Does the Presence of Missing Data Affect the Performance of the SORG Machine-learning Algorithm for Patients With Spinal Metastasis? Development of an Internet Application Algorithm.

BACKGROUND: The Skeletal Oncology Research Group machine-learning algorithm (SORG-MLA) was developed to predict the survival of patients with spinal metastasis. The algorithm was successfully tested in five international institutions using 1101 patients from different continents. The incorporation of 18 prognostic factors strengthens its predictive ability but limits its clinical utility because some prognostic factors might not be clinically available when a clinician wishes to make a prediction. QUESTIONS/PURPOSES: We performed this study to (1) evaluate the SORG-MLA's performance with data and (2) develop an internet-based application to impute the missing data. METHODS: A total of 2768 patients were included in this study. The data of 617 patients who were treated surgically were intentionally erased, and the data of the other 2151 patients who were treated with radiotherapy and medical treatment were used to impute the artificially missing data. Compared with those who were treated nonsurgically, patients undergoing surgery were younger (median 59 years [IQR 51 to 67 years] versus median 62 years [IQR 53 to 71 years]) and had a higher proportion of patients with at least three spinal metastatic levels (77% [474 of 617] versus 72% [1547 of 2151]), more neurologic deficit (normal American Spinal Injury Association [E] 68% [301 of 443] versus 79% [1227 of 1561]), higher BMI (23 kg/m2 [IQR 20 to 25 kg/m2] versus 22 kg/m2 [IQR 20 to 25 kg/m2]), higher platelet count (240 × 103/µL [IQR 173 to 327 × 103/µL] versus 227 × 103/µL [IQR 165 to 302 × 103/µL], higher lymphocyte count (15 × 103/µL [IQR 9 to 21× 103/µL] versus 14 × 103/µL [IQR 8 to 21 × 103/µL]), lower serum creatinine level (0.7 mg/dL [IQR 0.6 to 0.9 mg/dL] versus 0.8 mg/dL [IQR 0.6 to 1.0 mg/dL]), less previous systemic therapy (19% [115 of 617] versus 24% [526 of 2151]), fewer Charlson comorbidities other than cancer (28% [170 of 617] versus 36% [770 of 2151]), and longer median survival. The two patient groups did not differ in other regards. These findings aligned with our institutional philosophy of selecting patients for surgical intervention based on their level of favorable prognostic factors such as BMI or lymphocyte counts and lower levels of unfavorable prognostic factors such as white blood cell counts or serum creatinine level, as well as the degree of spinal instability and severity of neurologic deficits. This approach aims to identify patients with better survival outcomes and prioritize their surgical intervention accordingly. Seven factors (serum albumin and alkaline phosphatase levels, international normalized ratio, lymphocyte and neutrophil counts, and the presence of visceral or brain metastases) were considered possible missing items based on five previous validation studies and clinical experience. Artificially missing data were imputed using the missForest imputation technique, which was previously applied and successfully tested to fit the SORG-MLA in validation studies. Discrimination, calibration, overall performance, and decision curve analysis were applied to evaluate the SORG-MLA's performance. The discrimination ability was measured with an area under the receiver operating characteristic curve. It ranges from 0.5 to 1.0, with 0.5 indicating the worst discrimination and 1.0 indicating perfect discrimination. An area under the curve of 0.7 is considered clinically acceptable discrimination. Calibration refers to the agreement between the predicted outcomes and actual outcomes. An ideal calibration model will yield predicted survival rates that are congruent with the observed survival rates. The Brier score measures the squared difference between the actual outcome and predicted probability, which captures calibration and discrimination ability simultaneously. A Brier score of 0 indicates perfect prediction, whereas a Brier score of 1 indicates the poorest prediction. A decision curve analysis was performed for the 6-week, 90-day, and 1-year prediction models to evaluate their net benefit across different threshold probabilities. Using the results from our analysis, we developed an internet-based application that facilitates real-time data imputation for clinical decision-making at the point of care. This tool allows healthcare professionals to efficiently and effectively address missing data, ensuring that patient care remains optimal at all times. RESULTS: Generally, the SORG-MLA demonstrated good discriminatory ability, with areas under the curve greater than 0.7 in most cases, and good overall performance, with up to 25% improvement in Brier scores in the presence of one to three missing items. The only exceptions were albumin level and lymphocyte count, because the SORG-MLA's performance was reduced when these two items were missing, indicating that the SORG-MLA might be unreliable without these values. The model tended to underestimate the patient survival rate. As the number of missing items increased, the model's discriminatory ability was progressively impaired, and a marked underestimation of patient survival rates was observed. Specifically, when three items were missing, the number of actual survivors was up to 1.3 times greater than the number of expected survivors, while only 10% discrepancy was observed when only one item was missing. When either two or three items were omitted, the decision curves exhibited substantial overlap, indicating a lack of consistent disparities in performance. This finding suggests that the SORG-MLA consistently generates accurate predictions, regardless of the two or three items that are omitted. We developed an internet application (https://sorg-spine-mets-missing-data-imputation.azurewebsites.net/) that allows the use of SORG-MLA with up to three missing items. CONCLUSION: The SORG-MLA generally performed well in the presence of one to three missing items, except for serum albumin level and lymphocyte count (which are essential for adequate predictions, even using our modified version of the SORG-MLA). We recommend that future studies should develop prediction models that allow for their use when there are missing data, or provide a means to impute those missing data, because some data are not available at the time a clinical decision must be made. CLINICAL RELEVANCE: The results suggested the algorithm could be helpful when a radiologic evaluation owing to a lengthy waiting period cannot be performed in time, especially in situations when an early operation could be beneficial. It could help orthopaedic surgeons to decide whether to intervene palliatively or extensively, even when the surgical indication is clear.

