Joeky Senders

163 Summary SUMMARY Despite improved surgical and adjuvant treatment options, survival in patients with a malignant brain tumor remains dismal. Over the past decades, the volume and complexity of clinically-derived patient data (i.e., imaging, genomics, free-text etc.) is increasing exponentially. Machine learning provides a vast range of algorithms that can learn from this data and guide clinical decision-making by providing accurate patient- level predictions. The current thesis describes several studies along the continuum of the machine learning spectrum as it applies to neurosurgical oncology. Part I: Outcomes and risk factors in neurosurgical oncology This part characterizes postoperative outcomes and associated risk factors in patients undergoing craniotomy for a malignant brain tumor. In Chapter 2 , we have analyzed a cohort of 7376 patients identified through the National Surgical Quality Improvement Program (NSQIP) registry. Among patients undergoing craniotomy for a primary malignant brain tumor, 12.9% experienced a major complication within 30 days after surgery, most of which occurred during the initial hospital stay. The most common postoperativemajor complications were reoperation (5.1%), venous thromboembolism (VTE, 3.5%), anddeath (2.6%). Themost common reasons for reoperationandunplanned readmission were intracranial hemorrhage (18.5% of all re-operated patients) and wound-related complications (11.9% of all re-admitted patients). The American Society of Anesthesiologists (ASA)-classification and preoperative functional status were most frequently identified as predictors, as well as the strongest predictors of postoperative morbidity and mortality. Due to the high incidence and substantial impact on postoperative morbidity, a subsequent in-depth analysis ( Chapter 3 ) was performed in the same cohort to characterize the rates, timing, and predictors of VTE and intracranial hemorrhage. This study demonstrated that the increased risk of VTE extends beyond the period of hospitalization, especially for pulmonary embolism, whereas intracranial hemorrhages occurred predominantly within the first days after surgery. Although age and body mass index (BMI) were identified as overall predictors of VTE, distinct risk profiles were observed in patients who developed a pulmonary embolism versus deep venous thrombosis, as well as patients who developed a VTE in-hospital versus post-discharge. Given the persistent risk of pulmonary embolism beyond hospitalization, as well as the accumulation of intracranial hemorrhage events in the first days after surgery, a single-institutional retrospective cohort study ( Chapter 4 ) was performed to