A Review of Machine Learning Approaches for Brain Positron Emission Tomography Data Analysis.

Positron emission tomography (PET) imaging has moved forward the development of medical diagnostics and research across various domains, including cardiology, neurology, infection detection, and oncology. The integration of machine learning (ML) algorithms into PET data analysis has further enhanced their capabilities of including disease diagnosis and classification, image segmentation, and quantitative analysis. ML algorithms empower researchers and clinicians to extract valuable insights from complex big PET datasets, which enabling automated pattern recognition, predictive health outcome modeling, and more efficient data analysis. This review explains the basic knowledge of PET imaging, statistical methods for PET image analysis, and challenges of PET data analysis. We also discussed the improvement of analysis capabilities by combining PET data with machine learning algorithms and the application of this combination in various aspects of PET image research. This review also highlights current trends and future directions in PET imaging, emphasizing the driving and critical role of machine learning and big PET image data analytics in improving diagnostic accuracy and personalized medical approaches. Integration between PET imaging will shape the future of medical diagnosis and research.

Graph Autoencoders for Embedding Learning in Brain Networks and Major Depressive Disorder Identification.

Brain functional connectivity (FC) networks inferred from functional magnetic resonance imaging (fMRI) have shown altered or aberrant brain functional connectome in various neuropsychiatric disorders. Recent application of deep neural networks to connectome-based classification mostly relies on traditional convolutional neural networks (CNNs) using input FCs on a regular Euclidean grid to learn spatial maps of brain networks neglecting the topological information of the brain networks, leading to potentially sub-optimal performance in brain disorder identification. We propose a novel graph deep learning framework that leverages non-Euclidean information inherent in the graph structure for classifying brain networks in major depressive disorder (MDD). We introduce a novel graph autoencoder (GAE) architecture, built upon graph convolutional networks (GCNs), to embed the topological structure and node content of large fMRI networks into low-dimensional representations. For constructing the brain networks, we employ the Ledoit-Wolf (LDW) shrinkage method to efficiently estimate high-dimensional FC metrics from fMRI data. We explore both supervised and unsupervised techniques for graph embedding learning. The resulting embeddings serve as feature inputs for a deep fully-connected neural network (FCNN) to distinguish MDD from healthy controls (HCs). Evaluating our model on resting-state fMRI MDD dataset, we observe that the GAE-FCNN outperforms several state-of-the-art methods for brain connectome classification, achieving the highest accuracy when using LDW-FC edges as node features. The graph embeddings of fMRI FC networks also reveal significant group differences between MDD and HCs. Our framework demonstrates the feasibility of learning graph embeddings from brain networks, providing valuable discriminative information for diagnosing brain disorders.

Volumetric brain MRI signatures of heart failure with preserved ejection fraction in the setting of dementia.

Heart failure with preserved ejection fraction (HFpEF) is an important, emerging risk factor for dementia, but it is not clear whether HFpEF contributes to a specific pattern of neuroanatomical changes in dementia. A major challenge to studying this is the relative paucity of datasets of patients with dementia, with/without HFpEF, and relevant neuroimaging. We sought to demonstrate the feasibility of using modern data mining tools to create and analyze clinical imaging datasets and identify the neuroanatomical signature of HFpEF-associated dementia. We leveraged the bioinformatics tools at Vanderbilt University Medical Center to identify patients with a diagnosis of dementia with and without comorbid HFpEF using the electronic health record. We identified high resolution, clinically-acquired neuroimaging data on 30 dementia patients with HFpEF (age 76.9 ± 8.12 years, 61% female) as well as 301 age- and sex-matched patients with dementia but without HFpEF to serve as comparators (age 76.2 ± 8.52 years, 60% female). We used automated image processing pipelines to parcellate the brain into 132 structures and quantify their volume. We found six regions with significant atrophy associated with HFpEF: accumbens area, amygdala, posterior insula, anterior orbital gyrus, angular gyrus, and cerebellar white matter. There were no regions with atrophy inversely associated with HFpEF. Patients with dementia and HFpEF have a distinct neuroimaging signature compared to patients with dementia only. Five of the six regions identified in are in the temporo-parietal region of the brain. Future studies should investigate mechanisms of injury associated with cerebrovascular disease leading to subsequent brain atrophy.

Effects of open-label transdermal nicotine antidepressant augmentation on affective symptoms and executive function in late-life depression.

