Artificial intelligence in neuropathology: deep learning-based assessment of tauopathy.

Accumulation of abnormal tau in neurofibrillary tangles (NFT) occurs in Alzheimer disease (AD) and a spectrum of tauopathies. These tauopathies have diverse and overlapping morphological phenotypes that obscure classification and quantitative assessments. Recently, powerful machine learning-based approaches have emerged, allowing the recognition and quantification of pathological changes from digital images. Here, we applied deep learning to the neuropathological assessment of NFT in postmortem human brain tissue to develop a classifier capable of recognizing and quantifying tau burden. The histopathological material was derived from 22 autopsy brains from patients with tauopathies. We used a custom web-based informatics platform integrated with an in-house information management system to manage whole slide images (WSI) and human expert annotations as ground truth. We utilized fully annotated regions to train a deep learning fully convolutional neural network (FCN) implemented in PyTorch against the human expert annotations. We found that the deep learning framework is capable of identifying and quantifying NFT with a range of staining intensities and diverse morphologies. With our FCN model, we achieved high precision and recall in naive WSI semantic segmentation, correctly identifying tangle objects using a SegNet model trained for 200 epochs. Our FCN is efficient and well suited for the practical application of WSIs with average processing times of 45 min per WSI per GPU, enabling reliable and reproducible large-scale detection of tangles. We measured performance on test data of 50 pre-annotated regions on eight naive WSI across various tauopathies, resulting in the recall, precision, and an F1 score of 0.92, 0.72, and 0.81, respectively. Machine learning is a useful tool for complex pathological assessment of AD and other tauopathies. Using deep learning classifiers, we have the potential to integrate cell- and region-specific annotations with clinical, genetic, and molecular data, providing unbiased data for clinicopathological correlations that will enhance our knowledge of the neurodegeneration.

Surgical technique and short-term outcome for experimental laparoscopic closure of the epiploic foramen in 6 horses.

OBJECTIVE: To describe a laparoscopic technique for, and short-term outcome after, closure of the epiploic foramen (EF) in horses. STUDY DESIGN: Descriptive, experimental study. ANIMALS: Healthy, adult horses (n = 6). METHODS: Laparoscopic portals to approach the EF were identified in standing horses. Under laparoscopic observation, the gastropancreatic fold and right lobe of the pancreas were grasped with Babcock forceps and secured to the caudate hepatic lobe using helical titanium coils to obliterate the EF. Surgical procedure time and intra- and postoperative complications were recorded. Serial analysis of select serum enzymes was used as an indication of involvement of the pancreas and liver. Closure was reevaluated at 4 weeks using repeat laparoscopy, and necropsy was performed immediately after. RESULTS: At initial surgery, EF closure was successful in all 6 horses; median surgical time was 40.5 minutes (range, 22-110 minutes). Serum gamma-glutamyl transferase (GGT) and sorbitol dehydrogenase (SDH) were not significantly altered by the surgical procedure; however, aspartate aminotransferase (AST) and amylase (AMY) were transiently increased. At repeat laparoscopic reevaluation, closure was complete in 5 horses, with partial closure of the EF observed in 1 horse. No complications related to the procedure were noted during or after surgery in any horse. CONCLUSIONS: EF closure in the standing horse can be accomplished without complications to the surrounding organs and vessels.

Telehealth Adoption by Mental Health and Substance Use Disorder Treatment Facilities in the COVID-19 Pandemic.

OBJECTIVE: The study examined temporal and geographic trends in telehealth availability at U.S. behavioral health treatment facilities and risk factors for not offering telehealth. METHODS: Longitudinal data on 15,691 outpatient behavioral health treatment facilities were extracted daily from the Substance Abuse and Mental Health Services Administration's Behavioral Health Treatment Services Locator between January 20, 2020, and January 20, 2021. Facilities operated by the Department of Veterans Affairs were excluded. Bivariate analyses were used to assess trends in telehealth availability in 2020 and 2021. Multivariable regression analysis was used to examine facility- and county-level characteristics associated with telehealth availability in 2021. RESULTS: Telehealth availability increased by 77% from 2020 to 2021 for mental health treatment facilities and by 143% for substance use disorder treatment facilities. By January 2021, 68% of outpatient mental health facilities and 57% of substance use disorder treatment facilities in the sample were offering telehealth. Mental health and substance use disorder treatment facilities that did not accept Medicaid as a form of payment were less likely to offer telehealth in 2021, compared with facilities that accepted Medicaid. Mental health and substance use disorder treatment facilities that accepted private insurance were more likely to offer telehealth in 2021, compared with facilities that did not accept private insurance. CONCLUSIONS: Although 2020 saw a dramatic increase in telehealth availability at behavioral health treatment facilities, 32% of mental health treatment facilities and 43% of substance use disorder treatment facilities did not offer telehealth in January 2021, nearly 1 year into the pandemic.

