scholarly journals Extracting biological age from biomedical data via deep learning: too much of a good thing?

2017 ◽  
Author(s):  
Tim Pyrkov ◽  
Konstantin Slipensky ◽  
Mikhail Barg ◽  
Alexey Kondrashin ◽  
Boris Zhurov ◽  
...  

Aging-related physiological changes are systemic and, at least in humans, are linearly associated with age. Therefore, linear combinations of physiological measures trained to estimate chronological age have recently emerged as a practical way to quantify aging in the form of biological age. Aging acceleration, defined as the difference between the predicted and chronological age was found to be elevated in patients with major diseases and is predictive of mortality. In this work, we compare three increasingly accurate biological age models: metrics derived from unsupervised Principal Components Analysis (PCA), alongside two supervised biological age models; a multivariate linear regression and a state-of-the-art deep convolution neural network (CNN). All predictions were made using one-week long locomotor activity records from a 2003-2006 National Health and Nutrition Examination Survey (NHANES) dataset. We found that application of the supervised approaches improves the accuracy of the chronological age estimation at the expense of a loss of the association between the aging acceleration predicted by the model and all-cause mortality. Instead, we turned to the NHANES death register and introduced a novel way to train parametric proportional hazards models in a form suitable for out-of-the-box implementation with any modern machine learning software. Finally, we characterized a proof-of-concept example, a separate deep CNN trained to predict mortality risks that outperformed any of the biological age or simple linear proportional hazards models. Our findings demonstrate the emerging potential of combined wearable sensors and deep learning technologies for applications involving continuous health risk monitoring and real-time feedback to patients and care providers.

2021 ◽  
Vol 5 (2) ◽  
pp. 48
Author(s):  
Otty Ratna Wahyuni ◽  
Deny Saputra ◽  
Nastiti Faradilla Ramadhani ◽  
Dennaya Listya Dias

Objectives: The principle of measurement using the TCI (Tooth Coronal Index) method is to compare the pulp chamber height with a person's chronological age based on the formation of secondary dentin. The purpose of this study is to estimate age based on pulp chamber height in lower canines using periapical radiographs with TCI measurement. Materials and Methods: This study is an observational analytic study using 42 samples of periapical radiographs with the parallel technique of the lower canines. Samples were measured for CH and CPCH heights to determine TCI values and then linear regression was made to determine their biological age. Finally, the difference between biological and chronological age is calculated to determine the approximate age. Results: The mean difference between chronological age and biological age was ± 5.05 years and an average biological age of 29.38 years. Conclusion: TCI method based on pulp chamber height in lower canines using periapical radiographs can be used to estimate age with the difference between chronological age and biological age of ±5.05 years.


2019 ◽  
Vol 3 (Supplement_1) ◽  
pp. S322-S322
Author(s):  
Sarah N Forrester ◽  
David D McManus ◽  
Jane S Saczynski ◽  
Catarina I Kiefe

Abstract Atrial Fibrillation (AF) is associated with dementia and cognitive decline. AF is less prevalent among Blacks than Whites, although AF-related complications are more common in Blacks. In the general population, all-cause cognitive decline and dementia are more prevalent among Blacks than Whites. Thus, studying diverse populations with AF may advance our understanding of racial disparities in cognitive functioning. We created a measure of multisystem dysregulation (weathering), which includes but is more encompassing than aging, and examined its association with racial differences in cognition using data from the SAGE-AF study, a prospective cohort of >65-year olds with AF, at high stroke risk, and eligible for anticoagulation. Biological (as opposed to chronological) age among 974 participants was calculated using the Klemera and Doubal method using biomarkers representing physiological functioning, metabolism, and blood pressure. We defined weathering as the difference between biological and chronological age (weathering >0 indicates that biological age is higher than chronological age). We measured the association between weathering and the Montreal Cognitive Assessment (MoCA) score. Mean weathering (SD) was -0.7 (11.5) and 4.3 (12.6) for whites and non-whites, respectively. There was an interaction between race/ethnicity and weathering on cognition (P=0.004). In stratified analyses, higher weathering was associated with a lower MoCA score among both Whites and non-Whites but more so among non-whites (B = -0.09, 95% CI: -0.17, -0.02) for Whites (B = -0.03, 95% CI: -0.06, -0.01) for non-whites. Aging-related multisystem dysregulation is more strongly associated with worse cognition in non-whites than in whites.


e-GIGI ◽  
2021 ◽  
Vol 9 (1) ◽  
Author(s):  
Niluh R. Woroprobosari ◽  
Devina V. Wisaputri ◽  
Muhammad H. Ni'am

Abstract: Unexpected incident such as natural disaster and accident often occur in many countries including Indonesia which causes many victims with unknown identity. Tooth is one of the indicators to assess and determine a person's identity. Blenkin-Taylor method is used for age estimation of an individual by using teeth. This study was aimed to obtain the estimation of biological age by using Blenkin-Taylor method in Semarang. This was a descriptive study with a cross sectional design. Samples were panoramic digital radiograph data of patients aged 5-15 years, copied in the form of a soft file. The observation and measurement were performed on seven teeth of right lower jaw by using the DICOM RadiAnt application. Data of observations and measurements of maturation scores were calculated and converted into the Blenkin-Taylor formula to determine the biological age. The results showed that the difference between biological and chronological age was ±0.32 years. This value was lower than the Blenkin-Taylor previous study result which was ±0,6 years. In conclusion, by using the Blenkin-Taylor method, there was a difference between biological age and chronological age as many as ±0,32 years in individuals aged 5-15 years old in Semarang.Keywords: biological age, the Blenkin-Taylor method, panoramic radiography Abstrak: Kejadian tidak terduga seperti bencana alam dan kecelakaan sering terjadi di berbagai negara, salah satunya di Indonesia yang menimbulkan banyak korban jiwa yang tidak diketahui identitasnya. Gigi merupakan salah satu indikator untuk menilai dan menentukan identitas seseorang. Salah satu metode dalam menentukan estimasi usia dengan menggunakan gigi ialah metode Blenkin-Taylor. Penelitian ini bertujuan untuk mendapatkan gambaran estimasi usia biologis dengan menggunakan metode Blenkin-Taylor di Kota Semarang. Jenis penelitian ialah deskriptif dengan desain potong lintang. Sampel penelitian ialah data file digital radiograf panoramik pasien berusia 5-15 tahun yang disalin ke dalam bentuk soft file, kemudian dilakukan pengamatan dan pengukuran pada 7 gigi  regio  kanan  rahang  bawah  dengan  menggunakan  aplikasi  RadiAnt DICOM. Hasil pengamatan dan pengukuran skor maturasi dihitung dan dikonversikan ke dalam rumus metode Blenkin-Taylor untuk menentukan usia biologis. Hasil penelitian menunjukkan bahwa selisih usia biologis dan usia kronologis sebesar 0,32 tahun. Hal ini lebih kecil dibandingkan penelitian Blenkin-Taylor terdahulu sebesar 0,6 tahun. Simpulan penelitian ini ialah dengan mengggunakan metode Blenkin-Taylor terdapat selisih rerata usia kronologis dan usia biologis sebesar ± 0,32 tahun pada individu usia 5-15 tahun di Kota Semarang.Kata kunci: usia biologis, metode Blenkin-Taylor, radiograf panoramik


2018 ◽  
Vol 8 (1) ◽  
Author(s):  
Timothy V. