-
Ophthalmology Science 2024To gain an understanding of data labeling requirements to train deep learning models for measurement of geographic atrophy (GA) with fundus autofluorescence (FAF) images.
PURPOSE
To gain an understanding of data labeling requirements to train deep learning models for measurement of geographic atrophy (GA) with fundus autofluorescence (FAF) images.
DESIGN
Evaluation of artificial intelligence (AI) algorithms.
SUBJECTS
The Age-Related Eye Disease Study 2 (AREDS2) images were used for training and cross-validation, and GA clinical trial images were used for testing.
METHODS
Training data consisted of 2 sets of FAF images; 1 with area measurements only and no indication of GA location (Weakly labeled) and the second with GA segmentation masks (Strongly labeled).
MAIN OUTCOME MEASURES
Bland-Altman plots and scatter plots were used to compare GA area measurement between ground truth and AI measurements. The Dice coefficient was used to compare accuracy of segmentation of the Strong model.
RESULTS
In the cross-validation AREDS2 data set (n = 601), the mean (standard deviation [SD]) area of GA measured by human grader, Weakly labeled AI model, and Strongly labeled AI model was 6.65 (6.3) mm, 6.83 (6.29) mm, and 6.58 (6.24) mm, respectively. The mean difference between ground truth and AI was 0.18 mm (95% confidence interval, [CI], -7.57 to 7.92) for the Weakly labeled model and -0.07 mm (95% CI, -1.61 to 1.47) for the Strongly labeled model. With GlaxoSmithKline testing data (n = 156), the mean (SD) GA area was 9.79 (5.6) mm, 8.82 (4.61) mm, and 9.55 (5.66) mm for human grader, Strongly labeled AI model, and Weakly labeled AI model, respectively. The mean difference between ground truth and AI for the 2 models was -0.97 mm (95% CI, -4.36 to 2.41) and -0.24 mm (95% CI, -4.98 to 4.49), respectively. The Dice coefficient was 0.99 for intergrader agreement, 0.89 for the cross-validation data, and 0.92 for the testing data.
CONCLUSIONS
Deep learning models can achieve reasonable accuracy even with Weakly labeled data. Training methods that integrate large volumes of Weakly labeled images with small number of Strongly labeled images offer a promising solution to overcome the burden of cost and time for data labeling.
FINANCIAL DISCLOSURES
Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.
PubMed: 38827491
DOI: 10.1016/j.xops.2024.100477 -
AMIA Joint Summits on Translational... 2024Relation Extraction (RE) is a natural language processing (NLP) task for extracting semantic relations between biomedical entities. Recent developments in pre-trained...
Relation Extraction (RE) is a natural language processing (NLP) task for extracting semantic relations between biomedical entities. Recent developments in pre-trained large language models (LLM) motivated NLP researchers to use them for various NLP tasks. We investigated GPT-3.5-turbo and GPT-4 on extracting the relations from three standard datasets, EU-ADR, Gene Associations Database (GAD), and ChemProt. Unlike the existing approaches using datasets with masked entities, we used three versions for each dataset for our experiment: a version with masked entities, a second version with the original entities (unmasked), and a third version with abbreviations replaced with the original terms. We developed the prompts for various versions and used the chat completion model from GPT API. Our approach achieved a F1-score of 0.498 to 0.809 for GPT-3.5-turbo, and a highest F1-score of 0.84 for GPT-4. For certain experiments, the performance of GPT, BioBERT, and PubMedBERT are almost the same.
PubMed: 38827097
DOI: No ID Found -
Research Square May 2024Traditional feature dimension reduction methods have been widely used to uncover biological patterns or structures within individual spatial transcriptomics data....
Traditional feature dimension reduction methods have been widely used to uncover biological patterns or structures within individual spatial transcriptomics data. However, these methods are designed to yield feature representations that emphasize patterns or structures with dominant high variance, such as the normal tissue spatial pattern in a precancer setting. Consequently, they may inadvertently overlook patterns of interest that are potentially masked by these high-variance structures. Herein we present our graph contrastive feature representation method called CoCo-ST (Comparing and Contrasting Spatial Transcriptomics) to overcome this limitation. By incorporating a background data set representing normal tissue, this approach enhances the identification of interesting patterns in a target data set representing precancerous tissue. Simultaneously, it mitigates the influence of dominant common patterns shared by the background and target data sets. This enables discerning biologically relevant features crucial for capturing tissue-specific patterns, a capability we showcased through the analysis of serial mouse precancerous lung tissue samples.
PubMed: 38826463
DOI: 10.21203/rs.3.rs-4359834/v1 -
Journal of Infection and Public Health Jul 2024Alzheimer's disease (AD) is a neurodegenerative disorder influenced by age, sex, genetic factors, immune alterations, and infections. Multiple lines of evidence suggest...
