Using a Machine Learning Approach to Predict Outcomes after Radiosurgery for Cerebral Arteriovenous Malformations

CNS-CLIP: Transforming a Neurosurgical Journal Into a Multimodal Medical Model

BACKGROUND AND OBJECTIVES

Classical biomedical data science models are trained on a single modality and aimed at one specific task. However, the exponential increase in the size and capabilities of the foundation models inside and outside medicine shows a shift toward task-agnostic models using large-scale, often internet-based, data. Recent research into smaller foundation models trained on specific literature, such as programming textbooks, demonstrated that they can display capabilities similar to or superior to large generalist models, suggesting a potential middle ground between small task-specific and large foundation models. This study attempts to introduce a domain-specific multimodal model, Congress of Neurological Surgeons (CNS)-Contrastive Language-Image Pretraining (CLIP), developed for neurosurgical applications, leveraging data exclusively from Neurosurgery Publications.

METHODS

We constructed a multimodal data set of articles from Neurosurgery Publications through PDF data collection and figure-caption extraction using an artificial intelligence pipeline for quality control. Our final data set included 24 021 figure-caption pairs. We then developed a fine-tuning protocol for the OpenAI CLIP model. The model was evaluated on tasks including neurosurgical information retrieval, computed tomography imaging classification, and zero-shot ImageNet classification.

RESULTS

CNS-CLIP demonstrated superior performance in neurosurgical information retrieval with a Top-1 accuracy of 24.56%, compared with 8.61% for the baseline. The average area under receiver operating characteristic across 6 neuroradiology tasks achieved by CNS-CLIP was 0.95, slightly superior to OpenAI's Contrastive Language-Image Pretraining at 0.94 and significantly outperforming a vanilla vision transformer at 0.62. In generalist classification, CNS-CLIP reached a Top-1 accuracy of 47.55%, a decrease from the baseline of 52.37%, demonstrating a catastrophic forgetting phenomenon.

CONCLUSION

This study presents a pioneering effort in building a domain-specific multimodal model using data from a medical society publication. The results indicate that domain-specific models, while less globally versatile, can offer advantages in specialized contexts. This emphasizes the importance of using tailored data and domain-focused development in training foundation models in neurosurgery and general medicine.

The application of machine learning for treatment selection of unruptured brain arteriovenous malformations: A secondary analysis of the ARUBA trial data

OBJECTIVE

To build a supervised machine learning (ML) model that selects the best first-line treatment strategy for unruptured bAVMs.

METHODS

A Randomized Trial of Unruptured Brain Arteriovenous Malformations (ARUBA) trial data was obtained from the National Institute of Neurological Disorders and Stroke (NINDS). A team of five clinicians examined the demographic, clinical, and radiological details of each patient at baseline and reached a consensus on the best first-line treatment for bAVMs. Their treatment choice was used to train an automated supervised ML (autoML) model to select treatments for bAVMs for the training dataset. The accuracy and AUC of the algorithm in selecting the treatment strategy were measured for the test dataset, and feature importance scores of the included variables were calculated.

RESULTS

Among the 100,000 combinations of supervised ML algorithms and their hyperparameters attempted by autoML, gradient boosting classifier had the best predictive performance with an overall accuracy of 0.74 and an area under the curve (AUC) of 0.88. The treatment-specific accuracies were 0.96, 0.85, 0.84, and 0.82; and AUCs were 0.75, 0.95, 0.80, and 0.88 for medical management, surgery, endovascular embolization, and gamma-knife radiosurgery, respectively. Spetzler-Martin score, followed by eloquent AVM location and AVM size, were the three most important features in determining treatments.

CONCLUSION

ML could reliably select the best first-line treatment strategy for bAVMs as per multidisciplinary expert consensus. This study can be replicated for larger population-based AVM registries, with the inclusion of outcome data, thus helping address the bias involved in the management of unruptured bAVMs.

Prediction of implant failure risk due to periprosthetic femoral fracture after primary elective total hip arthroplasty : a simplified and validated model based on 154,519 total hip arthroplasties from the Swedish Arthroplasty Register

Aims

While cementless fixation offers potential advantages over cemented fixation, such as a shorter operating time, concerns linger over its higher cost and increased risk of periprosthetic fractures. If the risk of fracture can be forecasted, it would aid the shared decision-making process related to cementless stems. Our study aimed to develop and validate predictive models of periprosthetic femoral fracture (PPFF) necessitating revision and reoperation after elective total hip arthroplasty (THA).

Methods

We included 154,519 primary elective THAs from the Swedish Arthroplasty Register (SAR), encompassing 21 patient-, surgical-, and implant-specific features, for model derivation and validation in predicting 30-day, 60-day, 90-day, and one-year revision and reoperation due to PPFF. Model performance was tested using the area under the curve (AUC), and feature importance was identified in the best-performing algorithm.

Results

The Lasso regression excelled in predicting 30-day revisions (area under the receiver operating characteristic curve (AUC) = 0.85), while the Gradient Boosting Machine (GBM) model outperformed other models by a slight margin for all remaining endpoints (AUC range: 0.79 to 0.86). Predictive factors for revision and reoperation were identified, with patient features such as increasing age, higher American Society of Anesthesiologists grade (> III), and World Health Organization obesity classes II to III associated with elevated risks. A preoperative diagnosis of idiopathic necrosis increased revision risk. Concerning implant design, factors such as cementless femoral fixation, reverse-hybrid fixation, hip resurfacing, and small (< 35 mm) or large (> 52 mm) femoral heads increased both revision and reoperation risks.

Conclusion

This is the first study to develop machine-learning models to forecast the risk of PPFF necessitating secondary surgery. Future studies are required to externally validate our algorithm and assess its applicability in clinical practice.

Cerebral vascular malformations: pathogenesis and therapy

Cerebral vascular malformations (CVMs), particularly cerebral cavernous malformations and cerebral arteriovenous malformations, pose significant neurological challenges due to their complex etiologies and clinical implications. Traditionally viewed as congenital conditions with structural abnormalities, CVMs have been treated primarily through resection, embolization, and stereotactic radiosurgery. While these approaches offer some efficacy, they often pose risks to neurological integrity due to their invasive nature. Advances in next-generation sequencing, particularly high-depth whole-exome sequencing and bioinformatics, have facilitated the identification of gene variants from neurosurgically resected CVMs samples. These advancements have deepened our understanding of CVM pathogenesis. Somatic mutations in key mechanistic pathways have been identified as causative factors, leading to a paradigm shift in CVM treatment. Additionally, recent progress in noninvasive and minimally invasive techniques, including gene imaging genomics, liquid biopsy, or endovascular biopsies (endovascular sampling of blood vessel lumens), has enabled the identification of gene variants associated with CVMs. These methods, in conjunction with clinical data, offer potential for early detection, dynamic monitoring, and targeted therapies that could be used as monotherapy or adjuncts to surgery. This review highlights advancements in CVM pathogenesis and precision therapies, outlining the future potential of precision medicine in CVM management.

Radiation-induced malignancies after stereotactic radiosurgery for brain arteriovenous malformations: a large single-center retrospective study and systematic review

Stereotactic radiosurgery (SRS) is widely utilized to treat small- and medium-sized brain arteriovenous malformations (BAVMs); however, radiation-induced malignancies (RIMs) have been reported as extremely rare yet potentially life-threatening complications of SRS. This study aimed to investigate the risk of RIMs after SRS for BAVMs. The outcomes of patients who underwent single-session SRS for BAVMs at our institution and were followed for ≥ 5 years were analyzed to calculate the incidence of RIMs. In addition, a systematic review was conducted using the existing literature reporting RIMs after SRS for BAVMs in compliance with the PRISMA guideline. Regarding the in-hospital analysis, only one (0.18%) RIM (gliosarcoma) was observed among 569 patients, with a median follow-up period of 151 months (interquartile range, 103-255 months). The 15, 20, and 25-year cumulative incidences of RIMs were 0%, 0%, and 1.01%, respectively, whereas the overall incidence rate was 0.12 per 1,000 patient-years. In the systematic review, 14 studies were included, with the incidence of RIMs ranging from 0.00 to 0.24%. Eight patients with RIMs were identified, and the most common pathology was glioblastoma. The median time until the diagnosis of RIM was 7.1 years (range, 4-19 years) after SRS, and their clinical courses were largely dismal, with the post-diagnosis survival periods being 1-10 months. RIM constitutes an extremely rare but potentially fatal complication following SRS for BAVMs, with its incidence rate being at most 0.24%.

DSA Quantitative Analysis and Predictive Modeling of Obliteration in Cerebral AVM following Stereotactic Radiosurgery

BACKGROUND AND PURPOSE

Stereotactic radiosurgery is a key treatment modality for cerebral AVMs, particularly for small lesions and those located in eloquent brain regions. Predicting obliteration remains challenging due to evolving treatment paradigms and complex AVM presentations. With digital subtraction angiography (DSA) being the gold standard for outcome evaluation, radiomic approaches offer potential for more objective and detailed analysis. We aimed to develop machine learning modeling using DSA quantitative features for post-SRS obliteration prediction.