Patterns in Geographic Distribution of Substance Use Disorder Treatment Facilities in the US and Accepted Forms of Payment From 2010 to 2021.

Importance: The drug overdose crisis is a continuing public health problem and is expected to grow substantially in older adults. Understanding the geographic accessibility to a substance use disorder (SUD) treatment facility that accepts Medicare can inform efforts to address this crisis in older adults. Objective: To assess whether geographic accessibility of services was limited for older adults despite the increasing need for SUD and opioid use disorder treatments in this population. Design, Setting, and Participants: This longitudinal cross-sectional study obtained data on all licensed SUD treatment facilities for all US counties and Census tracts listed in the National Directory of Drug and Alcohol Abuse Treatment Programs from 2010 to 2021. Main Outcomes and Measures: Measures included the national proportion of treatment facilities accepting Medicare, Medicaid, private insurance, or cash as a form of payment; the proportion of counties with a treatment facility accepting each form of payment; and the proportion of the national population with Medicare, Medicaid, private insurance, or cash payment residing within a 15-, 30-, or 60-minute driving time from an SUD treatment facility accepting their form of payment in 2021. Results: A total of 11 709 SUD treatment facilities operated across the US per year between 2010 and 2021 (140 507 facility-year observations). Cash was the most commonly accepted form of payment (increasing slightly from 91.0% in 2010 to 91.6% by 2021), followed by private insurance (increasing from 63.5% to 75.3%), Medicaid (increasing from 54.0% to 71.8%), and Medicare (increasing from 32.1% to 41.9%). The proportion of counties with a treatment facility that accepted Medicare as a form of payment also increased over the same study period from 41.2% to 53.8%, whereas the proportion of counties with a facility that accepted Medicaid as a form of payment increased from 53.5% to 67.1%. The proportion of Medicare beneficiaries with a treatment facility that accepted Medicare as a form of payment within a 15-minute driving time increased from 53.3% to 57.0%. The proportion of individuals with a treatment facility within a 15-minute driving time that accepted their respective form of payment was 73.2% for those with Medicaid, 69.8% for those with private insurance, and 71.4% for those with cash payment in 2021. Conclusions and Relevance: Results of this study suggest that Medicare beneficiaries have less geographic accessibility to SUD treatment facilities given that acceptance of Medicare is low compared with other forms of payment. Policy makers need to consider increasing reimbursement rates and using additional incentives to encourage the acceptance of Medicare.

SARS-CoV-2 detection with aptamer-functionalized gold nanoparticles.

A rapid detection test for SARS-CoV-2 is urgently required to monitor virus spread and containment. Here, we describe a test that uses nanoprobes, which are gold nanoparticles functionalized with an aptamer specific to the spike membrane protein of SARS-CoV-2. An enzyme-linked immunosorbent assay confirms aptamer binding with the spike protein on gold surfaces. Protein recognition occurs by adding a coagulant, where nanoprobes with no bound protein agglomerate while those with sufficient bound protein do not. Using plasmon absorbance spectra, the nanoprobes detect 16 nM and higher concentrations of spike protein in phosphate-buffered saline. The time-varying light absorbance is examined at 540 nm to determine the critical coagulant concentration required to agglomerates the nanoprobes, which depends on the protein concentration. This approach detects 3540 genome copies/µl of inactivated SARS-CoV-2.

Availability and Accessibility of Mental Health Services in New York City.

