An Artificial Neural Network Model for the Prediction of Perioperative Blood Transfusion in Adult Spinal Deformity Surgery

Recent Outcomes and Challenges of Artificial Intelligence, Machine Learning, and Deep Learning in Neurosurgery

Neurosurgeons receive extensive technical training, which equips them with the knowledge and skills to specialise in various fields and manage the massive amounts of information and decision-making required throughout the various stages of neurosurgery, including preoperative, intraoperative, and postoperative care and recovery. Over the past few years, artificial intelligence (AI) has become more useful in neurosurgery. AI has the potential to improve patient outcomes by augmenting the capabilities of neurosurgeons and ultimately improving diagnostic and prognostic outcomes as well as decision-making during surgical procedures. By incorporating AI into both interventional and non-interventional therapies, neurosurgeons may provide the best care for their patients. AI, machine learning (ML), and deep learning (DL) have made significant progress in the field of neurosurgery. These cutting-edge methods have enhanced patient outcomes, reduced complications, and improved surgical planning.

The Utility of the Validated Intraoperative Bleeding Scale in Thoracolumbar Spine Surgery: A Single-Center Prospective Study

STUDY DESIGN

Prospective, single-center study.

OBJECTIVE

To evaluate the clinical relevance of the validated intraoperative bleeding severity scale (VIBe) in thoracolumbar spine surgery.

METHODS

Adult patients aged 18 through 88 undergoing elective decompression, instrumentation, and fusion of the thoracolumbar spine were prospectively enrolled after informed consent was provided and written consent was obtained. Validated intraoperative bleeding severity scores were recorded intraoperatively. Univariate analysis consisted of Student T-tests, Pearson's χ2 Tests, Fisher's Exact Tests, linear regression, and binary logistic regression. Multivariable regression was conducted to adjust for baseline characteristics and potential confounding variables.

RESULTS

A total of N = 121 patients were enrolled and included in the analysis. After adjusting for confounders, VIBe scores were correlated with an increased likelihood of intraoperative blood transfusion (β = 2.46, P = .012), postoperative blood transfusion (β = 2.36, P = .015), any transfusion (β = 2.49, P < .001), total transfusion volume (β = 180.8, P = .020), and estimated blood loss (EBL) (β = 409, P < .001). Validated intraoperative bleeding severity scores had no significant association with length of hospital stay, 30-day readmission, 30-day reoperation, 30-day emergency department visit, change in pre- to post-op hemoglobin and hematocrit, total drain output, or length of surgery.

CONCLUSION

The VIBe scale is associated with perioperative transfusion rates and EBL in patients undergoing thoracolumbar spine surgery. Overall, the VIBe scale has clinically relevant meaning in spine surgery, and shows potential utility in clinical research.

LEVEL OF EVIDENCE

Level II.

Prediction of Transfusion among In-patient Population using Temporal Pattern based Clinical Similarity Graphs

Intelligent prediction of risk of blood transfusion among hospitalized patients can identify at-risk patients and provide timely information to the hospital to plan and reserve resources to meet the demand of blood transfusion. While previously proposed solutions focus on sub-populations such as patients admitted to ICU after gastrointestinal bleeding or postpartum patients with hemorrhage, we design a predictive model applicable to complete in-patient population. Our model relies on patients' similarity graph based on temporal patterns among clinical history of the patients. These graphs are processed through graph convolutional neural network (GCNN) to estimate node or patient level risk of blood transfusion. Thus, our model not only learns from the patient's own clinical history but also from other patients with similar clinical history. The model is also capable of fusing diverse data elements from electronic health records (EHR) such as demographic information, billing codes, and recorded vital signs. Our model was validated on both internal and external sets and outperformed all comparative baseline models.

Advances and Evolving Challenges in Spinal Deformity Surgery

BACKGROUND

Surgical intervention is a critical tool to address adult spinal deformity (ASD). Given the evolution of spinal surgical techniques, we sought to characterize developments in ASD correction and barriers impacting clinical outcomes.

METHODS

We conducted a literature review utilizing PubMed, Embase, Web of Science, and Google Scholar to examine advances in ASD surgical correction and ongoing challenges from patient and clinician perspectives. ASD procedures were examined across pre-, intra-, and post-operative phases.

RESULTS

Several factors influence the effectiveness of ASD correction. Standardized radiographic parameters and three-dimensional modeling have been used to guide operative planning. Complex minimally invasive procedures, targeted corrections, and staged procedures can tailor surgical approaches while minimizing operative time. Further, improvements in osteotomy technique, intraoperative navigation, and enhanced hardware have increased patient safety. However, challenges remain. Variability in patient selection and deformity undercorrection have resulted in heterogenous clinical responses. Surgical complications, including blood loss, infection, hardware failure, proximal junction kyphosis/failure, and pseudarthroses, pose barriers. Although minimally invasive approaches are being utilized more often, clinical validation is needed.

CONCLUSIONS

The growing prevalence of ASD requires surgical solutions that can lead to sustained symptom resolution. Leveraging computational and imaging advances will be necessary as we seek to provide comprehensive treatment plans for patients.