To develop and validate a deep learning algorithm that predicts the final diagnosis of Alzheimer disease (AD), mild cognitive impairment, or neither at fluorine 18 (18 F) fluorodeoxyglucose (FDG) PET of the brain and compare its performance to that of radiologic readers. Materials and Methods: Prospective 18 F-FDG PET brain images from the Alzheimer's Disease Neuroimaging Initiative (ADNI) (2109 imaging studies from 2005 to 2017, 1002 patients) and retrospective independent test set (40 imaging studies from 2006 to 2016, 40 patients) were collected. Final clinical diagnosis at follow-up was recorded. Convolutional neural network of InceptionV3 architecture was trained on 90% of ADNI data set and tested on the remaining 10%, as well as the independent test set, with performance compared to radiologic readers. Model was analyzed with sensitivity, specificity, receiver operating characteristic (ROC), saliency map, and t-distributed stochastic neighbor embedding. Results: The algorithm achieved area under the ROC curve of 0.98 (95% confidence interval: 0.94, 1.00) when evaluated on predicting the final clinical diagnosis of AD in the independent test set (82% specificity at 100% sensitivity), an average of 75.8 months prior to the final diagnosis, which in ROC space outperformed reader performance (57% [four of seven] sensitivity, 91% [30 of 33] specificity; P , .05). Saliency map demonstrated attention to known areas of interest but with focus on the entire brain. Conclusion: By using fluorine 18 fluorodeoxyglucose PET of the brain, a deep learning algorithm developed for early prediction of Alzheimer disease achieved 82% specificity at 100% sensitivity, an average of 75.8 months prior to the final diagnosis.

Study Design. Retrospective cohort. Objective. We evaluate the correlation of the Patient-Reported Outcomes Measurement Information System for physical function (PROMIS-PF) with legacy patient-reported outcome measures (PROMs) in patients undergoing minimally invasive transforaminal lumbar interbody fusion (MIS TLIF) up to 2 years postoperatively. Summary of Background Data. PROMIS-PF has not been validated past 6 months following MIS TLIF. Methods. A surgical registry was retrospectively reviewed for eligible MIS TLIFs between May 2015 and September 2017. Inclusion criteria were primary, one- or two-level MIS TLIFs for degenerative spinal pathology. Patients without preoperative or 2-year follow up PROMIS-PF surveys were excluded. Demographic, perioperative, and PROMs including Visual Analog Scale (VAS) back, VAS leg, Oswestry Disability Index (ODI), 12-Item Short Form (SF-12) physical component summary (PCS) scores, and PROMIS-PF at preoperative and postoperative timepoint (e.g., 6 weeks, 12 weeks, 6 months, 1 year, and 2 years). A paired t test evaluated PROM improvement from baseline. The relationship of PROMIS-PF with VAS back, VAS leg, SF-12 PCS, and ODI was evaluated with a Pearson correlation coefficient. Results. The 68-subject cohort was 41.2% female, with an average age of 52.9 years; 44.1% were obese, and the majority underwent one-level fusions (95.6%). Pain (VAS back, VAS leg) and disability metrics (ODI) demonstrated significant improvement at all timepoints following MIS TLIF when compared to baseline (all P < 0.001). Physical function (SF-12 PCS, PROMIS-PF) demonstrated significant postoperative improvement at 12 weeks, 6 months, 1 year, and 2 years (all P < 0.001). All evaluated timepoints, with the exception of preoperative VAS back scores, revealed strong PROMIS-PF correlations with VAS back, VAS leg, ODI, and SF-12 PCS. Conclusion. PROMIS-PF demonstrated a strong correlation with pain (VAS back, VAS leg), disability (ODI) and physical function (SF-12) at all postoperative follow-ups through 2 years. Our study provides longitudinal evidence for utilizing PROMIS-PF as a valid physical function measure among patients undergoing MIS TLIF. Level of Evidence: 4.

Study Design. Literature review. Objective. To discern which social media outlets contribute most to spine surgery literature dissemination and to investigate how popular articles compare to articles with most citations. Summary of Background Data. Scientific literature is increasingly disseminated through social media. The Altmetric Attention Score (AAS) is defined as an automated, weighted score calculation from counts of all online attention received by various research outputs. Increasing AAS values indicate more online attention. For example, the overall top 100 Altmetric spine surgery peer-reviewed articles since 2010 had an AAS range from 78 to 1537. Among all spine surgery literature reviewed since 2010, the mean AAS was 5.3 with a median of 1.0. Methods. We performed an Altmetric database search of nine spine surgery journals from January 2010 to October 2019. Mean AAS was summarized alongside metrics including citation count and impact factor. We assessed correlations between AAS and online sources, readers, and citations. Journals were grouped by impact factor, and analysis-of-variance compared mean AAS. The 100 highest AAS articles were compared to the most cited. Results. 13,601 articles were included. The mean AAS was 5.3, with Twitter contributing the most. The three highest associations were news (P < 0.001), Twitter (P < 0.001), and Facebook (P < 0.001). There was no significant association between impact factor and AAS. Compared with the most cited articles, the top 100 AAS articles had significantly more article types, more prospective studies, fewer retrospective studies, fewer reviews, and fewer systematic reviews (P < 0.001 for all). Spine contributed the most articles in both top 100 sets. Conclusion. Our evaluation revealed Twitter, newsfeeds, and Facebook were the most significant social media outlets. Compared to articles with the most citations, the most popular articles are prospective and encompass broader study designs. Social media plays an integral role in dissemination, both within spine literature and the public sphere. Level of Evidence: 3