Mental health services are critical components of public health infrastructure that provide essential supports to people living with psychiatric disorders. In a typical year, about 20 percent of people will have a psychiatric disorder, and about 5 percent will experience serious psychological distress, indicating a potentially serious mental illness. Nationally, the use of mental health services is low, and the use of care is not equitably distributed. In the United States as a whole and in New York City (NYC), non-Hispanic white individuals are more likely to use mental health services than non-Hispanic black individuals or Hispanic individuals. The challenges of ensuring the availability of mental health services for all groups in NYC are particularly acute, given the size of the population and its diversity in income, culture, ethnicity, and language. Adding to these underlying challenges, the coronavirus disease 2019 (COVID-19) pandemic has disrupted established patterns of care. To advance policy strategy for addressing gaps in the mental health services system, RAND researchers investigate the availability and accessibility of mental health services in NYC. The RAND team used two complementary approaches to address these issues. First, the team conducted interviews with a broad group of professionals and patients in the mental health system to identify barriers to care and potential strategies for improving access and availability. Second, the team investigated geographic variations in the availability of mental health services by compiling and mapping data on the locations and service characteristics of mental health treatment facilities in NYC.

Artificial intelligence-derived neurofibrillary tangle burden is associated with antemortem cognitive impairment.

Tauopathies are a category of neurodegenerative diseases characterized by the presence of abnormal tau protein-containing neurofibrillary tangles (NFTs). NFTs are universally observed in aging, occurring with or without the concomitant accumulation of amyloid-beta peptide (Aß) in plaques that typifies Alzheimer disease (AD), the most common tauopathy. Primary age-related tauopathy (PART) is an Aß-independent process that affects the medial temporal lobe in both cognitively normal and impaired subjects. Determinants of symptomology in subjects with PART are poorly understood and require clinicopathologic correlation; however, classical approaches to staging tau pathology have limited quantitative reproducibility. As such, there is a critical need for unbiased methods to quantitatively analyze tau pathology on the histological level. Artificial intelligence (AI)-based convolutional neural networks (CNNs) generate highly accurate and precise computer vision assessments of digitized pathology slides, yielding novel histology metrics at scale. Here, we performed a retrospective autopsy study of a large cohort (n = 706) of human post-mortem brain tissues from normal and cognitively impaired elderly individuals with mild or no Aß plaques (average age of death of 83.1 yr, range 55-110). We utilized a CNN trained to segment NFTs on hippocampus sections immunohistochemically stained with antisera recognizing abnormal hyperphosphorylated tau (p-tau), which yielded metrics of regional NFT counts, NFT positive pixel density, as well as a novel graph-theory based metric measuring the spatial distribution of NFTs. We found that several AI-derived NFT metrics significantly predicted the presence of cognitive impairment in both the hippocampus proper and entorhinal cortex (p < 0.0001). When controlling for age, AI-derived NFT counts still significantly predicted the presence of cognitive impairment (p = 0.04 in the entorhinal cortex; p = 0.04 overall). In contrast, Braak stage did not predict cognitive impairment in either age-adjusted or unadjusted models. These findings support the hypothesis that NFT burden correlates with cognitive impairment in PART. Furthermore, our analysis strongly suggests that AI-derived metrics of tau pathology provide a powerful tool that can deepen our understanding of the role of neurofibrillary degeneration in cognitive impairment.

Low-level blast exposure disrupts gliovascular and neurovascular connections and induces a chronic vascular pathology in rat brain.

Much concern exists over the role of blast-induced traumatic brain injury (TBI) in the chronic cognitive and mental health problems that develop in veterans and active duty military personnel. The brain vasculature is particularly sensitive to blast injury. The aim of this study was to characterize the evolving molecular and histologic alterations in the neurovascular unit induced by three repetitive low-energy blast exposures (3 × 74.5 kPa) in a rat model mimicking human mild TBI or subclinical blast exposure. High-resolution two-dimensional differential gel electrophoresis (2D-DIGE) and matrix-assisted laser desorption/ionization (MALDI) mass spectrometry of purified brain vascular fractions from blast-exposed animals 6 weeks post-exposure showed decreased levels of vascular-associated glial fibrillary acidic protein (GFAP) and several neuronal intermediate filament proteins (α-internexin and the low, middle, and high molecular weight neurofilament subunits). Loss of these proteins suggested that blast exposure disrupts gliovascular and neurovascular interactions. Electron microscopy confirmed blast-induced effects on perivascular astrocytes including swelling and degeneration of astrocytic endfeet in the brain cortical vasculature. Because the astrocyte is a major sensor of neuronal activity and regulator of cerebral blood flow, structural disruption of gliovascular integrity within the neurovascular unit should impair cerebral autoregulation. Disrupted neurovascular connections to pial and parenchymal blood vessels might also affect brain circulation. Blast exposures also induced structural and functional alterations in the arterial smooth muscle layer. Interestingly, by 8 months after blast exposure, GFAP and neuronal intermediate filament expression had recovered to control levels in isolated brain vascular fractions. However, despite this recovery, a widespread vascular pathology was still apparent at 10 months after blast exposure histologically and on micro-computed tomography scanning. Thus, low-level blast exposure disrupts gliovascular and neurovascular connections while inducing a chronic vascular pathology.