MATERIALS AND METHODS

A prospective registry of patients with cerebral AVMs was screened to include patients with digital prestereotactic radiosurgery DSA. Anterior-posterior and lateral views were retrieved and manually segmented. Quantitative features were computed from the lesion ROI. Following feature selection, machine learning models were developed to predict unsuccessful 2-year total obliteration using processed radiomics features in comparison with clinical and radiosurgical features. When we evaluated through area under the receiver operating characteristic curve (AUROC), accuracy, area under the precision-recall curve F1, recall, and precision, the best performing model predictions on the test set were interpreted using the Shapley additive explanations approach.

RESULTS

DSA images of 100 included patients were retrieved and analyzed. The best-performing clinical radiosurgical model was a gradient boosting classifier with an AUROC of 68% and a recall of 67%. When we used radiomics variables as input, the AdaBoost classifier had the best evaluation metrics with an AUROC of 79% and a recall of 75%. The most important clinico-radiosurgical features, ranked by model contribution, were lesion volume, patient age, treatment dose rate, the presence of seizure at presentation, and prior resection. The most important ranked radiomics features were the following: gray-level size zone matrix, gray-level nonuniformity, kurtosis, sphericity, skewness, and gray-level dependence matrix dependence nonuniformity.

CONCLUSIONS

The combination of radiomics with machine learning is a promising approach for predicting cerebral AVM obliteration status following stereotactic radiosurgery. DSA could enhance prognostication of stereotactic radiosurgery-treated AVMs due to its high spatial resolution. Model interpretation is essential for building transparent models and establishing clinically valid radiomic signatures.

The new era of artificial intelligence in neuroradiology: current research and promising tools

Radiology has a number of characteristics that make it an especially suitable medical discipline for early artificial intelligence (AI) adoption. These include having a well-established digital workflow, standardized protocols for image storage, and numerous well-defined interpretive activities. The more than 200 commercial radiologic AI-based products recently approved by the Food and Drug Administration (FDA) to assist radiologists in a number of narrow image-analysis tasks such as image enhancement, workflow triage, and quantification, corroborate this observation. However, in order to leverage AI to boost efficacy and efficiency, and to overcome substantial obstacles to widespread successful clinical use of these products, radiologists should become familiarized with the emerging applications in their particular areas of expertise. In light of this, in this article we survey the existing literature on the application of AI-based techniques in neuroradiology, focusing on conditions such as vascular diseases, epilepsy, and demyelinating and neurodegenerative conditions. We also introduce some of the algorithms behind the applications, briefly discuss a few of the challenges of generalization in the use of AI models in neuroradiology, and skate over the most relevant commercially available solutions adopted in clinical practice. If well designed, AI algorithms have the potential to radically improve radiology, strengthening image analysis, enhancing the value of quantitative imaging techniques, and mitigating diagnostic errors.

Machine learning models on a web application to predict short-term postoperative outcomes following anterior cervical discectomy and fusion

BACKGROUND

The frequency of anterior cervical discectomy and fusion (ACDF) has increased up to 400% since 2011, underscoring the need to preoperatively anticipate adverse postoperative outcomes given the procedure's expanding use. Our study aims to accomplish two goals: firstly, to develop a suite of explainable machine learning (ML) models capable of predicting adverse postoperative outcomes following ACDF surgery, and secondly, to embed these models in a user-friendly web application, demonstrating their potential utility.

METHODS

We utilized data from the National Surgical Quality Improvement Program database to identify patients who underwent ACDF surgery. The outcomes of interest were four short-term postoperative adverse outcomes: prolonged length of stay (LOS), non-home discharges, 30-day readmissions, and major complications. We utilized five ML algorithms - TabPFN, TabNET, XGBoost, LightGBM, and Random Forest - coupled with the Optuna optimization library for hyperparameter tuning. To bolster the interpretability of our models, we employed SHapley Additive exPlanations (SHAP) for evaluating predictor variables' relative importance and used partial dependence plots to illustrate the impact of individual variables on the predictions generated by our top-performing models. We visualized model performance using receiver operating characteristic (ROC) curves and precision-recall curves (PRC). Quantitative metrics calculated were the area under the ROC curve (AUROC), balanced accuracy, weighted area under the PRC (AUPRC), weighted precision, and weighted recall. Models with the highest AUROC values were selected for inclusion in a web application.

RESULTS

The analysis included 57,760 patients for prolonged LOS [11.1% with prolonged LOS], 57,780 for non-home discharges [3.3% non-home discharges], 57,790 for 30-day readmissions [2.9% readmitted], and 57,800 for major complications [1.4% with major complications]. The top-performing models, which were the ones built with the Random Forest algorithm, yielded mean AUROCs of 0.776, 0.846, 0.775, and 0.747 for predicting prolonged LOS, non-home discharges, readmissions, and complications, respectively.

CONCLUSIONS

Our study employs advanced ML methodologies to enhance the prediction of adverse postoperative outcomes following ACDF. We designed an accessible web application to integrate these models into clinical practice. Our findings affirm that ML tools serve as vital supplements in risk stratification, facilitating the prediction of diverse outcomes and enhancing patient counseling for ACDF.

Prediction of Obliteration After the Gamma Knife Radiosurgery of Arteriovenous Malformations Using Hand-Crafted Radiomics and Deep-Learning Methods

INTRODUCTION

Brain arteriovenous malformations (bAVMs) are vascular abnormalities that can be treated with embolization or radiotherapy to prevent the risk of future rupture. In this study, we use hand-crafted radiomics and deep learning techniques to predict favorable vs. unfavorable outcomes following Gamma Knife radiosurgery (GKRS) of bAVMs and compare their prediction performances.

METHODS

One hundred twenty-six patients seen at one academic medical center for GKRS obliteration of bAVMs over 15 years were retrospectively reviewed. Forty-two patients met the inclusion criteria. Favorable outcomes were defined as complete nidus obliteration demonstrated on cerebral angiogram and asymptomatic recovery. Unfavorable outcomes were defined as incomplete obliteration or complications relating to the AVM that developed after GKRS. Outcome predictions were made using a random forest model with hand-crafted radiomic features and a fine-tuned ResNet-34 convolutional neural network (CNN) model. The performance was evaluated by using a ten-fold cross-validation technique.

RESULTS

The average accuracy and area-under-curve (AUC) values of the Random Forest Classifier (RFC) with radiomics features were 68.5 ±9.80% and 0.705 ±0.086, whereas those of the ResNet-34 model were 60.0 ±11.9% and 0.694 ±0.124. Four radiomics features used with RFC discriminated unfavorable response cases from favorable response cases with statistical significance. When cropped images were used with ResNet-34, the accuracy and AUC decreased to 59.3 ± 14.2% and 55.4 ±10.4%, respectively.

CONCLUSIONS

A hand-crafted radiomics model and a pre-trained CNN model can be fine-tuned on pre-treatment MRI scans to predict clinical outcomes of AVM patients undergoing GKRS with equivalent prediction performance. The outcome predictions are promising but require further external validation on more patients.

Prediction of delayed cerebral ischemia followed aneurysmal subarachnoid hemorrhage. A machine-learning based study

INTRODUCTION

Delayed Cerebral Ischemia (DCI) is a significant complication following aneurysmal subarachnoid hemorrhage (aSAH) that can lead to poor outcomes. Machine learning techniques have shown promise in predicting DCI and improving risk stratification.

METHODS

In this study, we aimed to develop machine learning models to predict the occurrence of DCI in patients with aSAH. Patient data, including various clinical variables and co-factors, were collected. Six different machine learning models, including logistic regression, multilayer perceptron, decision tree, random forest, gradient boosting machine, and extreme gradient boosting (XGB), were trained and evaluated using performance metrics such as accuracy, area under the curve (AUC), precision, recall, and F1 score.

RESULTS

After data augmentation, the random forest model demonstrated the best performance, with an AUC of 0.85. The multilayer perceptron neural network model achieved an accuracy of 0.93 and an F1 score of 0.85, making it the best performing model. The presence of positive clinical vasospasm was identified as the most important feature for predicting DCI.

CONCLUSIONS

Our study highlights the potential of machine learning models in predicting the occurrence of DCI in patients with aSAH. The multilayer perceptron model showed excellent performance, indicating its utility in risk stratification and clinical decision-making. However, further validation and refinement of the models are necessary to ensure their generalizability and applicability in real-world settings. Machine learning techniques have the potential to enhance patient care and improve outcomes in aSAH, but their implementation should be accompanied by careful evaluation and validation.

Application of artificial intelligence in brain arteriovenous malformations: Angioarchitectures, clinical symptoms and prognosis prediction

BACKGROUND

Artificial intelligence (AI) has rapidly advanced in the medical field, leveraging its intelligence and automation for the management of various diseases. Brain arteriovenous malformations (AVM) are particularly noteworthy, experiencing rapid development in recent years and yielding remarkable results. This paper aims to summarize the applications of AI in the management of AVMs management.

METHODS

Literatures published in PubMed during 1999-2022, discussing AI application in AVMs management were reviewed.

RESULTS

AI algorithms have been applied in various aspects of AVM management, particularly in machine learning and deep learning models. Automatic lesion segmentation or delineation is a promising application that can be further developed and verified. Prognosis prediction using machine learning algorithms with radiomic-based analysis is another meaningful application.

CONCLUSIONS

AI has been widely used in AVMs management. This article summarizes the current research progress, limitations and future research directions.