A Dynamic Interbody Cage Improves Bone Formation in Anterior Cervical Surgery: A Porcine Biomechanical Study.

BACKGROUND: Anterior cervical discectomy and fusion (ACDF) with a rigid interbody spacer is commonly used in the treatment of cervical degenerative disc disease. Although ACDF relieves clinical symptoms, it is associated with several complications such as pseudoarthrosis and adjacent segment degeneration. The concept of dynamic fusion has been proposed to enhance fusion and reduce implant subsidence rate and post-fusion stiffness; this pilot preclinical animal study was conducted to begin to compare rigid and dynamic fusion in ACDF. QUESTIONS/PURPOSES: Using a pig model, we asked, is there (1) decreased subsidence, (2) reduced axial stiffness in compression, and (3) improved likelihood of bone growth with a dynamic interbody cage compared with a rigid interbody cage in ACDF? METHODS: ACDF was performed at two levels, C3/4 and C5/6, in 10 pigs weighing 48 to 55 kg at the age of 14 to 18 months (the pigs were skeletally mature). One level was implanted with a conventional rigid interbody cage, and the other level was implanted with a dynamic interbody cage. The conventional rigid interbody cage was implanted in the upper level in the first five pigs and in the lower level in the next five pigs. Both types of interbody cages were implanted with artificial hydroxyapatite and tricalcium phosphate bone grafts. To assess subsidence, we took radiographs at 0, 7, and 14 weeks postoperatively. Subsidence less than 10% of the disc height was considered as no radiologic abnormality. The animals were euthanized at 14 weeks, and each operated-on motion segment was harvested. Five specimens from each group were biomechanically tested under axial compression loading to determine stiffness. The other five specimens from each group were used for microCT evaluation of bone ingrowth and ongrowth and histologic investigation of bone formation. Sample size was determined based on 80% power and an α of 0.05 to detect a between-group difference of successful bone formation of 15%. RESULTS: With the numbers available, there was no difference in subsidence between the two groups. Seven of 10 operated-on levels with rigid cages had subsidence on a follow-up radiograph at 14 weeks, and subsidence occurred in two of 10 operated-on levels with dynamic cages (Fisher exact test; p = 0.07). The stiffness of the unimplanted rigid interbody cages was higher than the unimplanted dynamic interbody cages. After harvesting, the median (range) stiffness of the motion segments fused with dynamic interbody cages (531 N/mm [372 to 802]) was less than that of motion segments fused with rigid interbody cages (1042 N/mm [905 to 1249]; p = 0.002). Via microCT, we observed bone trabecular formation in both groups. The median (range) proportions of specimens showing bone ongrowth (88% [85% to 92%]) and bone volume fraction (87% [72% to 100%]) were higher in the dynamic interbody cage group than bone ongrowth (79% [71% to 81%]; p < 0.001) and bone volume fraction (66% [51% to 78%]; p < 0.001) in the rigid interbody cage group. The percentage of the cage with bone ingrowth was higher in the dynamic interbody cage group (74% [64% to 90%]) than in the rigid interbody cage group (56% [32% to 63%]; p < 0.001), and the residual bone graft percentage was lower (6% [5% to 8%] versus 13% [10% to 20%]; p < 0.001). In the dynamic interbody cage group, more bone formation was qualitatively observed inside the cages than in the rigid interbody cage group, with a smaller area of fibrotic tissue under histologic investigation. CONCLUSION: The dynamic interbody cage provided satisfactory stabilization and percentage of bone ongrowth in this in vivo model of ACDF in pigs, with lower stiffness after bone ongrowth and no difference in subsidence. CLINICAL RELEVANCE: The dynamic interbody cage appears to be worthy of further investigation. An animal study with larger numbers, with longer observation time, with multilevel surgery, and perhaps in the lumbar spine should be considered.