BACKGROUND
Alzheimer's disease (AD) is a neurodegenerative disorder influenced by age, sex, genetic factors, immune alterations, and infections. Multiple lines of evidence suggest that changes in antibody response are linked to AD pathology.
METHODS
To elucidate the mechanisms underlying AD development, we investigated antibodies that target autoimmune epitopes using high-resolution epitope microarrays. Our study compared two groups: individuals with AD (n = 19) and non-demented (ND) controls (n = 19). To validate the results, we measured antibody levels in plasma samples from AD patients (n = 96), mild cognitive impairment (MCI; n = 91), and ND controls (n = 97). To further explore the invlovement of EBV, we performed epitope masking immunofluorescence microscopy analysis and tests to induce lytic replication using the B95-8 cell line.
RESULTS
In this study, we analyzed high-resolution epitope-specific serum antibody levels in AD, revealing significant disparities in antibodies targeting multiple epitopes between the AD and control groups. Particularly noteworthy was the significant down-regulation of antibody (anti-DG#29) targeting an epitope of Epstein-Barr virus nuclear antigen 1 (EBNA1). This down-regulation increased AD risk in female patients (odds ratio up to 6.6), but not in male patients. Our investigation further revealed that the down-regulation of the antibody (anti-DG#29) is associated with EBV reactivation in AD, as indicated by the analysis of EBV VCA IgG or IgM levels. Additionally, our data demonstrated that the epitope region on EBNA1 for the antibody is hidden during the EBV lytic reactivation of B95-8 cells.
CONCLUSION
Our findings suggest a potential relationship of EBV in the development of AD in female. Moreover, we propose that antibodies targeting the epitope (DG#29) of EBNA1 could serve as valuable indicators of AD risk in female.
Topics: Humans; Alzheimer Disease; Female; Male; Epstein-Barr Virus Nuclear Antigens; Aged; Antibodies, Viral; Epitopes; Herpesvirus 4, Human; Cognitive Dysfunction; Aged, 80 and over; Epstein-Barr Virus Infections; Middle Aged
PubMed: 38824738
DOI: 10.1016/j.jiph.2024.05.050 -
Psychiatry Research Aug 2024This study examined trajectories of suicide-risk and their relationship to symptoms, recovery, and quality of life over time. Data was obtained from the Recovery after...
This study examined trajectories of suicide-risk and their relationship to symptoms, recovery, and quality of life over time. Data was obtained from the Recovery after an Initial Schizophrenia Episode Early Treatment Program (RAISE-ETP) study. 404 individuals with first-episode psychosis (FEP) completed measures of suicide-risk, depression, positive symptoms, recovery, and quality of life at baseline, 6mo, 12mo, 18mo, and 24mo. Latent class analysis was used to identify temporal trajectories of suicide-risk. General linear mixed models for repeated measures were used to examine the relationship between the latent trajectories of suicide-risk and clinical variables. Results identified three latent trajectories of suicide-risk (low-risk, worsening, and improving). The low-risk and improving classes experienced improvements in depression, positive symptoms, quality of life, and recovery over time. The worsening class experienced improvements in positive symptoms and quality of life, but no change in depression or recovery. These results suggest that some individuals with FEP are at risk for persistent depression and worsening suicide-risk during treatment despite experiencing improvements in positive symptoms and quality of life. These findings have important clinical implications, as persistent depression and worsening suicide-risk might be masked by the primary focus on positive symptoms and quality of life in most FEP clinics.
Topics: Humans; Psychotic Disorders; Female; Male; Adult; Young Adult; Quality of Life; Depression; Adolescent; Suicide; Schizophrenia
PubMed: 38823163
DOI: 10.1016/j.psychres.2024.115978 -
Plant Methods May 2024Traditional Chinese Medicinal Plants (CMPs) hold a significant and core status for the healthcare system and cultural heritage in China. It has been practiced and...
BACKGROUND
Traditional Chinese Medicinal Plants (CMPs) hold a significant and core status for the healthcare system and cultural heritage in China. It has been practiced and refined with a history of exceeding thousands of years for health-protective affection and clinical treatment in China. It plays an indispensable role in the traditional health landscape and modern medical care. It is important to accurately identify CMPs for avoiding the affected clinical safety and medication efficacy by the different processed conditions and cultivation environment confusion.