Invertible and Variable Augmented Network for Pretreatment Patient-Specific Quality Assurance Dose Prediction

Pretreatment patient-specific quality assurance (prePSQA) is conducted to confirm the accuracy of the radiotherapy dose delivered. However, the process of prePSQA measurement is time consuming and exacerbates the workload for medical physicists. The purpose of this work is to propose a novel deep learning (DL) network to improve the accuracy and efficiency of prePSQA. A modified invertible and variable augmented network was developed to predict the three-dimensional (3D) measurement-guided dose (MDose) distribution of 300 cancer patients who underwent volumetric modulated arc therapy (VMAT) between 2018 and 2021, in which 240 cases were randomly selected for training, and 60 for testing. For simplicity, the present approach was termed as "IVPSQA." The input data include CT images, radiotherapy dose exported from the treatment planning system, and MDose distribution extracted from the verification system. Adam algorithm was used for first-order gradient-based optimization of stochastic objective functions. The IVPSQA model obtained high-quality 3D prePSQA dose distribution maps in head and neck, chest, and abdomen cases, and outperformed the existing U-Net-based prediction approaches in terms of dose difference maps and horizontal profiles comparison. Moreover, quantitative evaluation metrics including SSIM, MSE, and MAE demonstrated that the proposed approach achieved a good agreement with ground truth and yield promising gains over other advanced methods. This study presented the first work on predicting 3D prePSQA dose distribution by using the IVPSQA model. The proposed method could be taken as a clinical guidance tool and help medical physicists to reduce the measurement work of prePSQA.

Current perspectives and trends in the treatment of brain arteriovenous malformations: a review and bibliometric analysis

Background

Currently, there is a lack of intuitive analysis regarding the development trend, main authors, and research hotspots in the field of cerebral arteriovenous malformation treatment, as well as a detailed elaboration of possible research hotspots.

Methods

A bibliometric analysis was conducted on data retrieved from the Web of Science core collection database between 2000 and 2022. The analysis was performed using R, VOSviewer, CiteSpace software, and an online bibliometric platform.

Results

A total of 1,356 articles were collected, and the number of publications has increased over time. The United States and the University of Pittsburgh are the most prolific countries and institutions in the field. The top three cited authors are Kondziolka D, Sheehan JP, and Lunsford LD. The Journal of Neurosurgery and Neurosurgery are two of the most influential journals in the field of brain arteriovenous malformation treatment research, with higher H-index, total citations, and number of publications. Furthermore, the analysis of keywords indicates that "aruba trial," "randomised trial," "microsurgery," "onyx embolization," and "Spetzler-Martin grade" may become research focal points. Additionally, this paper discusses the current research status, existing issues, and potential future research directions for the treatment of brain arteriovenous malformations.

Conclusion

This bibliometric study comprehensively analyses the publication trend of cerebral arteriovenous malformation treatment in the past 20 years. It covers the trend of international cooperation, publications, and research hotspots. This information provides an important reference for scholars to further study cerebral arteriovenous malformation.

Machine Learning Approach for Analyzing 3-Year Outcomes of Patients with Brain Arteriovenous Malformation (AVM) after Stereotactic Radiosurgery (SRS)

A cerebral arteriovenous malformation (AVM) is a tangle of abnormal blood vessels that irregularly connects arteries and veins. Stereotactic radiosurgery (SRS) has been shown to be an effective treatment for AVM patients, but the factors associated with AVM obliteration remains a matter of debate. In this study, we aimed to develop a model that can predict whether patients with AVM will be cured 36 months after intervention by means of SRS and identify the most important predictors that explain the probability of being cured. A machine learning (ML) approach was applied using decision tree (DT) and logistic regression (LR) techniques on historical data (sociodemographic, clinical, treatment, angioarchitecture, and radiosurgery procedure) of 202 patients with AVM who underwent SRS at the Instituto de Radiocirugía del Perú (IRP) between 2005 and 2018. The LR model obtained the best results for predicting AVM cure with an accuracy of 0.92, sensitivity of 0.93, specificity of 0.89, and an area under the curve (AUC) of 0.98, which shows that ML models are suitable for predicting the prognosis of medical conditions such as AVM and can be a support tool for medical decision-making. In addition, several factors were identified that could explain whether patients with AVM would be cured at 36 months with the highest likelihood: the location of the AVM, the occupation of the patient, and the presence of hemorrhage.

Developing an Automated Registry (Autoregistry) of Spine Surgery Using Natural Language Processing and Health System Scale Databases

BACKGROUND AND OBJECTIVES

Clinical registries are critical for modern surgery and underpin outcomes research, device monitoring, and trial development. However, existing approaches to registry construction are labor-intensive, costly, and prone to manual error. Natural language processing techniques combined with electronic health record (EHR) data sets can theoretically automate the construction and maintenance of registries. Our aim was to automate the generation of a spine surgery registry at an academic medical center using regular expression (regex) classifiers developed by neurosurgeons to combine domain expertise with interpretable algorithms.

METHODS

We used a Hadoop data lake consisting of all the information generated by an academic medical center. Using this database and structured query language queries, we retrieved every operative note written in the department of neurosurgery since our transition to EHR. Notes were parsed using regex classifiers and compared with a random subset of 100 manually reviewed notes.

RESULTS

A total of 31 502 operative cases were downloaded and processed using regex classifiers. The codebase required 5 days of development, 3 weeks of validation, and less than 1 hour for the software to generate the autoregistry. Regex classifiers had an average accuracy of 98.86% at identifying both spinal procedures and the relevant vertebral levels, and it correctly identified the entire list of defined surgical procedures in 89% of patients. We were able to identify patients who required additional operations within 30 days to monitor outcomes and quality metrics.

CONCLUSION

This study demonstrates the feasibility of automatically generating a spine registry using the EHR and an interpretable, customizable natural language processing algorithm which may reduce pitfalls associated with manual registry development and facilitate rapid clinical research.

A machine learning approach to predict the glaucoma filtration surgery outcome

This study aimed at predicting the filtration surgery (FS) outcome using a machine learning (ML) approach. 102 glaucomatous patients undergoing FS were enrolled and underwent ocular surface clinical tests (OSCTs), determination of surgical site-related biometric parameters (SSPs) and conjunctival vascularization. Break-up-time, Schirmer test I, corneal fluorescein staining, Meibomian gland expressibility; conjunctival hyperemia, upper bulbar conjunctiva area of exposure, limbus to superior eyelid distance; and conjunctival epithelial and stromal (CET, CST) thickness and reflectivity (ECR, SCR) at AS-OCT were considered. Successful FS required a 30% baseline intraocular pressure reduction, with values ≤ 18 mmHg with or without medications. The classification tree (CT) was the ML algorithm used to analyze data. At the twelfth month, FS was successful in 60.8% of cases, whereas failed in 39.2%. At the variable importance ranking, CST and SCR were the predictors with the greater relative importance to the CART tree construction, followed by age. CET and ECR showed less relative importance, whereas OSCTs and SSPs were not important features. Within the CT, CST turned out the most important variable for discriminating success from failure, followed by SCR and age, with cut-off values of 75 µm, 169 on gray scale, and 62 years, respectively. The ROC curve for the classifier showed an AUC of 0.784 (0.692-0.860). In this ML approach, CT analysis found that conjunctival stroma thickness and reflectivity, along with age, can predict the FS outcome with good accuracy. A pre-operative thick and hyper-reflective stroma, and a younger age increase the risk of FS failure.

Using Machine Learning to Predict Nonachievement of Clinically Significant Outcomes After Rotator Cuff Repair

Background

Although some evidence suggests that machine learning algorithms may outperform classical statistical methods in prognosis prediction for several orthopaedic surgeries, to our knowledge, no study has yet used machine learning to predict patient-reported outcome measures after rotator cuff repair.

Purpose

To determine whether machine learning algorithms using preoperative data can predict the nonachievement of the minimal clinically important difference (MCID) of disability at 2 years after rotator cuff surgical repair with a similar performance to that of other machine learning studies in the orthopaedic surgery literature.

Study Design

Case-control study; Level of evidence, 3.

Methods

We evaluated 474 patients (n = 500 shoulders) with rotator cuff tears who underwent arthroscopic rotator cuff repair between January 2013 and April 2019. The study outcome was the difference between the preoperative and 24-month postoperative American Shoulder and Elbow Surgeons (ASES) score. A cutoff score was calculated based on the established MCID of 15.2 points to separate success (higher than the cutoff) from failure (lower than the cutoff). Routinely collected imaging, clinical, and demographic data were used to train 8 machine learning algorithms (random forest classifier; light gradient boosting machine [LightGBM]; decision tree classifier; extra trees classifier; logistic regression; extreme gradient boosting [XGBoost]; k-nearest neighbors [KNN] classifier; and CatBoost classifier). We used a random sample of 70% of patients to train the algorithms, and 30% were left for performance assessment, simulating new data. The performance of the models was evaluated with the area under the receiver operating characteristic curve (AUC).

Results

The AUCs for all algorithms ranged from 0.58 to 0.68. The random forest classifier and LightGBM presented the highest AUC values (0.68 [95% CI, 0.48-0.79] and 0.67 [95% CI, 0.43-0.75], respectively) of the 8 machine learning algorithms. Most of the machine learning algorithms outperformed logistic regression (AUC, 0.59 [95% CI, 0.48-0.81]); nonetheless, their performance was lower than that of other machine learning studies in the orthopaedic surgery literature.