Validity of the I­FEED classification in assessing postoperative gastrointestinal impairment in patients undergoing elective lumbar spinal surgery with general anesthesia: a prospective observational study.

BACKGROUND: The I-FEED classification, scored 0-8, was reported to accurately describe the clinical manifestations of gastrointestinal impairment after colorectal surgery. Therefore, it is interesting to determine whether the I-FEED scoring system is also applicable to patients undergoing lumbar spine surgery. METHODS: Adult patients undergoing elective lumbar spine surgery were enrolled, and the I-FEED score was measured for 4 days after surgery. The I-FEED scoring system incorporates five elements: intake (score: 0, 1, 3), feeling nauseated (score: 0, 1, 3), emesis (score: 0, 1, 3), results of physical exam (score: 0, 1, 3), and duration of symptoms (score: 0, 1, 2). Daily I-FEED scores were summed, and the highest overall score is used to categorize patients into one of three categories: normal (0-2 points), postoperative gastrointestinal intolerance (POGI; 3-5 points), and postoperative gastrointestinal dysfunction (POGD; 6 + points). The construct validity hypothesis testing determines whether the I-FEED category is consistent with objective clinical findings relevant to gastrointestinal impairment, namely, the longer length of hospital stay (LOS), higher inhospital medical cost, more postoperative gastrointestinal medical treatment, and more postoperative non-gastrointestinal complications. RESULTS: A total of 156 patients were enrolled, and 25.0% of patients were categorized as normal, 49.4% POGI, and 25.6% POGD. Patients with higher I-FEED scores agreed with the four validity hypotheses. Patients with POGD had a significantly longer length of hospital stay (1 day longer median stay; p = 0.049) and more inhospital medical costs (approximately 500 Taiwanese dollars; p = 0.037), and more patients with POGD required rectal laxatives (10.3% vs. 32.5% vs. 32.5%; p = 0.026). In addition, more patients with POGD had non-gastrointestinal complications (5.1% vs. 11.7% vs. 30.0%; p = 0.034). CONCLUSION: This study contributes preliminary validity evidence for the I-FEED score as a measure for postoperative gastrointestinal impairment after elective lumbar spine surgery.

Metastatic spinal cord compression as the first manifestation of malignancy: A retrospective study of surgical outcome from single institution.

BACKGROUND: Given the limited studies addressing the issue about the effect of different surgical modalities for metastatic spinal cord compression (MSCC) as the first malignancy manifestation, we conducted a retrospective case-control study to evaluate the surgical outcome of MSCC as the first malignancy manifestation. METHODS: A total of 128 patients who were suspected of having metastatic spinal cord compression and underwent surgery from 2008 to 2021 were enrolled in the study. All patients were categorized into either 'debulking group' or 'palliative group'. RESULTS: The primary outcome was progression-free survival (PFS). The secondary outcomes were overall survival (OS), Frankel scale, and Karnofsky scores. All the outcomes were analyzed with a data cutoff of December 31, 2021. There was a significant difference between groups in progression-free survival (PFS) (p = 0.0094). However, there was no significant difference between groups in the overall survival (OS) (p = 0.0746). Age of onset, gender, duration of symptoms, and location of spinal metastasis, initial Frankel, initial Tomita scores, and initial Karnofsky performance scale showed no significant differences between groups. CONCLUSION: In conclusion, debulking surgery was shown to provide better neurological recoveries and could be considered first in patients with metastatic spinal cord compression as the first malignancy manifestation.

Prognostic significance of lab data and performance comparison by validating survival prediction models for patients with spinal metastases after radiotherapy.