RESULTS
In this study, we utilize a self-developed device to obtain high-resolution data. Furthermore, we constructed a visual multi-varieties CMPs image dataset. Firstly, a random local data enhancement preprocessing method is proposed to enrich the feature representation for imbalanced data by random cropping and random shadowing. Then, a novel hybrid supervised pre-training network is proposed to expand the integration of global features within Masked Autoencoders (MAE) by incorporating a parallel classification branch. It can effectively enhance the feature capture capabilities by integrating global features and local details. Besides, the newly designed losses are proposed to strengthen the training efficiency and improve the learning capacity, based on reconstruction loss and classification loss.
CONCLUSIONS
Extensive experiments are performed on our dataset as well as the public dataset. Experimental results demonstrate that our method achieves the best performance among the state-of-the-art methods, highlighting the advantages of efficient implementation of plant technology and having good prospects for real-world applications.
PubMed: 38822406
DOI: 10.1186/s13007-024-01202-6 -
PloS One 2024Health personnel (HP) are on the frontlines during response to public health emergencies like COVID-19. This risk of exposure suggests the need for safety in responding...
SARS-CoV-2 active infection and antibodies amongst health personnel during the outbreak in Cameroon: Strengthening the health system for response to future public health emergencies.
BACKGROUND
Health personnel (HP) are on the frontlines during response to public health emergencies like COVID-19. This risk of exposure suggests the need for safety in responding to any pandemic. Therefore, to ascertain the rate of SARS-CoV-2 infection and immunity, and their determinants amongst HP become relevant.
METHODS
A cross sectional health facility-based study was carried-out amongst HP in the Centre Region of Cameroon from 1st February to 30th June 2021. Characteristics and access to preventive tools were collected using face-to-face administered questionnaire. Nasopharyngeal swabs and whole blood were collected for PCR, IgG and IgM testing respectively. STATA version 17 software was used for data analysis. Determinants of COVID-19 infection were explored by estimating crude and adjusted Odd Ratio.
RESULTS
Out of 510 HP reached, 458 were enrolled with mean age of 35 (±10) years. Thirty-four (7.4%) were PCR-positive to SARS-CoV-2 with 73.5% being clinicians versus 9 (26.4%) non-clinicians (p = 0.05). Sero-positivity to SARS-CoV-2 IgG/IgM was 40.2% (184/458), with 84.2% being clinicians versus 29 (15.8%) non-clinicians (p = 0.733). Amongst the 34 HP with PCR-positivity, 16 (47%) had no antibodies, while, 15 (44%) were IgG only. An estimate of HP (43.7%) had at least an evidence of PCR, IgG or IgM contact to COVID-19. Determinants of PCR-positivity was being clinical staff (AOR = 0.29, P = 0.039); and that of IgG/IgM were being non clinical staff (AOR = 0.41, p = 0.018) and regular use of face masks (AOR = 0.44, p = 0.001). HP trained on IPC (24%) were mainly from peripheral level (74.7%, p = 0.002).
CONCLUSION
Active infections were within the range of pandemic control (<10%). However, around two-fifths of participants have had contact with the virus, indicating that HP remains a population at risk of COVID-19 and other similarly-transmitted epidemic prone diseases, and also an important source of transmission. There is need of vaccine to achieve protectiveness, and optimal response also requires capacity building to improve the health system when challenged by a future pandemic.
Topics: Humans; COVID-19; Cameroon; Health Personnel; Male; Adult; Female; SARS-CoV-2; Cross-Sectional Studies; Antibodies, Viral; Middle Aged; Public Health; Disease Outbreaks; Immunoglobulin G; Immunoglobulin M
PubMed: 38820301
DOI: 10.1371/journal.pone.0304477 -
BMC Medical Informatics and Decision... May 2024Securing adequate data privacy is critical for the productive utilization of data. De-identification, involving masking or replacing specific values in a dataset, could...
BACKGROUND
Securing adequate data privacy is critical for the productive utilization of data. De-identification, involving masking or replacing specific values in a dataset, could damage the dataset's utility. However, finding a reasonable balance between data privacy and utility is not straightforward. Nonetheless, few studies investigated how data de-identification efforts affect data analysis results. This study aimed to demonstrate the effect of different de-identification methods on a dataset's utility with a clinical analytic use case and assess the feasibility of finding a workable tradeoff between data privacy and utility.
METHODS
Predictive modeling of emergency department length of stay was used as a data analysis use case. A logistic regression model was developed with 1155 patient cases extracted from a clinical data warehouse of an academic medical center located in Seoul, South Korea. Nineteen de-identified datasets were generated based on various de-identification configurations using ARX, an open-source software for anonymizing sensitive personal data. The variable distributions and prediction results were compared between the de-identified datasets and the original dataset. We examined the association between data privacy and utility to determine whether it is feasible to identify a viable tradeoff between the two.