Conclusion

Machine learning algorithms demonstrated some ability to predict the nonachievement of the MCID on the ASES 2 years after rotator cuff repair surgery.

Role of artificial intelligence and machine learning in the diagnosis of cerebrovascular disease

Introduction

Cerebrovascular diseases are known to cause significant morbidity and mortality to the general population. In patients with cerebrovascular disease, prompt clinical evaluation and radiographic interpretation are both essential in optimizing clinical management and in triaging patients for critical and potentially life-saving neurosurgical interventions. With recent advancements in the domains of artificial intelligence (AI) and machine learning (ML), many AI and ML algorithms have been developed to further optimize the diagnosis and subsequent management of cerebrovascular disease. Despite such advances, further studies are needed to substantively evaluate both the diagnostic accuracy and feasibility of these techniques for their application in clinical practice. This review aims to analyze the current use of AI and MI algorithms in the diagnosis of, and clinical decision making for cerebrovascular disease, and to discuss both the feasibility and future applications of utilizing such algorithms.

Methods

We review the use of AI and ML algorithms to assist clinicians in the diagnosis and management of ischemic stroke, hemorrhagic stroke, intracranial aneurysms, and arteriovenous malformations (AVMs). After identifying the most widely used algorithms, we provide a detailed analysis of the accuracy and effectiveness of these algorithms in practice.

Results

The incorporation of AI and ML algorithms for cerebrovascular patients has demonstrated improvements in time to detection of intracranial pathologies such as intracerebral hemorrhage (ICH) and infarcts. For ischemic and hemorrhagic strokes, commercial AI software platforms such as RapidAI and Viz.AI have bene implemented into routine clinical practice at many stroke centers to expedite the detection of infarcts and ICH, respectively. Such algorithms and neural networks have also been analyzed for use in prognostication for such cerebrovascular pathologies. These include predicting outcomes for ischemic stroke patients, hematoma expansion, risk of aneurysm rupture, bleeding of AVMs, and in predicting outcomes following interventions such as risk of occlusion for various endovascular devices. Preliminary analyses have yielded promising sensitivities when AI and ML are used in concert with imaging modalities and a multidisciplinary team of health care providers.

Conclusion

The implementation of AI and ML algorithms to supplement clinical practice has conferred a high degree of accuracy, efficiency, and expedited detection in the clinical and radiographic evaluation and management of ischemic and hemorrhagic strokes, AVMs, and aneurysms. Such algorithms have been explored for further purposes of prognostication for these conditions, with promising preliminary results. Further studies should evaluate the longitudinal implementation of such techniques into hospital networks and residency programs to supplement clinical practice, and the extent to which these techniques improve patient care and clinical outcomes in the long-term.

Development of machine learning models for predicting outcome in patients with distal medium vessel occlusions: a retrospective study

Background

While numerous prognostic factors have been reported for large vessel occlusion (LVO)-acute ischemic stroke (AIS) patients, the same cannot be said for distal medium vessel occlusions (DMVOs). We used machine learning (ML) algorithms to develop a model predicting the short-term outcome of AIS patients with DMVOs using demographic, clinical, and laboratory variables and baseline computed tomography (CT) perfusion (CTP) postprocessing quantitative parameters.

Methods

In this retrospective cohort study, consecutive patients with AIS admitted to two comprehensive stroke centers between January 1, 2017, and September 1, 2022, were screened. Demographic, clinical, and radiological data were extracted from electronic medical records. The clinical outcome was divided into two categories, with a cut-off defined by the median National Institutes of Health Stroke Scale (NIHSS) shift score. Data preprocessing involved addressing missing values through imputation, scaling with a robust scaler, normalization using min-max normalization, and encoding of categorical variables. The data were split into training and test sets (70:30), and recursive feature elimination (RFE) was employed for feature selection. For ML analyses, XGBoost, LightGBM, CatBoost, multi-layer perceptron, random forest, and logistic regression algorithms were utilized. Performance evaluation involved the receiver operating characteristic (ROC) curve, precision-recall curve (PRC), the area under these curves, accuracy, precision, recall, and Matthews correlation coefficient (MCC). The relative weights of predictor variables were examined using Shapley additive explanations (SHAP).

Results

Sixty-nine patients were included and divided into two groups: 35 patients with favorable outcomes and 34 patients with unfavorable outcomes. Utilizing ten selected features, the XGBoost algorithm achieved the best performance in predicting unfavorable outcomes, with an area under the ROC curve (AUROC) of 0.894 and an area under the PRC curve (AUPRC) of 0.756. The SHAP analysis ranked the following features in order of importance for the XGBoost model: mismatch volume, time-to-maximum of the tissue residue function (Tmax) >6 s, diffusion-weighted imaging (DWI) volume, neutrophil-to-platelet ratio (NPR), mean corpuscular volume (MCV), Tmax >10 s, hemoglobin, potassium, hypoperfusion index (HI), and Tmax >8 s.

Conclusions

Our ML models, trained on baseline quantitative laboratory and CT parameters, accurately predicted the short-term outcome in patients with DMVOs. These findings may aid clinicians in predicting prognosis and may be helpful for future research.

Preoperative flow analysis of arteriovenous malformations and obliteration response after stereotactic radiosurgery

OBJECTIVE

Morphological and angioarchitectural features of cerebral arteriovenous malformations (AVMs) have been widely described and associated with outcomes; however, few studies have conducted a quantitative analysis of AVM flow. The authors examined brain AVM flow and transit time on angiograms using direct visual analysis and a computer-based method and correlated these factors with the obliteration response after Gamma Knife radiosurgery.

METHODS

A retrospective analysis was conducted at a single institution using a prospective registry of patients managed from January 2013 to December 2019: 71 patients were analyzed using a visual method of flow determination and 38 were analyzed using a computer-based method. After comparison and validation of the two methods, obliteration response was correlated to flow analysis, demographic, angioarchitectural, and dosimetric data.

RESULTS

The mean AVM volume was 3.84 cm3 (range 0.64-19.8 cm3), 32 AVMs (45%) were in critical functional locations, and the mean margin radiosurgical dose was 18.8 Gy (range 16-22 Gy). Twenty-seven AVMs (38%) were classified as high flow, 37 (52%) as moderate flow, and 7 (10%) as low flow. Complete obliteration was achieved in 44 patients (62%) at the time of the study; the mean time to obliteration was 28 months for low-flow, 34 months for moderate-flow, and 47 months for high-flow AVMs. Univariate and multivariate analyses of factors predicting obliteration included AVM nidus volume, age, and flow. Adverse radiation effects were identified in 5 patients (7%), and 67 patients (94%) remained free of any functional deterioration during follow-up.

CONCLUSIONS

AVM flow analysis and categorization in terms of transit time are useful predictors of the probability of and the time to obliteration. The authors believe that a more quantitative understanding of flow can help to guide stereotactic radiosurgery treatment and set accurate outcome expectations.

Machine Learning Models to Forecast Outcomes of Pituitary Surgery: A Systematic Review in Quality of Reporting and Current Evidence

Background: The complex nature and heterogeneity involving pituitary surgery results have increased interest in machine learning (ML) applications for prediction of outcomes over the last decade. This study aims to systematically review the characteristics of ML models involving pituitary surgery outcome prediction and assess their reporting quality. Methods: We searched the PubMed, Scopus, and Web of Knowledge databases for publications on the use of ML to predict pituitary surgery outcomes. We used the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) to assess report quality. Our search strategy was based on the terms “artificial intelligence”, “machine learning”, and “pituitary”. Results: 20 studies were included in this review. The principal models reported in each article were post-surgical endocrine outcomes (n = 10), tumor management (n = 3), and intra- and postoperative complications (n = 7). Overall, the included studies adhered to a median of 65% (IQR = 60–72%) of TRIPOD criteria, ranging from 43% to 83%. The median reported AUC was 0.84 (IQR = 0.80–0.91). The most popular algorithms were support vector machine (n = 5) and random forest (n = 5). Only two studies reported external validation and adherence to any reporting guideline. Calibration methods were not reported in 15 studies. No model achieved the phase of actual clinical applicability. Conclusion: Applications of ML in the prediction of pituitary outcomes are still nascent, as evidenced by the lack of any model validated for clinical practice. Although studies have demonstrated promising results, greater transparency in model development and reporting is needed to enable their use in clinical practice. Further adherence to reporting guidelines can help increase AI’s real-world utility and improve clinical practice.

A Machine Learning-Based Online Prediction Tool for Predicting Short-Term Postoperative Outcomes Following Spinal Tumor Resections

Background: Preoperative prediction of short-term postoperative outcomes in spinal tumor patients can lead to more precise patient care plans that reduce the likelihood of negative outcomes. With this study, we aimed to develop machine learning algorithms for predicting short-term postoperative outcomes and implement these models in an open-source web application. Methods: Patients who underwent surgical resection of spinal tumors were identified using the American College of Surgeons, National Surgical Quality Improvement Program. Three outcomes were predicted: prolonged length of stay (LOS), nonhome discharges, and major complications. Four machine learning algorithms were developed and integrated into an open access web application to predict these outcomes. Results: A total of 3073 patients that underwent spinal tumor resection were included in the analysis. The most accurately predicted outcomes in terms of the area under the receiver operating characteristic curve (AUROC) was the prolonged LOS with a mean AUROC of 0.745 The most accurately predicting algorithm in terms of AUROC was random forest, with a mean AUROC of 0.743. An open access web application was developed for getting predictions for individual patients based on their characteristics and this web application can be accessed here: huggingface.co/spaces/MSHS-Neurosurgery-Research/NSQIP-ST. Conclusion: Machine learning approaches carry significant potential for the purpose of predicting postoperative outcomes following spinal tumor resections. Development of predictive models as clinically useful decision-making tools may considerably enhance risk assessment and prognosis as the amount of data in spinal tumor surgery continues to rise.

