithms in the development of clinical prediction models has increased, researchers are becoming more aware of the deleterious effects that stem from the lack of reporting standards. One of the most obvious consequences is the insufficient reproducibility found in current prediction models. In an attempt to characterize methods to improve reproducibility and to allow for better clinical performance, we utilize a previously proposed taxonomy that separates reproducibility into 3 components: technical, statistical, and conceptual reproducibility. By following this framework, we discuss common errors that lead to poor reproducibility, highlight the importance of generalizability when evaluating a ML model's performance, and provide suggestions to optimize generalizability to ensure adequate performance. These efforts are a necessity before such models are applied to patient care.

Predictive modeling of long-term opioid and benzodiazepine use after intradural tumor resection

BACKGROUND CONTEXT

Despite increased awareness of the ongoing opioid epidemic, opioid and benzodiazepine use remain high after spine surgery. In particular, long-term co-prescription of opioids and benzodiazepines have been linked to high risk of overdose-associated death. Tumor patients represent a unique subset of spine surgery patients and few studies have attempted to develop predictive models to anticipate long-term opioid and benzodiazepine use after spinal tumor resection.

METHODS

The IBM Watson Health MarketScan Database and Medicare Supplement were assessed to identify admissions for intradural tumor resection between 2007 and 2015. Adult patients were required to have at least 6 months of continuous preadmission baseline data and 12 months of continuous postdischarge follow-up. Primary outcomes were long-term opioid and benzodiazepine use, defined as at least 6 prescriptions within 12 months. Secondary outcomes were durations of opioid and benzodiazepine prescribing. Logistic regression models, with and without regularization, were trained on an 80% training sample and validated on the withheld 20%.

RESULTS

A total of 1,942 patients were identified. The majority of tumors were extramedullary (74.8%) and benign (62.5%). A minority of patients received arthrodesis (9.2%) and most patients were discharged to home (79.1%). Factors associated with postdischarge opioid use duration include tumor malignancy (vs benign, B=19.8 prescribed-days/year, 95% confidence interval [CI] 1.1-38.5) and intramedullary compartment (vs extramedullary, B=18.1 prescribed-days/year, 95% CI 3.3-32.9). Pre- and perioperative use of prescribed nonsteroidal anti-inflammatory drugs and gabapentin/pregabalin were associated with shorter and longer duration opioid use, respectively. History of opioid and history of benzodiazepine use were both associated with increased postdischarge opioid and benzodiazepine use. Intramedullary location was associated with longer duration postdischarge benzodiazepine use (B=10.3 prescribed-days/year, 95% CI 1.5-19.1). Among assessed models, elastic net regularization demonstrated best predictive performance in the withheld validation cohort when assessing both long-term opioid use (area under curve [AUC]=0.748) and long-term benzodiazepine use (AUC=0.704). Applying our model to the validation set, patients scored as low-risk demonstrated a 4.8% and 2.4% risk of long-term opioid and benzodiazepine use, respectively, compared to 35.2% and 11.1% of high-risk patients.

CONCLUSIONS

We developed and validated a parsimonious, predictive model to anticipate long-term opioid and benzodiazepine use early after intradural tumor resection, providing physicians opportunities to consider alternative pain management strategies.