RESULTS
All 19 de-identification scenarios significantly decreased re-identification risk. Nevertheless, the de-identification processes resulted in record suppression and complete masking of variables used as predictors, thereby compromising dataset utility. A significant correlation was observed only between the re-identification reduction rates and the ARX utility scores.
CONCLUSIONS
As the importance of health data analysis increases, so does the need for effective privacy protection methods. While existing guidelines provide a basis for de-identifying datasets, achieving a balance between high privacy and utility is a complex task that requires understanding the data's intended use and involving input from data users. This approach could help find a suitable compromise between data privacy and utility.
Topics: Humans; Confidentiality; Data Anonymization; Emergency Service, Hospital; Length of Stay; Republic of Korea; Male
PubMed: 38816848
DOI: 10.1186/s12911-024-02545-9 -
Irish Veterinary Journal May 2024The nutritional status in alpacas is often masked by their dense fibre coat. Its assessment is commonly approached by different body condition scores (BCS) that rely on...
BACKGROUND
The nutritional status in alpacas is often masked by their dense fibre coat. Its assessment is commonly approached by different body condition scores (BCS) that rely on manual palpation of defined anatomical regions. However, BCS is an important diagnostic tool to aid recognition of diseased South American camelids (SACs) and low BCS has been associated with conditions like anaemia and neutrophilia. For dose-dependent veterinary treatment, body weight (BW), that should be as accurate as possible, is required. As on-site weighing with scales is often not possible, BW can mostly only be roughly estimated. To date, it remains unclear whether BCS in alpacas provides reliable information on BW or the ratios of BW to body length commonly known as Body Mass Index (BMI) or Ponderal Index (PI). Equations to estimate BW based on body measurements are available in the literature. Nonetheless, respective equations were developed in growing alpacas or adult llamas and BCS was not included.
RESULTS
To compare six different BCS approaches and to examine the relationship between BCS and BW, body measurements and BCS scores were recorded in a herd of 105 alpacas. The examined BCS approaches showed significant (p < 0.05) but poor to moderate positive correlations to BW, BMI or PI. A solely visual inspection of BCS, in contrast, was not correlated with BW, BMI or PI. Equations previously developed in other studies provided an accurate estimation of BW. Multiple linear regression showed that the accuracy in predicting BW could be further increased by adding BCS data and sex.
CONCLUSION
Our observations indicate that most selected BCS approaches are not only important measures of nutritional status but can also be used to create more accurate models for BW calculation in alpacas. The study also supports the claim that a purely visual inspection of alpacas is not an adequate method to evaluate the nutritional status of these animals.
PubMed: 38816833
DOI: 10.1186/s13620-024-00274-z -
Turkish Journal of Medical Sciences 2023It was aimed to evaluate the positive effects of health behaviors (general hygiene, wearing face masks, physical distancing, and travel restrictions) acquired during the...
BACKGROUND/AIM
It was aimed to evaluate the positive effects of health behaviors (general hygiene, wearing face masks, physical distancing, and travel restrictions) acquired during the coronavirus disease 2019 (COVID-19) pandemic on the prevention of other infectious diseases in Ankara Province, Türkiye.
MATERIALS AND METHODS
This study was designed retrospectively. Among the notifiable group A infectious diseases, acute intestinal infections (AIIs) with International Classification of Diseases, Tenth Revision diagnosis codes A09 (diarrhea and gastroenteritis presumed to be of infectious origin), R11 (nausea and vomiting), and K52 (other noninfectious gastroenteritis and colitis), as well as influenza, tuberculosis, measles, varicella, malaria, and meningococcal meningitis were included in the scope of this study.The data of the selected infectious diseases in Ankara Province for the last 2 years before the pandemic (January 2018-December 2019) and for the 2-year period of the pandemic (January 2020-December 2021) were analyzed after checking the data. The number of cases were presented as frequencies, the 1-sample chi-squared test was used in the statistical analysis and the statistical significance level (α) was taken as 0.05.
RESULTS
The findings for each disease/disease group were discussed under separate headings. Comparing the prepandemic period (2018-2019) with the pandemic period (2020-2021), the decreases in the number of cases of selected infectious diseases, except influenza, were statistically significant.
CONCLUSION
Undoubtedly, the experience gained from the pandemic struggle will guide us in shaping our future lives. From this point forward, we should be aware that living in crowded environments and as a highly mobile population, that unhygienic habits are unfavorable for the spread of all infectious diseases, and we should take care to continuously apply the precautions for healthy living in our new lifestyle.
Topics: Humans; COVID-19; Retrospective Studies; Health Behavior; Turkey; Communicable Diseases; SARS-CoV-2; Hygiene; Masks; Physical Distancing; Pandemics; Communicable Disease Control; Travel
PubMed: 38813503
DOI: 10.55730/1300-0144.5745