Artificial Intelligence-Assisted Evaluation of the Spatial Relationship between Brain Arteriovenous Malformations and the Corticospinal Tract to Predict Postsurgical Motor Defects

BACKGROUND AND PURPOSE

Preoperative evaluation of brain AVMs is crucial for the selection of surgical candidates. Our goal was to use artificial intelligence to predict postsurgical motor defects in patients with brain AVMs involving motor-related areas.

MATERIALS AND METHODS

Eighty-three patients who underwent microsurgical resection of brain AVMs involving motor-related areas were retrospectively reviewed. Four artificial intelligence-based indicators were calculated with artificial intelligence on TOF-MRA and DTI, including FN5mm/50mm (the proportion of fiber numbers within 5-50mm from the lesion border), FN10mm/50mm (the same but within 10-50mm), FP5mm/50mm (the proportion of fiber voxel points within 5-50mm from the lesion border), and FP10mm/50mm (the same but within 10-50mm). The association between the variables and long-term postsurgical motor defects was analyzed using univariate and multivariate analyses. Least absolute shrinkage and selection operator regression with the Pearson correlation coefficient was used to select the optimal features to develop the machine learning model to predict postsurgical motor defects. The area under the curve was calculated to evaluate the predictive performance.

RESULTS

In patients with and without postsurgical motor defects, the mean FN5mm/50mm, FN10mm/50mm, FP5mm/50mm, and FP10mm/50mm were 0.24 (SD, 0.24) and 0.03 (SD, 0.06), 0.37 (SD, 0.27) and 0.06 (SD, 0.08), 0.06 (SD, 0.10) and 0.01 (SD, 0.02), and 0.10 (SD, 0.12) and 0.02 (SD, 0.05), respectively. Univariate and multivariate logistic analyses identified FN10mm/50mm as an independent risk factor for long-term postsurgical motor defects (P = .002). FN10mm/50mm achieved a mean area under the curve of 0.86 (SD, 0.08). The mean area under the curve of the machine learning model consisting of FN10mm/50mm, diffuseness, and the Spetzler-Martin score was 0.88 (SD, 0.07).

CONCLUSIONS

The artificial intelligence-based indicator, FN10mm/50mm, can reflect the lesion-fiber spatial relationship and act as a dominant predictor for postsurgical motor defects in patients with brain AVMs involving motor-related areas.

Machine Learning-Enabled Determination of Diffuseness of Brain Arteriovenous Malformations from Magnetic Resonance Angiography

The diffuseness of brain arteriovenous malformations (bAVMs) is a significant factor in surgical outcome evaluation and hemorrhagic risk prediction. However, there are still predicaments in identifying diffuseness, such as the judging variety resulting from different experience and difficulties in quantification. The purpose of this study was to develop a machine learning (ML) model to automatically identify the diffuseness of bAVM niduses using three-dimensional (3D) time-of-flight magnetic resonance angiography (TOF-MRA) images. A total of 635 patients with bAVMs who underwent TOF-MRA imaging were enrolled. Three experienced neuroradiologists delineated the bAVM lesions and identified the diffuseness on TOF-MRA images, which were considered the ground-truth reference. The U-Net-based segmentation model was trained to segment lesion areas. Eight mainstream ML models were trained through the radiomic features of segmented lesions to identify diffuseness, based on which an integrated model was built and yielded the best performance. In the test set, the Dice score, F2 score, precision, and recall for the segmentation model were 0.80 [0.72-0.84], 0.80 [0.71-0.86], 0.84 [0.77-0.93], and 0.82 [0.69-0.89], respectively. For the diffuseness identification model, the ensemble-based model was applied with an area under the Receiver-operating characteristic curves (AUC) of 0.93 (95% CI 0.87-0.99) in the training set. The AUC, accuracy, precision, recall, and F1 score for the diffuseness identification model were 0.95, 0.90, 0.81, 0.84, and 0.83, respectively, in the test set. The ML models showed good performance in automatically detecting bAVM lesions and identifying diffuseness. The method may help to judge the diffuseness of bAVMs objectively, quantificationally, and efficiently.

A stacking ensemble machine learning model to predict alpha-1 antitrypsin deficiency-associated liver disease clinical outcomes based on UK Biobank data

Alpha-1 antitrypsin deficiency associated liver disease (AATD-LD) is a rare genetic disorder and not well-recognized. Predicting the clinical outcomes of AATD-LD and defining patients more likely to progress to advanced liver disease are crucial for better understanding AATD-LD progression and promoting timely medical intervention. We aimed to develop a tailored machine learning (ML) model to predict the disease progression of AATD-LD. This analysis was conducted through a stacking ensemble learning model by combining five different ML algorithms with 58 predictor variables using nested five-fold cross-validation with repetitions based on the UK Biobank data. Performance of the model was assessed through prediction accuracy, area under the receiver operating characteristic (AUROC), and area under the precision-recall curve (AUPRC). The importance of predictor contributions was evaluated through a feature importance permutation method. The proposed stacking ensemble ML model showed clinically meaningful accuracy and appeared superior to any single ML algorithms in the ensemble, e.g., the AUROC for AATD-LD was 68.1%, 75.9%, 91.2%, and 67.7% for all-cause mortality, liver-related death, liver transplant, and all-cause mortality or liver transplant, respectively. This work supports the use of ML to address the unanswered clinical questions with clinically meaningful accuracy using real-world data.

A SuperLearner Approach to Predict Run-In Selection in Clinical Trials

A critical early step in a clinical trial is defining the study sample that appropriately represents the target population from which the sample will be drawn. Envisaging a “run-in” process in study design may accomplish this task; however, the traditional run-in requires additional patients, increasing times, and costs. The possible use of the available a-priori data could skip the run-in period. In this regard, ML (machine learning) techniques, which have recently shown considerable promising usage in clinical research, can be used to construct individual predictions of therapy response probability conditional on patient characteristics. An ensemble model of ML techniques was trained and validated on twin randomized clinical trials to mimic a run-in process within this framework. An ensemble ML model composed of 26 algorithms was trained on the twin clinical trials. SuperLearner (SL) performance for the Verum (Treatment) arm is above 70% sensitivity. The Positive Predictive Value (PPP) achieves a value of 80%. Results show good performance in the direction of being useful in the simulation of the run-in period; the trials conducted in similar settings can train an optimal patient selection algorithm minimizing the run-in time and costs of conduction.

Machine learning for predicting hemorrhage in pediatric patients with brain arteriovenous malformation

OBJECTIVE

Ruptured brain arteriovenous malformations (bAVMs) in a child are associated with substantial morbidity and mortality. Prior studies investigating predictors of hemorrhagic presentation of a bAVM during childhood are limited. Machine learning (ML), which has high predictive accuracy when applied to large data sets, can be a useful adjunct for predicting hemorrhagic presentation. The goal of this study was to use ML in conjunction with a traditional regression approach to identify predictors of hemorrhagic presentation in pediatric patients based on a retrospective cohort study design.

METHODS

Using data obtained from 186 pediatric patients over a 19-year study period, the authors implemented three ML algorithms (random forest models, gradient boosted decision trees, and AdaBoost) to identify features that were most important for predicting hemorrhagic presentation. Additionally, logistic regression analysis was used to ascertain significant predictors of hemorrhagic presentation as a comparison.

RESULTS

All three ML models were consistent in identifying bAVM size and patient age at presentation as the two most important factors for predicting hemorrhagic presentation. Age at presentation was not identified as a significant predictor of hemorrhagic presentation in multivariable logistic regression. Gradient boosted decision trees/AdaBoost and random forest models identified bAVM location and a concurrent arterial aneurysm as the third most important factors, respectively. Finally, logistic regression identified a left-sided bAVM, small bAVM size, and the presence of a concurrent arterial aneurysm as significant risk factors for hemorrhagic presentation.

CONCLUSIONS

By using an ML approach, the authors found predictors of hemorrhagic presentation that were not identified using a conventional regression approach.

A Review of Artificial Intelligence in Cerebrovascular Disease Imaging: Applications and Challenges

BACKGROUND

A variety of emerging medical imaging technologies based on artificial intelligence have been widely applied in many diseases, but they are still limitedly used in the cerebrovascular field even though the diseases can lead to catastrophic consequences.

OBJECTIVE

This work aims to discuss the current challenges and future directions of artificial intelligence technology in cerebrovascular diseases through reviewing the existing literature related to applications in terms of computer-aided detection, prediction and treatment of cerebrovascular diseases.

METHODS

Based on artificial intelligence applications in four representative cerebrovascular diseases including intracranial aneurysm, arteriovenous malformation, arteriosclerosis and moyamoya disease, this paper systematically reviews studies published between 2006 and 2021 in five databases: National Center for Biotechnology Information, Elsevier Science Direct, IEEE Xplore Digital Library, Web of Science and Springer Link. And three refinement steps were further conducted after identifying relevant literature from these databases.