In silico prediction of tumor antigens derived from functional missense mutations of the cancer gene census.

Antigen-specific immune responses against peptides derived from missense gene mutations have been identified in multiple cancers. The application of personalized peptide vaccines based on the tumor mutation repertoire of each cancer patient is a near-term clinical reality. These peptides can be identified for pre-validation by leveraging the results of massive gene sequencing efforts in cancer. In this study, we utilized NetMHC 3.2 to predict nanomolar peptide binding affinity to 57 human HLA-A and B alleles. All peptides were derived from 5,685 missense mutations in 312 genes annotated as functionally relevant in the Cancer Genome Project. Of the 26,672,189 potential 8-11 mer peptide-HLA pairs evaluated, 0.4% (127,800) display binding affinities < 50 nM, predicting high affinity interactions. These peptides can be segregated into two groups based on the binding affinity to HLA proteins relative to germline-encoded sequences: peptides for which both the mutant and wild-type forms are high affinity binders, and peptides for which only the mutant form is a high affinity binder. Current evidence directs the attention to mutations that increase HLA binding affinity, as compared with cognate wild-type peptide sequences, as these potentially are more relevant for vaccine development from a clinical perspective. Our analysis generated a database including all predicted HLA binding peptides and the corresponding change in binding affinity as a result of point mutations. Our study constitutes a broad foundation for the development of personalized peptide vaccines that hone-in on functionally relevant targets in multiple cancers in individuals with diverse HLA haplotypes.

Evaluation of gastric pressures as an indirect method for measurement of intraabdominal pressures in the horse.

OBJECTIVE: To develop an indirect method for measurement of intraabdominal pressures in the standing horse using measurement of gastric pressures as a less invasive technique, and to compare this method with direct intraabdominal pressures obtained from the peritoneal cavity. DESIGN: Prospective, experimental study. SETTING: University-based equine research facility. ANIMALS: Ten healthy adult horses, 7 geldings and 3 mares. INTERVENTIONS: Gastric pressures were measured using a nasogastric tube with a U-tube manometry technique, while intraperitoneal pressures were measured with a peritoneal cannula. Measurements of intraabdominal pressure were obtained by both methods, simultaneously, and were evaluated using 5 increasing volumes of fluid infused into the stomach (0, 400, 1,000, 2,000, and 3,000 mL). Bias and agreement between the 2 methods were determined using Bland-Altman analysis and Lin's concordance correlation coefficients. MEASUREMENTS AND MAIN RESULTS: Mean gastric pressure was 14.44 ± 4.69 cm H(2)O and ranged from 0 to 25.8 cm H(2)O. Intraperitoneal pressure measurements were generally subatmospheric, and ranged from -6.6 to 3.1 cm H(2) O (mean ± SD, -1.59 ± 2.09 cm H(2)O). Measurements of intraperitoneal pressures were repeatable; however, intra- and interindividual variance was significantly larger for measurements of gastric pressures. The mean and relative bias for comparison between the 2 techniques was 15.9 ± 5.3 cm H(2)O and 244.3 ± 199.2%, respectively. The Lin's concordance correlation coefficient between gastric and intraperitoneal pressures was -0.003 but this was not statistically significant (P=0.75). CONCLUSIONS: There was no statistical concordance between measurements of intraabdominal pressure using gastric and intraperitoneal pressure measurement, indicating that gastric pressures cannot be substituted for intraperitoneal pressure measurement. Direct measurement of intraperitoneal pressures may be a more consistent method for comparison of intraabdominal pressures between horses, due to less variability within and between individuals.