BACKGROUND: Late-life depression (LLD) is characterized by a poor response to antidepressant medications and diminished cognitive performance, particularly in executive functioning. There is currently no accepted pharmacotherapy for LLD that effectively treats both mood and cognitive symptoms. This study investigated whether transdermal nicotine augmentation of standard antidepressant medications benefitted mood and cognitive symptoms in LLD. METHODS: Nonsmoking participants aged 60 years or older with unremitted LLD on stable SSRI or SNRI medications (N = 29) received transdermal nicotine patches up to a 21 mg daily dose over 12 weeks. Clinical measures assessed depression severity, secondary affective symptoms, and cognitive performance. Nicotine metabolite concentrations were obtained from blood samples. RESULTS: Depression severity significantly decreased over the trial, with a 76 % response rate and 59 % remission rate. Change in depression severity was positively associated with nicotine exposure. Participants also exhibited improvement in self-reported affective symptoms (apathy, insomnia, rumination, and generalized anxiety symptoms), negativity bias, and disability. Executive function test performance significantly improved, specifically in measures of cognitive control, as did subjective cognitive performance. Adverse events were generally mild, with 75 % of the sample tolerating the maximum dose. CONCLUSION: The current study extends our previous pilot open-label trial in LLD, supporting feasibility and tolerability of transdermal nicotine patches as antidepressant augmentation. Although preliminary, this open-label study supports the potential benefit of transdermal nicotine patches for both mood and cognitive symptoms of LLD. Further research, including definitive randomized, blinded trials, is warranted to confirm these findings and explore long-term risk and benefit. TRIAL REGISTRATION: The study was registered with clinicaltrials.gov (NCT04433767).

Polyethylene Glycol-Mediated Axonal Fusion Promotes Early Sensory Recovery After Digital Nerve Injury: A Randomized Clinical Trial.

BACKGROUND: Peripheral nerve repair is limited by Wallerian degeneration coupled with the slow and inconsistent rates of nerve regrowth. In more proximal injuries, delayed nerve regeneration can cause debilitating muscle atrophy. Topical application of polyethylene glycol (PEG) during neurorrhaphy facilitates the fusion of severed axonal membranes, immediately restoring action potentials across the coaptation site. In preclinical animal models, PEG-fusion resulted in remarkable early functional recovery. METHODS: This is the first randomized clinical trial comparing functional outcomes between PEG-fusion and standard neurorrhaphy. Participants with digital nerve transections were followed up at 2 weeks, 1 month, and 3 months postoperatively. The primary outcome was assessed using the Medical Research Council Classification (MRCC) rating for sensory recovery at each timepoint. Semmes-Weinstein monofilaments and static two-point discrimination determined MRCC ratings. Postoperative quality of life was measured using the Michigan Hand Questionnaire (MHQ). RESULTS: Forty-eight transected digital nerves (25 control, 23 PEG) across twenty-two patients were analyzed. PEG-fused nerves demonstrated significantly higher MRCC scores at 2 weeks (OR 16.95, 95% CI: 1.79 - 160.38, p = 0.008) and 1 month (OR 13.40, 95% CI: 1.64 - 109.77, p = 0.009). Participants in the PEG cohort also had significantly higher average MHQ scores at 2 weeks (Hodge's g 1.28, 95% CI: 0.23 - 2.30, p = 0.0163) and 1 month (Hodge's g 1.02, 95% CI: 0.04 - 1.99, p = 0.049). No participants had adverse events related to the study drug. CONCLUSION: PEG-fusion promotes early sensory recovery and improved patient well-being following peripheral nerve repair of digital nerves.

Overcome the Limitation of Phenome-Wide Association Studies (PheWAS): Extension of PheWAS to Efficient and Robust Large-Scale ICD Codes Analysis.

The Phenome-wide association studies (PheWAS) have become widely used for efficient, high-throughput evaluation of relationship between a genetic factor and a large number of disease phenotypes, typically extracted from a DNA biobank linked with electronic medical records (EMR). Phecodes, billing code-derived disease case-control status, are usually used as outcome variables in PheWAS and logistic regression has been the standard choice of analysis method. Since the clinical diagnoses in EMR are often inaccurate with errors which can lead to biases in the odds ratio estimates, much effort has been put to accurately define the cases and controls to ensure an accurate analysis. Specifically in order to correctly classify controls in the population, an exclusion criteria list for each Phecode was manually compiled to obtain unbiased odds ratios. However, the accuracy of the list cannot be guaranteed without extensive data curation process. The costly curation process limits the efficiency of large-scale analyses that take full advantage of all structured phenotypic information available in EMR. Here, we proposed to estimate relative risks (RR) instead. We first demonstrated the desired nature of RR that overcomes the inaccuracy in the controls via theoretical formula. With simulation and real data application, we further confirmed that RR is unbiased without compiling exclusion criteria lists. With RR as estimates, we are able to efficiently extend PheWAS to a larger-scale, phenome construction agnostic analysis of phenotypes, using ICD 9/10 codes, which preserve much more disease-related clinical information than Phecodes.