RESULTS

For the popular research topic, most of the included publications involved computer-aided detection and prediction of aneurysms, while studies about arteriovenous malformation, arteriosclerosis and moyamoya disease showed an upward trend in recent years. Both conventional machine learning and deep learning algorithms were utilized in these publications, but machine learning techniques accounted for a larger proportion.

CONCLUSION

Algorithms related to artificial intelligence, especially deep learning, are promising tools for medical imaging analysis and will enhance the performance of computer-aided detection, prediction and treatment of cerebrovascular diseases.

A Machine Learning Model Predicts the Outcome of SRS for Residual Arteriovenous malformations after partial embolization- A Real-World Clinical Obstacle

PURPOSE

To propose a machine learning (ML) model predicting the favorable outcome of stereotactic radiosurgery (SRS) for residual brain arteriovenous malformation (bAVM) after partial embolization.

MATERIALS AND METHODS

One hundred and thirty bAVM patients who underwent partial embolization followed by SRS were retrospectively reviewed. Patients were randomly split into training datasets (n=100) and testing datasets (n=30). Radiomics and dosimetric features were extracted from pre-SRS treatment images. Feature selection was performed to select appropriate radiomics and dosimetric features. Three ML algorithms were applied to construct models using selected features respectively. A total of 9 models were trained to predict favorable outcomes (obliteration without complication) of bAVMs. The efficacy of these models was evaluated on the testing dataset using mean accuracy (ACC) and area under the receiver operating characteristic curve (AUROC).

RESULTS

The obliteration rate of this cohort was 70.77% (92/130) with a mean follow-up period of 43.8 (Range 12-108 months) months. Favorable outcomes were achieved in 89 (68.46%) patients. Four radiomics features and 7 dosimetric features were selected for ML model construction. The dosimetric SVM showed the best performance on the training dataset, with an ACC and AUC of 0.74 and 0.78 respectively. The dosimetric SVM model also showed the best performance on the testing dataset where the ACC and AUC were 0.83 and 0.77 respectively.

CONCLUSION

Dosimetric features are good predictors of prognosis for patients with partially embolized bAVM followed by SRS therapy. The use of ML models is an innovative method for predicting favorable outcomes of partially embolized bAVM followed by SRS therapy.

Evaluation of machine learning algorithms for trabeculectomy outcome prediction in patients with glaucoma

The purpose of this study was to evaluate the performance of machine learning algorithms to predict trabeculectomy surgical outcomes. Preoperative systemic, demographic and ocular data from consecutive trabeculectomy surgeries from a single academic institution between January 2014 and December 2018 were incorporated into models using random forest, support vector machine, artificial neural networks and multivariable logistic regression. Mean area under the receiver operating characteristic curve (AUC) and accuracy were used to evaluate the discrimination of each model to predict complete success of trabeculectomy surgery at 1 year. The top performing model was optimized using recursive feature selection and hyperparameter tuning. Calibration and net benefit of the final models were assessed. Among the 230 trabeculectomy surgeries performed on 184 patients, 104 (45.2%) were classified as complete success. Random forest was found to be the top performing model with an accuracy of 0.68 and AUC of 0.74 using 5-fold cross-validation to evaluate the final optimized model. These results provide evidence that machine learning models offer value in predicting trabeculectomy outcomes in patients with refractory glaucoma.

Potential of machine learning to predict early ischemic events after carotid endarterectomy or stenting: a comparison with surgeon predictions

Carotid endarterectomy (CEA) and carotid artery stenting (CAS) are recommended for high stroke-risk patients with carotid artery stenosis to reduce ischemic events. However, we often face difficulty in determining the best treatment strategy. We aimed to develop an accurate post-CEA/CAS outcome prediction model using machine learning that will serve as a basis for a new decision support tool for patient-specific treatment planning. Retrospectively collected data from 165 consecutive patients with carotid stenosis underwent CEA or CAS and were divided into training and test samples. The following five machine learning algorithms were tuned, and their predictive performance was evaluated by comparison with surgeon predictions: an artificial neural network, logistic regression, support vector machine, random forest, and extreme gradient boosting (XGBoost). Seventeen clinical factors were introduced into the models. Outcome was defined as any ischemic stroke within 30 days after treatment including asymptomatic diffusion-weighted imaging abnormalities. The XGBoost model performed the best in the evaluation; its sensitivity, specificity, positive predictive value, and accuracy were 31.9%, 94.6%, 47.2%, and 86.2%, respectively. These statistical measures were comparable to those of surgeons. Internal carotid artery peak systolic velocity, low-density lipoprotein cholesterol, and procedure (CEA or CAS) were the most contributing factors according to the XGBoost algorithm. We were able to develop a post-procedural outcome prediction model comparable to surgeons in performance. The accurate outcome prediction model will make it possible to make a more appropriate patient-specific selection of CEA or CAS for the treatment of carotid artery stenosis.

Machine learning for predicting long-term kidney allograft survival: a scoping review

Supervised machine learning (ML) is a class of algorithms that "learn" from existing input-output pairs, which is gaining popularity in pattern recognition for classification and prediction problems. In this scoping review, we examined the use of supervised ML algorithms for the prediction of long-term allograft survival in kidney transplant recipients. Data sources included PubMed, the Cumulative Index to Nursing and Allied Health Literature, and the Institute for Electrical and Electronics Engineers (IEEE) Xplore libraries from inception to November 2019. We screened titles and abstracts and potentially eligible full-text reports to select studies and subsequently abstracted the data. Eleven studies were identified. Decision trees were the most commonly used method (n = 8), followed by artificial neural networks (ANN) (n = 4) and Bayesian belief networks (n = 2). The area under receiver operating curve (AUC) was the most common measure of discrimination (n = 7), followed by sensitivity (n = 5) and specificity (n = 4). Model calibration examining the reliability in risk prediction was performed using either the Pearson r or the Hosmer-Lemeshow test in four studies. One study showed that logistic regression had comparable performance to ANN, while another study demonstrated that ANN performed better in terms of sensitivity, specificity, and accuracy, as compared with a Cox proportional hazards model. We synthesized the evidence related to the comparison of ML techniques with traditional statistical approaches for prediction of long-term allograft survival in patients with a kidney transplant. The methodological and reporting quality of included studies was poor. Our study also demonstrated mixed results in terms of the predictive potential of the models.

Quantification of hemodynamics of cerebral arteriovenous malformations after stereotactic radiosurgery using 4D flow magnetic resonance imaging

Stereotactic radiosurgery (SRS) is used to treat cerebral arteriovenous malformations (AVMs). However, early evaluation of efficacy is difficult as structural magnetic resonance imaging (MRI)/magnetic resonance angiography (MRA) often does not demonstrate appreciable changes within the first 6 months. The aim of this study was to evaluate the use of four-dimensional (4D) flow MRI to quantify hemodynamic changes after SRS as early as 2 months. This was a retrospective observational study, which included 14 patients with both pre-SRS and post-SRS imaging obtained at multiple time points from 1 to 27 months after SRS. A 3T MRI Scanner was used to obtain T₂ single-shot fast spin echo, time-of-flight MRA, and postcontrast 4D flow with three-dimensional velocity encoding between 150 and 200 cm/s. Post-hoc two-dimensional cross-sectional flow was measured for the dominant feeding artery, the draining vein, and the corresponding contralateral artery as a control. Measurements were performed by two independent observers, and reproducibility was assessed. Wilcoxon signed-rank tests were used to compare differences in flow, circumference, and pulsatility between the feeding artery and the contralateral artery both before and after SRS; and differences in nidus size and flow and circumference of the feeding artery and draining vein before and after SRS. Arterial flow (L/min) decreased in the primary feeding artery (mean: 0.1 ± 0.07 vs. 0.3 ± 0.2; p < 0.05) and normalized in comparison to the contralateral artery (mean: 0.1 ± 0.07 vs. 0.1 ± 0.07; p = 0.068). Flow decreased in the draining vein (mean: 0.1 ± 0.2 vs. 0.2 ± 0.2; p < 0.05), and the circumference of the draining vein also decreased (mean: 16.1 ± 8.3 vs. 15.7 ± 6.7; p < 0.05). AVM volume decreased after SRS (mean: 45.3 ± 84.8 vs. 38.1 ± 78.7; p < 0.05). However, circumference (mm) of the primary feeding artery remained similar after SRS (mean: 15.7 ± 2.7 vs. 16.1 ± 3.1; p = 0.600). 4D flow may be able to demonstrate early hemodynamic changes in AVMs treated with radiosurgery, and these changes appear to be more pronounced and occur earlier than the structural changes on standard MRI/MRA. Level of Evidence: 4 Technical Efficacy Stage: 1.

Machine learning in neurosurgery: a global survey

Background

Recent technological advances have led to the development and implementation of machine learning (ML) in various disciplines, including neurosurgery. Our goal was to conduct a comprehensive survey of neurosurgeons to assess the acceptance of and attitudes toward ML in neurosurgical practice and to identify factors associated with its use.

Methods

The online survey consisted of nine or ten mandatory questions and was distributed in February and March 2019 through the European Association of Neurosurgical Societies (EANS) and the Congress of Neurosurgeons (CNS).