Anatomical phenotyping and staging of brain tumours

Unlike other tumors, the anatomical extent of brain tumors is not objectified and quantified through staging. Staging systems are based on understanding the anatomical sequence of tumor progression and its relationship to histopathological dedifferentiation and survival. The aim of this study was to describe the spatiotemporal phenotype of the most frequent brain tumor entities, to assess the association of anatomical tumor features with survival probability and to develop a staging system for WHO grade 2 and 3 gliomas and glioblastoma. Anatomical phenotyping was performed on a consecutive cohort of 1000 patients with first diagnosis of a primary or secondary brain tumor. Tumor probability in different topographic, phylogenetic and ontogenetic parcellation units was assessed on preoperative MRI through normalization of the relative tumor prevalence to the relative volume of the respective structure. We analyzed the spatiotemporal tumor dynamics by cross-referencing preoperative against preceding and subsequent MRIs of the respective patient. The association between anatomical phenotype and outcome defined prognostically critical anatomical tumor features at diagnosis. Based on a hypothesized sequence of anatomical tumor progression, we developed a three-level staging system for WHO grade 2 and 3 gliomas and glioblastoma. This staging system was validated internally in the original cohort and externally in an independent cohort of 300 consecutive patients. While primary central nervous system lymphoma showed highest probability along white matter tracts, metastases enriched along terminal arterial flow areas. Neuroepithelial tumors mapped along all sectors of the ventriculocortical axis, while adjacent units were spared, consistent with a transpallial behavior within phylo-ontogenetic radial units. Their topographic pattern correlated with morphogenetic processes of convergence and divergence of radial units during phylo- and ontogenesis. While a ventriculofugal growth dominated in neuroepithelial tumors, a gradual deviation from this neuroepithelial spatiotemporal behavior was found with progressive histopathological dedifferentiation. The proposed three-level staging system for WHO grade 2 and 3 gliomas and glioblastoma correlated with the degree of histological dedifferentiation and proved accurate in terms of survival upon both internal and external validation. In conclusion, this study identified specific spatiotemporal phenotypes in brain tumors through topographic probability and growth pattern assessment. The association of anatomical tumor features with survival defined critical steps in the anatomical sequence of neuroepithelial tumor progression, based on which a staging system for WHO grade 2 and 3 gliomas and glioblastoma was developed and validated.

Impact of intraoperative magnetic resonance imaging on gross total resection, extent of resection, and residual tumor volume in pituitary surgery: systematic review and meta-analysis

BACKGROUND

Residual tumor tissue after pituitary adenoma surgery, is linked with additional morbidity and mortality. Intraoperative magnetic resonance imaging (ioMRI) could improve resection. We aim to assess the improvement in gross total resection (GTR), extent of resection (EOR), and residual tumor volume (RV) achieved using ioMRI.

METHODS

A systematic review was carried out on PubMed/MEDLINE to identify any studies reporting intra- and postoperative (1) GTR, (2) EOR, or (3) RV in patients who underwent resection of pituitary adenomas with ioMRI. Random effects meta-analysis of the rate of improvement after ioMRI for these three surgical outcomes was intended.

RESULTS

Among 34 included studies (2130 patients), the proportion of patients with conversion to GTR (∆GTR) after ioMRI was 0.19 (95% CI 0.15-0.23). Mean ∆EOR was + 9.07% after ioMRI. Mean ∆RV was 0.784 cm3. For endoscopically treated patients, ∆GTR was 0.17 (95% CI 0.09-0.25), while microscopic ∆GTR was 0.19 (95% CI 0.15-0.23). Low-field ioMRI studies demonstrated a ∆GTR of 0.19 (95% CI 0.11-0.28), while high-field and ultra-high-field ioMRI demonstrated a ∆GTR of 0.19 (95% CI 0.15-0.24) and 0.20 (95% CI 0.13-0.28), respectively.

CONCLUSIONS

Our meta-analysis demonstrates that around one fifth of patients undergoing pituitary adenoma resection convert from non-GTR to GTR after the use of ioMRI. EOR and RV can also be improved to a certain extent using ioMRI. Endoscopic versus microscopic technique or field strength does not appear to alter the impact of ioMRI. Statistical heterogeneity was high, indicating that the improvement in surgical results due to ioMRI varies considerably by center.

Analysis of Factors Relevant to Revenue Enhancement in Hernia Interventions (SwissDRG G09)

Background: Since diagnosis-related groups (SwissDRG) were established in Switzerland in 2012, small and medium-size hospitals have encountered increasing financial troubles. Even though hernia repair operations are frequent, most hospitals fail to cover their costs with these procedures. Previous studies have focused mainly on analyzing costs and the contributing factors but less on variables that can be positively influenced. Therefore, this study aims to identify the relevant and influenceable factors for revenue growth in hernia repair surgery. Methods: Data from all patients who underwent the SwissDRG G09 surgery for a hernia in 2019 were analyzed. The contribution margin (CM4), as well as any over- or under-coverage, was correlated to case-specific costs. Results: A total of 168 patients received hernia repair surgery with the SwissDRG code G09. The average revenue/loss generated by one procedure was CHF −623.84. Procedures covered by the General Health Insurance (OKP) generated a loss of CHF −830.70 on average, whereas procedures covered by private insurance companies (VVG) generated revenue of CHF +1100 on average. Significant factors impacting the profitability of hernia repair operations were teaching during surgery (p < 0.005), the surgical operating time (p < 0.001), the total anesthesia time (p < 0.001), the number of surgeons present (p = 0.022), the insurance state of patients (p < 0.001), and the type of surgery (p < 0.01 for Lichtenstein’s procedure). Conclusions: This study reveals that hernia repair surgery performed under cost coverage by OKP is generally unprofitable. Our results further imply that the most important and influenceable factors for revenue enhancement are the quality and process optimization of the surgical department. To compensate for this deficit, hospitals should aim to increase the percentage of patients with private health insurance coverage in their procedures. Since outpatient surgery does not provide a valid alternative due to the low reimbursement by insurance companies, the cost efficiency of inpatient hernia repair needs to be increased by process optimization of the surgical department; for instance, by providing specialized hernia teams performing with shorter operation times and high quality.