Prophylactic Antibiotic Choice and Deep Infection in Lower Extremity Endoprosthetic Reconstruction: Comparison of Cefazolin, Cefazolin-Vancomycin, and Alternative Regimens.

INTRODUCTION: Infection is a common mode of failure in lower extremity endoprostheses. The Prophylactic Antibiotic Regimens in Tumor Surgery trial reported that 5 days of cefazolin had no difference in surgical site infection compared with 24 hours of cefazolin. Our purpose was to evaluate infection rates of patients receiving perioperative cefazolin monotherapy, cefazolin-vancomycin dual therapy, or alternative antibiotic regimens. METHODS: A single-center retrospective review was conducted on patients who received lower extremity endoprostheses from 2008 to 2021 with minimum 1-year follow-up. Three prophylactic antibiotic regimen groups were compared: cefazolin monotherapy, cefazolin-vancomycin dual therapy, and alternative regimens. The primary outcome was deep infection, defined by a sinus tract, positive culture, or clinical diagnosis. Secondary outcomes were revision surgery, microorganisms isolated, and superficial wound issues. RESULTS: The overall deep infection rate was 10% (30/294) at the median final follow-up of 3.0 years (IQR 1.7 to 5.4). The deep infection rates in the cefazolin, cefazolin-vancomycin, and alternative regimen groups were 8% (6/72), 10% (18/179), and 14% (6/43), respectively (P = 0.625). Patients not receiving cefazolin had an 18% deep infection rate (6/34) and 21% revision surgery rate (7/34) compared with a 9% deep infection rate (24/260) (P = 0.13) and 12% revision surgery rate (31/260) (P = 0.17) in patients receiving cefazolin. In those not receiving cefazolin, 88% (30/34) were due to a documented penicillin allergy, only two being anaphylaxis. All six patients in the alternative regimen group who developed deep infections did not receive cefazolin secondary to nonanaphylactic penicillin allergy. CONCLUSION: The addition of perioperative vancomycin to cefazolin in lower extremity endoprosthetic reconstructions was not associated with a lower deep infection rate. Patients who did not receive cefazolin trended toward higher rates of deep infection and revision surgery, although not statistically significant. The most common reason for not receiving cefazolin was a nonanaphylactic penicillin allergy, highlighting the continued practice of foregoing cefazolin unnecessarily.

Longitudinal Assessment of Pancreas Volume by MRI Predicts Progression to Stage 3 Type 1 Diabetes.

OBJECTIVE: This multicenter prospective cohort study compared pancreas volume as assessed by MRI, metabolic scores derived from oral glucose tolerance testing (OGTT), and a combination of pancreas volume and metabolic scores for predicting progression to stage 3 type 1 diabetes (T1D) in individuals with multiple diabetes-related autoantibodies. RESEARCH DESIGN AND METHODS: Pancreas MRI was performed in 65 multiple autoantibody-positive participants enrolled in the Type 1 Diabetes TrialNet Pathway to Prevention study. Prediction of progression to stage 3 T1D was assessed using pancreas volume index (PVI), OGTT-derived Index60 score and Diabetes Prevention Trial-Type 1 Risk Score (DPTRS), and a combination of PVI and DPTRS. RESULTS: PVI, Index60, and DPTRS were all significantly different at study entry in 11 individuals who subsequently experienced progression to stage 3 T1D compared with 54 participants who did not experience progression (P < 0.005). PVI did not correlate with metabolic testing across individual study participants. PVI declined longitudinally in the 11 individuals diagnosed with stage 3 T1D, whereas Index60 and DPTRS increased. The area under the receiver operating characteristic curve for predicting progression to stage 3 from measurements at study entry was 0.76 for PVI, 0.79 for Index60, 0.79 for DPTRS, and 0.91 for PVI plus DPTRS. CONCLUSIONS: These findings suggest that measures of pancreas volume and metabolism reflect distinct components of risk for developing stage 3 type 1 diabetes and that a combination of these measures may provide superior prediction than either alone.