BACKGROUND AND PURPOSE: Well-performing survival prediction models (SPMs) help patients and healthcare professionals to choose treatment aligning with prognosis. This retrospective study aims to investigate the prognostic impacts of laboratory data and to compare the performances of Metastases location, Elderly, Tumor primary, Sex, Sickness/comorbidity, and Site of radiotherapy (METSSS) model, New England Spinal Metastasis Score (NESMS), and Skeletal Oncology Research Group machine learning algorithm (SORG-MLA) for spinal metastases (SM). MATERIALS AND METHODS: From 2010 to 2018, patients who received radiotherapy (RT) for SM at a tertiary center were enrolled and the data were retrospectively collected. Multivariate logistic and Cox-proportional-hazard regression analyses were used to assess the association between laboratory values and survival. The area under receiver-operating characteristics curve (AUROC), calibration analysis, Brier score, and decision curve analysis were used to evaluate the performance of SPMs. RESULTS: A total of 2786 patients were included for analysis. The 90-day and 1-year survival rates after RT were 70.4% and 35.7%, respectively. Higher albumin, hemoglobin, or lymphocyte count were associated with better survival, while higher alkaline phosphatase, white blood cell count, neutrophil count, neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, or international normalized ratio were associated with poor prognosis. SORG-MLA has the best discrimination (AUROC 90-day, 0.78; 1-year 0.76), best calibrations, and the lowest Brier score (90-day 0.16; 1-year 0.18). The decision curve of SORG-MLA is above the other two competing models with threshold probabilities from 0.1 to 0.8. CONCLUSION: Laboratory data are of prognostic significance in survival prediction after RT for SM. Machine learning-based model SORG-MLA outperforms statistical regression-based model METSSS model and NESMS in survival predictions.

Effect of Bainite to Ferrite Yield Strength Ratio on the Deformability of Mesostructures for Ferrite/Bainite Dual-Phase Steels.

The strength and plasticity balance of F/B dual-phase X80 pipeline steels strongly depends on deformation compatibility between the soft phase of ferrite and the hard phase of bainite; thus, the tensile strength of ferrite and bainite, as non-negligible factors affecting the deformation compatibility, should be considered first. In this purely theoretical paper, an abstract representative volume elements (RVE) model was developed, based on the mesostructure of an F/B dual-phase X80 pipeline steel. The effect of the yield strength difference between bainite and ferrite on tensile properties and the strain hardening behaviors of the mesostructure was studied. The results show that deformation first occurs in ferrite, and strain and stress localize in ferrite prior to bainite. In the modified Crussard-Jaoul (C-J) analysis, as the yield strength ratio of bainite to ferrite (σy,B/σy,F) increases, the transition strain associated with the deformation transformation from ferrite soft phase deformation to uniform deformation of ferrite and bainite increases. Meanwhile, as the uncoordinated deformation of ferrite and bainite is enhanced, the strain localization factor (SLF) increases, especially the local strain concentration. Consequently, the yield, tensile strength, and yield ratio (yield strength/tensile strength) increase with the increase in σy,B/σy,F. Inversely, the strain hardening exponent and uniform elongation decrease.

Randomized multi-reader evaluation of automated detection and segmentation of brain tumors in stereotactic radiosurgery with deep neural networks.

BACKGROUND: Stereotactic radiosurgery (SRS), a validated treatment for brain tumors, requires accurate tumor contouring. This manual segmentation process is time-consuming and prone to substantial inter-practitioner variability. Artificial intelligence (AI) with deep neural networks have increasingly been proposed for use in lesion detection and segmentation but have seldom been validated in a clinical setting. METHODS: We conducted a randomized, cross-modal, multi-reader, multispecialty, multi-case study to evaluate the impact of AI assistance on brain tumor SRS. A state-of-the-art auto-contouring algorithm built on multi-modality imaging and ensemble neural networks was integrated into the clinical workflow. Nine medical professionals contoured the same case series in two reader modes (assisted or unassisted) with a memory washout period of 6 weeks between each section. The case series consisted of 10 algorithm-unseen cases, including five cases of brain metastases, three of meningiomas, and two of acoustic neuromas. Among the nine readers, three experienced experts determined the ground truths of tumor contours. RESULTS: With the AI assistance, the inter-reader agreement significantly increased (Dice similarity coefficient [DSC] from 0.86 to 0.90, P < 0.001). Algorithm-assisted physicians demonstrated a higher sensitivity for lesion detection than unassisted physicians (91.3% vs 82.6%, P = .030). AI assistance improved contouring accuracy, with an average increase in DSC of 0.028, especially for physicians with less SRS experience (average DSC from 0.847 to 0.865, P = .002). In addition, AI assistance improved efficiency with a median of 30.8% time-saving. Less-experienced clinicians gained prominent improvement on contouring accuracy but less benefit in reduction of working hours. By contrast, SRS specialists had a relatively minor advantage in DSC, but greater time-saving with the aid of AI. CONCLUSIONS: Deep learning neural networks can be optimally utilized to improve accuracy and efficiency for the clinical workflow in brain tumor SRS.