Machine Vision for Real-Time Intraoperative Anatomic Guidance: A Proof-of-Concept Study in Endoscopic Pituitary Surgery

BACKGROUND

Current intraoperative orientation methods either rely on preoperative imaging, are resource-intensive to implement, or difficult to interpret. Real-time, reliable anatomic recognition would constitute another strong pillar on which neurosurgeons could rest for intraoperative orientation.

OBJECTIVE

To assess the feasibility of machine vision algorithms to identify anatomic structures using only the endoscopic camera without prior explicit anatomo-topographic knowledge in a proof-of-concept study.

METHODS

We developed and validated a deep learning algorithm to detect the nasal septum, the middle turbinate, and the inferior turbinate during endoscopic endonasal approaches based on endoscopy videos from 23 different patients. The model was trained in a weakly supervised manner on 18 and validated on 5 patients. Performance was compared against a baseline consisting of the average positions of the training ground truth labels using a semiquantitative 3-tiered system.

RESULTS

We used 367 images extracted from the videos of 18 patients for training, as well as 182 test images extracted from the videos of another 5 patients for testing the fully developed model. The prototype machine vision algorithm was able to identify the 3 endonasal structures qualitatively well. Compared to the baseline model based on location priors, the algorithm demonstrated slightly but statistically significantly (P < .001) improved annotation performance.

CONCLUSION

Automated recognition of anatomic structures in endoscopic videos by means of a machine vision model using only the endoscopic camera without prior explicit anatomo-topographic knowledge is feasible. This proof of concept encourages further development of fully automated software for real-time intraoperative anatomic guidance during surgery.

Machine learning-based prediction of outcomes of the endoscopic endonasal approach in Cushing disease: is the future coming?

OBJECTIVE

Machine learning (ML) is an innovative method to analyze large and complex data sets. The aim of this study was to evaluate the use of ML to identify predictors of early postsurgical and long-term outcomes in patients treated for Cushing disease (CD).

METHODS

All consecutive patients in our center who underwent surgery for CD through the endoscopic endonasal approach were retrospectively reviewed. Study endpoints were gross-tumor removal (GTR), postsurgical remission, and long-term control of disease. Several demographic, radiological, and histological factors were assessed as potential predictors. For ML-based modeling, data were randomly divided into 2 sets with an 80% to 20% ratio for bootstrapped training and testing, respectively. Several algorithms were tested and tuned for the area under the curve (AUC).

RESULTS

The study included 151 patients. GTR was achieved in 137 patients (91%), and postsurgical hypersecretion remission was achieved in 133 patients (88%). At last follow-up, 116 patients (77%) were still in remission after surgery and in 21 patients (14%), CD was controlled with complementary treatment (overall, of 131 cases, 87% were under control at follow-up). At internal validation, the endpoints were predicted with AUCs of 0.81-1.00, accuracy of 81%-100%, and Brier scores of 0.035-0.151. Tumor size and invasiveness and histological confirmation of adrenocorticotropic hormone (ACTH)-secreting cells were the main predictors for the 3 endpoints of interest.

CONCLUSIONS

ML algorithms were used to train and internally validate robust models for all the endpoints, giving accurate outcome predictions in CD cases. This analytical method seems promising for potentially improving future patient care and counseling; however, careful clinical interpretation of the results remains necessary before any clinical adoption of ML. Moreover, further studies and increased sample sizes are definitely required before the widespread adoption of ML to the study of CD.