Results

Out of 7280 neurosurgeons who received the survey, we received 362 responses, with a response rate of 5%, mainly in Europe and North America. In total, 103 neurosurgeons (28.5%) reported using ML in their clinical practice, and 31.1% in research. Adoption rates of ML were relatively evenly distributed, with 25.6% for North America, 30.9% for Europe, 33.3% for Latin America and the Middle East, 44.4% for Asia and Pacific and 100% for Africa with only two responses. No predictors of clinical ML use were identified, although academic settings and subspecialties neuro-oncology, functional, trauma and epilepsy predicted use of ML in research. The most common applications were for predicting outcomes and complications, as well as interpretation of imaging.

Conclusions

This report provides a global overview of the neurosurgical applications of ML. A relevant proportion of the surveyed neurosurgeons reported clinical experience with ML algorithms. Future studies should aim to clarify the role and potential benefits of ML in neurosurgery and to reconcile these potential advantages with bioethical considerations.

Development of machine learning-based preoperative predictive analytics for unruptured intracranial aneurysm surgery: a pilot study

BACKGROUND

The decision to treat unruptured intracranial aneurysms (UIAs) or not is complex and requires balancing of risk factors and scores. Machine learning (ML) algorithms have previously been effective at generating highly accurate and comprehensive individualized preoperative predictive analytics in transsphenoidal pituitary and open tumor surgery. In this pilot study, we evaluate whether ML-based prediction of clinical endpoints is feasible for microsurgical management of UIAs.

METHODS

Based on data from a prospective registry, we developed and internally validated ML models to predict neurological outcome at discharge, as well as presence of new neurological deficits and any complication at discharge. Favorable neurological outcome was defined as modified Rankin scale (mRS) 0 to 2. According to the Clavien-Dindo grading (CDG), every adverse event during the post-operative course (surgery and not surgery related) is recorded as a complication. Input variables included age; gender; aneurysm complexity, diameter, location, number, and prior treatment; prior subarachnoid hemorrhage (SAH); presence of anticoagulation, antiplatelet therapy, and hypertension; microsurgical technique and approach; and various unruptured aneurysm scoring systems (PHASES, ELAPSS, UIATS).

RESULTS

We included 156 patients (26.3% male; mean [SD] age, 51.7 [11.0] years) with UIAs: 37 (24%) of them were treated for multiple aneurysm and 39 (25%) were treated for a complex aneurysm. Poor neurological outcome (mRS ≥ 3) was seen in 12 patients (7.7%) at discharge. New neurological deficits were seen in 10 (6.4%), and any kind of complication occurred in 20 (12.8%) patients. In the internal validation cohort, area under the curve (AUC) and accuracy values of 0.63-0.77 and 0.78-0.91 were observed, respectively.

CONCLUSIONS

Application of ML enables prediction of early clinical endpoints after microsurgery for UIAs. Our pilot study lays the groundwork for development of an externally validated multicenter clinical prediction model.

Prediction of 1-Year Mortality from Acute Myocardial Infarction Using Machine Learning

Risk stratification at hospital discharge could be instrumental in guiding postdischarge care. In this study, the risk models for 1-year mortality using machine learning (ML) were evaluated for guiding management of acute myocardial infarction (AMI) patients. From the Korea Acute Myocardial Infarction Registry (KAMIR) dataset, 22,182 AMI patients were selected. The 1-year all-cause mortality was recorded at 12-month follow-up periods. Anomaly detection was conducted for removing outliers; principal component analysis for dimensionality reduction, recursive feature elimination algorithm for feature selection. Model selection and training were conducted with 70% of the dataset after the creation and cross-validation of hundreds of models with decision trees, ensembles, logistic regressions, and deepnets algorithms. The rest of the dataset (30%) was used for comparison between the ML and KAMIR score-based models. The mean age of the AMI patients was 64 years, 71.8% were male, and 56.7% were eventually diagnosed with ST-elevation myocardial infarction. There were 1,332 patients suffering from all-cause mortality (6%) during a median 338 days of follow-up. The ML models for 1-year mortality were well-calibrated (Hosmer-Lemeshow p >0.05) and showed good discrimination (area under the curve for test cohort: 0.918). Compared with the performance of the KAMIR score model, the ML model had a higher area under the curve, net reclassification improvement, and integrated discrimination improvement. The ML model for 1-year mortality was well-calibrated and had excellent discriminatory ability and higher performance. In a comprehensive clinical evaluation process, this model could support risk stratification and management in postdischarge AMI patients.

Pituitary Tumors in the Computational Era, Exploring Novel Approaches to Diagnosis, and Outcome Prediction with Machine Learning

BACKGROUND

Machine learning has emerged as a viable asset in the setting of pituitary surgery. In the past decade, the number of machine learning models developed to aid in the diagnosis of pituitary lesions and predict intraoperative and postoperative complications following transsphenoidal surgery has increased exponentially. As computational processing power continues to increase, big data sets continue to expand, and learning algorithms continue to surpass gold standard predictive tools, machine learning will serve to become an important component in improving patient care and outcomes.

METHODS

Relevant studies were identified based on a literature search in PubMed and MEDLINE databases, as well as from other sources including reference lists of published articles.

RESULTS

Radiomics and artificial neural networks comprise the majority of machine learning-based applications in pituitary surgery. Radiomics serves to quantify specific imaging features, which can then be used to noninvasively identify tumor characteristics and make definitive diagnoses, circumventing presurgical biopsy altogether. Neural networks can be adapted to predict intraoperative changes in visual evoked potentials or cerebral spinal fluid leak. In addition, these algorithms may be combined with others to predict tumor aggressiveness, gross total resection, recurrence and remission, and even total cost burden.

CONCLUSIONS

The field of machine learning is broad, with radiomics and artificial neural networks comprising 2 commonly used supervised learning methods in pituitary surgery. Given the large heterogeneity of pituitary and sellar lesions, the promise of machine learning lies in its ability to identify relationships and patterns that are otherwise hidden from standard statistical methods. While machine learning has great potential as a clinical adjunct during the surgical preplanning process and in predicting complications and outcomes, challenges moving forward include standardization and validation of these paradigms.

Microsurgery Versus Stereotactic Radiosurgery for Brain Arteriovenous Malformations: A Matched Cohort Study

BACKGROUND

Microsurgery (MS) and stereotactic radiosurgery (SRS) remain the preferred interventions for the curative treatment of brain arteriovenous malformations (AVM), but their relative efficacy remains incompletely defined.

OBJECTIVE

To compare the outcomes of MS to SRS for AVMs through a retrospective, matched cohort study.

METHODS

We evaluated institutional databases of AVM patients who underwent MS and SRS. MS-treated patients were matched, in a 1:1 ratio based on patient and AVM characteristics, to SRS-treated patients. Statistical analyses were performed to compare outcomes data between the 2 cohorts. The primary outcome was defined as AVM obliteration without a new permanent neurological deficit.

RESULTS

The matched MS and SRS cohorts were each comprised of 59 patients. Both radiological (85 vs 11 mo; P < .001) and clinical (92 vs 12 mo; P < .001) follow-up were significantly longer for the SRS cohort. The primary outcome was achieved in 69% of each cohort. The MS cohort had a significantly higher obliteration rate (98% vs 72%; P = .001), but also had a significantly higher rate of new permanent deficit (31% vs 10%; P = .011). The posttreatment hemorrhage rate was significantly higher for the SRS cohort (10% for SRS vs 0% for MS; P = .027). In subgroup analyses of ruptured and unruptured AVMs, no significant differences between the primary outcomes were observed.

CONCLUSION

For patients with comparable AVMs, MS and SRS afford similar rates of deficit-free obliteration. Nidal obliteration is more frequently achieved with MS, but this intervention also incurs a greater risk of new permanent neurological deficit.

Stereotactic radiosurgery for arteriovenous malformations of the basal ganglia and thalamus: an international multicenter study

OBJECTIVE

Arteriovenous malformations (AVMs) of the basal ganglia (BG) and thalamus are associated with elevated risks of both hemorrhage if left untreated and neurological morbidity after resection. Therefore, stereotactic radiosurgery (SRS) has become a mainstay in the management of these lesions, although its safety and efficacy remain incompletely understood. The aim of this retrospective multicenter cohort study was to evaluate the outcomes of SRS for BG and thalamic AVMs and determine predictors of successful endpoints and adverse radiation effects.

METHODS

The authors retrospectively reviewed data on patients with BG or thalamic AVMs who had undergone SRS at eight institutions participating in the International Gamma Knife Research Foundation (IGKRF) from 1987 to 2014. Favorable outcome was defined as AVM obliteration, no post-SRS hemorrhage, and no permanently symptomatic radiation-induced changes (RICs). Multivariable models were developed to identify independent predictors of outcome.

RESULTS

The study cohort comprised 363 patients with BG or thalamic AVMs. The mean AVM volume and SRS margin dose were 3.8 cm3 and 20.7 Gy, respectively. The mean follow-up duration was 86.5 months. Favorable outcome was achieved in 58.5% of patients, including obliteration in 64.8%, with rates of post-SRS hemorrhage and permanent RIC in 11.3% and 5.6% of patients, respectively. Independent predictors of favorable outcome were no prior AVM embolization (p = 0.011), a higher margin dose (p = 0.008), and fewer isocenters (p = 0.044).