Decompressive Cranioplasty (Osteoplastic Hinged Craniectomy): A Novel Technique for Increased Intracranial Pressure-Initial Experience and Outcome.

BACKGROUND: We redesigned decompressive craniectomy and cranioplasty procedures to decrease the inherent risk of complications. This novel technique, called decompressive cranioplasty, not only may decrease the complication rate but also may improve the cosmetic result, obviate the need for artificial skull implant, and increase the decompressive volume compared with traditional craniectomy. METHODS: In decompressive cranioplasty, the Agnes Fast craniotomy was adopted without cutting the temporalis muscle from the underlying bone flap. After opening the dura with or without removal of intracranial hematomas, duraplasty was performed with an intracranial pressure monitor inserted. Four miniplates were bent into a "Z" shape, and the vascularized bone flap was elevated approximately 1.2-1.5 cm above the outer cortex of the skull and fixed with the miniplates. Subsequent cranioplasty was done with a mini-incision on the miniplate sites and reshaping of the miniplate to align the outer cortex of the bone flap. RESULTS: We successfully performed decompressive cranioplasty in 3 emergent cases-2 traumatic subdural hematomas and 1 malignant middle cerebral artery infarction. Postoperative brain computed tomography demonstrated adequate decompression in all cases. Cosmetic outcome was excellent, and there was no temporal hallowing. Mastication function was not affected. At 6-month follow-up, there was no bone flap shrinkage and no hydrocephalus. CONCLUSIONS: Decompressive cranioplasty is a safe and effective method in the management of patients with brain edema and intracranial hypertension. It is simple to perform and may reduce the morbidity associated with traditional decompressive craniectomy and subsequent cranioplasty.

Acute Infarction in the Artery of Percheron Distribution during Cerebral Angiography: A Case Report and Literature Review.

Improvements in techniques, contrast agents, and catheter design have significantly decreased angiography-related neurological deficits and complications. This article reports a case involving an angiographic total obliteration arteriovenous malformation (AVM) in a patient with an acute infarction in the artery of Percheron (AOP) distribution following angiography. Furthermore, imaging of an AOP acute infarction in cerebral angiography is presented.

Symptom distress and quality of life after stereotactic radiosurgery in patients with pituitary tumors: a questionnaire survey.

BACKGROUND: Stereotactic radiosurgery (SRS) is a common treatment for recurrent or residual pituitary adenomas. The persistence of symptoms and treatment related complications may impair the patient's quality of life (QOL). PURPOSE: The purpose of this study was to examine symptom distress, QOL, and the relationship between them among patients with pituitary tumors who had undergone SRS. METHODS: This study used a cross-sectional design and purposive sampling. We enrolled patients diagnosed with pituitary tumors who had undergone SRS. Data were collected at the CyberKnife Center at a medical center in Northern Taiwan in 2012. A questionnaire survey was used for data collection. Our questionnaire consisted of 3 parts the Pituitary Tumor Symptom Distress Questionnaire, the World Health Organization Quality of Life Instrument Short-Form (WHOQOL-BREF), and a demographic questionnaire. RESULTS: Sixty patients were enrolled in the study. The most common symptoms reported by patients after SRS were memory loss, fatigue, blurred vision, headache, sleep problems, and altered libido. The highest and lowest scores for QOL were in the environmental and psychological domains, respectively. Age was positively correlated with general health and the psychological domains. Level of symptom distress was negatively correlated with overall QOL, general health, physical health, and the psychological and social relationships domains. The scores in the psychological and environmental domains were higher in males than in females. Patients with ≤6 symptoms had better overall QOL, general health, physical health, and psychological and social relationships than those with >6 symptoms. CONCLUSION: Symptom distress can affect different aspects of patient QOL. Levels of symptom distress, number of symptoms, age, and gender were variables significantly correlated with patient QOL. These results may be utilized by healthcare personnel to design educational and targeted interventional programs for symptom management to improve patient QOL.