Topographic volume-standardization atlas of the human brain

Specific anatomical patterns are seen in various diseases affecting the brain. Clinical studies on the topography of pathologies are often limited by the absence of a normalization of the prevalence of pathologies to the relative volume of the affected anatomical structures. A comprehensive reference on the relative volumes of clinically relevant anatomical structures serving for such a normalization, is currently lacking. The analyses are based on anatomical high-resolution three-dimensional T1-weighted magnetic resonance imaging data of 30 healthy Caucasian volunteers, including 14 females (mean age 37.79 years, SD 13.04) and 16 males (mean age 38.31 years, SD 16.91). Semi-automated anatomical segmentation was used, guided by a neuroanatomical parcellation algorithm differentiating 96 structures. Relative volumes were derived by normalizing parenchymal structures to the total individual encephalic volume and ventricular segments to the total individual ventricular volume. The present investigation provides the absolute and relative volumes of 96 anatomical parcellation units of the human encephalon. A larger absolute volume in males than in females is found for almost all parcellation units. While parenchymal structures display a trend towards decreasing volumes with increasing age, a significant inverse effect is seen with the ventricular system. The variances in volumes as well as the effects of gender and age are given for each structure before and after normalization. The provided atlas constitutes an anatomically detailed and comprehensive analysis of the absolute and relative volumes of the human encephalic structures using a clinically oriented parcellation algorithm. It is intended to serve as a reference for volume-standardization in clinical studies on the topographic prevalence of pathologies.

Enhanced Recovery After Surgery strategies for elective craniotomy: a systematic review

OBJECTIVE

Enhanced Recovery After Surgery (ERAS) has led to a paradigm shift in perioperative care through multimodal interventions. Still, ERAS remains a relatively new concept in neurosurgery, and there is no summary of evidence on ERAS applications in cranial neurosurgery.

METHODS

The authors systematically reviewed the literature using the PubMed/MEDLINE, Embase, Scopus, and Cochrane Library databases for ERAS protocols and elements. Studies had to assess at least one pre-, peri-, or postoperative ERAS element and evaluate at least one of the following outcomes: 1) length of hospital stay, 2) length of ICU stay, 3) postoperative pain, 4) direct and indirect healthcare cost, 5) complication rate, 6) readmission rate, or 7) patient satisfaction.

RESULTS

A final 27 articles were included in the qualitative analysis, with mixed quality of evidence ranging from high in 3 cases to very low in 1 case. Seventeen studies reported a complete ERAS protocol. Preoperative ERAS elements include patient selection through multidisciplinary team discussion, patient counseling and education to adjust expectations of the postoperative period, and mental state assessment; antimicrobial, steroidal, and antiepileptic prophylaxes; nutritional assessment, as well as preoperative oral carbohydrate loading; and postoperative nausea and vomiting (PONV) prophylaxis. Anesthesiology interventions included local anesthesia for pin sites, regional field block or scalp block, avoidance or minimization of the duration of invasive monitoring, and limitation of intraoperative mannitol. Other intraoperative elements include absorbable skin sutures and avoidance of wound drains. Postoperatively, the authors identified early extubation, observation in a step-down unit instead of routine ICU admission, early mobilization, early fluid de-escalation, early intake of solid food and liquids, early removal of invasive monitoring, professional nutritional assessment, PONV management, nonopioid rescue analgesia, and early postoperative imaging. Other postoperative interventions included discharge criteria standardization and home visits or progress monitoring by a nurse.

CONCLUSIONS

A wide range of evidence-based interventions are available to improve recovery after elective craniotomy, although there are few published ERAS protocols. Patient-centered optimization of neurosurgical care spanning the pre-, intra-, and postoperative periods is feasible and has already provided positive results in terms of improved outcomes such as postoperative pain, patient satisfaction, reduced length of stay, and cost reduction with an excellent safety profile. Although fast-track recovery protocols and ERAS studies are gaining momentum for elective craniotomy, prospective trials are needed to provide stronger evidence.

Development of a Complication- and Treatment-Aware Prediction Model for Favorable Functional Outcome in Aneurysmal Subarachnoid Hemorrhage Based on Machine Learning

BACKGROUND

Current prognostic tools in aneurysmal subarachnoid hemorrhage (aSAH) are constrained by being primarily based on patient and disease characteristics on admission.

OBJECTIVE

To develop and validate a complication- and treatment-aware outcome prediction tool in aSAH.