CONCLUSIONS

SRS is the preferred intervention for the majority of BG and thalamic AVMs. Patients with morphologically compact AVMs that have not been previously embolized are more likely to have a favorable outcome, which may be related to the use of a higher margin dose.

Predictive Modeling of Outcomes After Traumatic and Nontraumatic Spinal Cord Injury Using Machine Learning: Review of Current Progress and Future Directions

Machine learning represents a promising frontier in epidemiological research on spine surgery. It consists of a series of algorithms that determines relationships between data. Machine learning maintains numerous advantages over conventional regression techniques, such as a reduced requirement for a priori knowledge on predictors and better ability to manage large datasets. Current studies have made extensive strides in employing machine learning to a greater capacity in spinal cord injury (SCI). Analyses using machine learning algorithms have been done on both traumatic SCI and nontraumatic SCI, the latter of which typically represents degenerative spine disease resulting in spinal cord compression, such as degenerative cervical myelopathy. This article is a literature review of current studies published in traumatic and nontraumatic SCI that employ machine learning for the prediction of a host of outcomes. The studies described utilize machine learning in a variety of capacities, including imaging analysis and prediction in large epidemiological data sets. We discuss the performance of these machine learning-based clinical prognostic models relative to conventional statistical prediction models. Finally, we detail the future steps needed for machine learning to become a more common modality for statistical analysis in SCI.

Optimization of treatment strategy by using a machine learning model to predict survival time of patients with malignant glioma after radiotherapy

The purpose of this study was to predict the survival time of patients with malignant glioma after radiotherapy with high accuracy by considering additional clinical factors and optimize the prescription dose and treatment duration for individual patient by using a machine learning model. A total of 35 patients with malignant glioma were included in this study. The candidate features included 12 clinical features and 192 dose-volume histogram (DVH) features. The appropriate input features and parameters of the support vector machine (SVM) were selected using the genetic algorithm based on Akaike's information criterion, i.e. clinical, DVH, and both clinical and DVH features. The prediction accuracy of the SVM models was evaluated through a leave-one-out cross-validation test with residual error, which was defined as the absolute difference between the actual and predicted survival times after radiotherapy. Moreover, the influences of various values of prescription dose and treatment duration on the predicted survival time were evaluated. The prediction accuracy was significantly improved with the combined use of clinical and DVH features compared with the separate use of both features (P < 0.01, Wilcoxon signed rank test). Mean ± standard deviation of the leave-one-out cross-validation using the combined clinical and DVH features, only clinical features and only DVH features were 104.7 ± 96.5, 144.2 ± 126.1 and 204.5 ± 186.0 days, respectively. The prediction accuracy could be improved with the combination of clinical and DVH features, and our results show the potential to optimize the treatment strategy for individual patients based on a machine learning model.

A machine learning approach to predict early outcomes after pituitary adenoma surgery

OBJECTIVEPituitary adenomas occur in a heterogeneous patient population with diverse perioperative risk factors, endocrinopathies, and other tumor-related comorbidities. This heterogeneity makes predicting postoperative outcomes challenging when using traditional scoring systems. Modern machine learning algorithms can automatically identify the most predictive risk factors and learn complex risk-factor interactions using training data to build a robust predictive model that can generalize to new patient cohorts. The authors sought to build a predictive model using supervised machine learning to accurately predict early outcomes of pituitary adenoma surgery.METHODSA retrospective cohort of 400 consecutive pituitary adenoma patients was used. Patient variables/predictive features were limited to common patient characteristics to improve model implementation. Univariate and multivariate odds ratio analysis was performed to identify individual risk factors for common postoperative complications and to compare risk factors with model predictors. The study population was split into 300 training/validation patients and 100 testing patients to train and evaluate four machine learning models using binary classification accuracy for predicting early outcomes.RESULTSThe study included a total of 400 patients. The mean ± SD patient age was 53.9 ± 16.3 years, 59.8% of patients had nonfunctioning adenomas and 84.7% had macroadenomas, and the mean body mass index (BMI) was 32.6 ± 7.8 (58.0% obesity rate). Multivariate odds ratio analysis demonstrated that age < 40 years was associated with a 2.86 greater odds of postoperative diabetes insipidus and that nonobese patients (BMI < 30) were 2.2 times more likely to develop postoperative hyponatremia. Using broad criteria for a poor early postoperative outcome-major medical and early surgical complications, extended length of stay, emergency department admission, inpatient readmission, and death-31.0% of patients met criteria for a poor early outcome. After model training, a logistic regression model with elastic net (LR-EN) regularization best predicted early postoperative outcomes of pituitary adenoma surgery on the 100-patient testing set-sensitivity 68.0%, specificity 93.3%, overall accuracy 87.0%. The receiver operating characteristic and precision-recall curves for the LR-EN model had areas under the curve of 82.7 and 69.5, respectively. The most important predictive variables were lowest perioperative sodium, age, BMI, highest perioperative sodium, and Cushing's disease.CONCLUSIONSEarly postoperative outcomes of pituitary adenoma surgery can be predicted with 87% accuracy using a machine learning approach. These results provide insight into how predictive modeling using machine learning can be used to improve the perioperative management of pituitary adenoma patients.

Stereotactic Radiosurgery for Unruptured Versus Ruptured Pediatric Brain Arteriovenous Malformations

Background and Purpose—

The effects of prior hemorrhage on stereotactic radiosurgery (SRS) outcomes for pediatric arteriovenous malformations (AVMs) are not well defined. The aim of this multicenter, retrospective cohort study is to compare the SRS outcomes for unruptured versus ruptured pediatric AVMs.

Methods—

The International Radiosurgery Research Foundation pediatric AVM database from 1987 to 2018 was reviewed retrospectively. Favorable outcome was defined as AVM obliteration, no post-SRS hemorrhage, and no permanently symptomatic radiation-induced changes. Associations between prior hemorrhage and outcomes were adjusted for baseline differences, inverse probability weights, and competing risks.

Results—

The study cohort comprised 153 unruptured and 386 ruptured AVMs. Favorable outcome was achieved in 48.4% and 60.4% of unruptured and ruptured AVMs, respectively (adjusted odds ratio, 1.353; P =0.190). Cumulative AVM obliteration probabilities were 51.2%, 59.4%, 64.2%, and 70.0% for unruptured and 61.0%, 69.3%, 74.0%, and 79.3% for ruptured AVMs at 4, 6, 8, and 10 years, respectively (subhazard ratio, 1.311; P =0.020). Cumulative post-SRS hemorrhage probabilities were 4.5%, 5.6%, 5.6%, and 9.8% for unruptured and 4.7%, 6.1%, 6.1%, and 10.6% for ruptured AVMs at 4, 6, 8, and 10 years, respectively (subhazard ratio, 1.086; P =0.825). Probabilities of AVM obliteration (adjusted subhazard ratio, 0.968; P =0.850) and post-SRS hemorrhage (adjusted subhazard ratio, 1.663; P =0.251) were comparable between the 2 cohorts after inverse probability weight adjustments. Symptomatic (15.8% versus 8.1%; adjusted odds ratio, 0.400; P =0.008) and permanent (9.2% versus 5.0%; adjusted odds ratio, 0.441; P =0.045) radiation-induced change were more common in unruptured AVMs.

Conclusions—

The overall outcomes after SRS for unruptured versus ruptured pediatric AVMs are comparable. However, symptomatic and permanent radiation-induced change occur more frequently in pediatric patients with unruptured AVMs.

Can machine learning algorithms accurately predict discharge to nonhome facility and early unplanned readmissions following spinal fusion? Analysis of a national surgical registry

OBJECTIVE

Nonhome discharge and unplanned readmissions represent important cost drivers following spinal fusion. The authors sought to utilize different machine learning algorithms to predict discharge to rehabilitation and unplanned readmissions in patients receiving spinal fusion.

METHODS

The authors queried the 2012-2013 American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) for patients undergoing cervical or lumbar spinal fusion. Outcomes assessed included discharge to nonhome facility and unplanned readmissions within 30 days after surgery. A total of 7 machine learning algorithms were evaluated. Predictive hierarchical clustering of procedure codes was used to increase model performance. Model performance was evaluated using overall accuracy and area under the receiver operating characteristic curve (AUC), as well as sensitivity, specificity, and positive and negative predictive values. These performance metrics were computed for both the imputed and unimputed (missing values dropped) datasets.

RESULTS

A total of 59,145 spinal fusion cases were analyzed. The incidence rates of discharge to nonhome facility and 30-day unplanned readmission were 12.6% and 4.5%, respectively. All classification algorithms showed excellent discrimination (AUC > 0.80, range 0.85-0.87) for predicting nonhome discharge. The generalized linear model showed comparable performance to other machine learning algorithms. By comparison, all models showed poorer predictive performance for unplanned readmission, with AUC ranging between 0.63 and 0.66. Better predictive performance was noted with models using imputed data.

CONCLUSIONS

In an analysis of patients undergoing spinal fusion, multiple machine learning algorithms were found to reliably predict nonhome discharge with modest performance noted for unplanned readmissions. These results provide early evidence regarding the feasibility of modern machine learning classifiers in predicting these outcomes and serve as possible clinical decision support tools to facilitate shared decision making.