METHODS

This cohort study included data from an ongoing prospective nationwide multicenter registry on all aSAH patients in Switzerland (Swiss SOS [Swiss Study on aSAH]; 2009-2015). We trained supervised machine learning algorithms to predict a binary outcome at discharge (modified Rankin scale [mRS] ≤ 3: favorable; mRS 4-6: unfavorable). Clinical and radiological variables on admission ("Early" Model) as well as additional variables regarding secondary complications and disease management ("Late" Model) were used. Performance of both models was assessed by classification performance metrics on an out-of-sample test dataset.

RESULTS

Favorable functional outcome at discharge was observed in 1156 (62.0%) of 1866 patients. Both models scored a high accuracy of 75% to 76% on the test set. The "Late" outcome model outperformed the "Early" model with an area under the receiver operator characteristics curve (AUC) of 0.85 vs 0.79, corresponding to a specificity of 0.81 vs 0.70 and a sensitivity of 0.71 vs 0.79, respectively.

CONCLUSION

Both machine learning models show good discrimination and calibration confirmed on application to an internal test dataset of patients with a wide range of disease severity treated in different institutions within a nationwide registry. Our study indicates that the inclusion of variables reflecting the clinical course of the patient may lead to outcome predictions with superior predictive power compared to a model based on admission data only.

Association of a prehabilitation program with anxiety and depression before colorectal surgery: a post hoc analysis of the pERACS randomized controlled trial

PURPOSE

Hospital-associated anxiety and depression are major preoperative stressors and common in colorectal cancer surgery and major abdominal surgery. The prehabilitation Enhanced Recovery After Colorectal Surgery (pERACS) study is a single-center, single-blinded randomized controlled trial (RCT) evaluating the effect of a structured prehabilitation program. We evaluate within this RCT the association of a prehabilitation program with anxiety and depression before colorectal surgery.

METHODS

Treatment allocation randomized and single-blinded. Regardless of group allocation, patients were treated according to our institutional Enhanced Recovery After Surgery (ERAS) protocol. Inclusion criteria consisted of adult patients suffering from colorectal disease requiring surgical treatment and who were treated according to the ERAS protocol. Anxiety and depression scores were assessed at baseline and at admission according to the Hospital Anxiety and Depression Scale (HADS), with its subcomponents for depression (HADS-D) and for anxiety (HADS-A).

RESULTS

A total of 23 patients randomized to prehabilitation (mean age: 64.8±11.5 years) and 25 patients randomized to the control group (64.0±11.9 years) were included. There was no statistically significant difference in HADS-Anxiety improvement (Prehabilitation: -1.7±2.8 points vs. control: -0.4±3.4 points, p=0.132). Similarly, the difference in HADS-Depression improvement among the prehabilitation (1.0±2.4 points) and control (-0.3 ± 4.0 points) groups (p = 0.543) was non-significant. Clinically meaningful improvement in anxiety (60.9%/40.0%, p=0.149) and depression (34.8%/20.0%, p=0.250) was similar among the groups.

CONCLUSION

In a post hoc analysis of a randomized trial, prehabilitation had no effect on preoperative reduction of anxiety and depression measures.

TRIAL REGISTRATION NUMBER

Clinicaltrials.gov NCT02746731. Date of registration: April 21, 2016.

Safety of the paramedian supracerebellar-transtentorial approach for selective amygdalohippocampectomy

OBJECTIVE

The goal of this study was to assess the reproducibility and safety of the recently introduced paramedian supracerebellar-transtentorial (PST) approach for selective amygdalohippocampectomy (SA).

METHODS

The authors performed a retrospective analysis of prospectively collected data originating from their surgical register of patients undergoing SA via a PST approach for lesional medial temporal lobe epilepsy. All patients received thorough pre- and postoperative clinical (neurological, neuropsychological, psychiatric) and instrumental (ictal and long-term EEG, invasive EEG if needed, MRI) workup. Surgery-induced complications were assessed at discharge and at every follow-up thereafter and were classified according to Clavien-Dindo grade (CDG). Epilepsy outcome was defined according to Engel classification. Data were reported according to common descriptive statistical methods.

RESULTS

Between May 2015 and May 2018, 17 patients underwent SA via a PST approach at the authors' institution (hippocampal sclerosis in 13 cases, WHO grade II glioma in 2 cases, and reactive gliosis in 2 cases). The median postoperative follow-up was 7 months (mean 9 months, range 3-19 months). There was no surgery-related mortality and no complication (CDG ≥ 2) in the whole series. Transitory CDG 1 surgical complications occurred in 4 patients and had resolved in all of them by the first postoperative follow-up. One patient showed a deterioration of neuropsychological performance with new slight mnestic deficits. No patient experienced a clinically relevant postoperative visual field defect. No morbidity due to semisitting position was recorded. At last follow-up 13/17 (76.4%) patients were in Engel class I (9/17 [52.9%] were in class IA).

CONCLUSIONS

The PST approach is a reproducible and safe surgical route for SA. The achievable complication rate is in line with the best results in the literature. Visual function outcome particularly benefits from this highly selective, neocortex-sparing approach. A larger patient sample and longer follow-up will show in the future if the seizure control rate and neuropsychological outcome also compare better than those achieved with current common surgical techniques.

The association of patient age with postoperative morbidity and mortality following resection of intracranial tumors

Introduction

The postoperative functional status of patients with intracranial tumors is influenced by patient-specific factors, including age.

Research question

This study aimed to elucidate the association between age and postoperative morbidity or mortality following the resection of brain tumors.

Material and methods

A multicenter database was retrospectively reviewed. Functional status was assessed before and 3–6 months after tumor resection by the Karnofsky Performance Scale (KPS). Uni- and multivariable linear regression were used to estimate the association of age with postoperative change in KPS. Logistic regression models for a ≥10-point decline in KPS or mortality were built for patients ≥75 years.

Results

The total sample of 4864 patients had a mean age of 56.4 ​± ​14.4 years. The mean change in pre-to postoperative KPS was −1.43. For each 1-year increase in patient age, the adjusted change in postoperative KPS was −0.11 (95% CI -0.14 - - 0.07). In multivariable analysis, patients ≥75 years had an odds ratio of 1.51 to experience postoperative functional decline (95%CI 1.21–1.88) and an odds ratio of 2.04 to die (95%CI 1.33–3.13), compared to younger patients.

Discussion

Patients with intracranial tumors treated surgically showed a minor decline in their postoperative functional status. Age was associated with this decline in function, but only to a small extent.

Conclusion

Patients ≥75 years were more likely to experience a clinically meaningful decline in function and about two times as likely to die within the first 6 months after surgery, compared to younger patients.

•

A multicenter database of patients with intracranial tumors is analyzed in this study.

•

Age is associated with a minor decline in the postoperative functional status & mortality.

•

Patients ≥75 years are more likely to experience a clinically meaningful decline in function and to die.

Magnetic resonance imaging-based synthetic computed tomography of the lumbar spine for surgical planning: a clinical proof-of-concept

OBJECTIVE

Computed tomography scanning of the lumbar spine incurs a radiation dose ranging from 3.5 mSv to 19.5 mSv as well as relevant costs and is commonly necessary for spinal neuronavigation. Mitigation of the need for treatment-planning CT scans in the presence of MRI facilitated by MRI-based synthetic CT (sCT) would revolutionize navigated lumbar spine surgery. The authors aim to demonstrate, as a proof of concept, the capability of deep learning-based generation of sCT scans from MRI of the lumbar spine in 3 cases and to evaluate the potential of sCT for surgical planning.

METHODS

Synthetic CT reconstructions were made using a prototype version of the "BoneMRI" software. This deep learning-based image synthesis method relies on a convolutional neural network trained on paired MRI-CT data. A specific but generally available 4-minute 3D radiofrequency-spoiled T1-weighted multiple gradient echo MRI sequence was supplemented to a 1.5T lumbar spine MRI acquisition protocol.

RESULTS

In the 3 presented cases, the prototype sCT method allowed voxel-wise radiodensity estimation from MRI, resulting in qualitatively adequate CT images of the lumbar spine based on visual inspection. Normal as well as pathological structures were reliably visualized. In the first case, in which a spiral CT scan was available as a control, a volume CT dose index (CTDIvol) of 12.9 mGy could thus have been avoided. Pedicle screw trajectories and screw thickness were estimable based on sCT findings.

CONCLUSIONS

The evaluated prototype BoneMRI method enables generation of sCT scans from MRI images with only minor changes in the acquisition protocol, with a potential to reduce workflow complexity, radiation exposure, and costs. The quality of the generated CT scans was adequate based on visual inspection and could potentially be used for surgical planning, intraoperative neuronavigation, or for diagnostic purposes in an adjunctive manner.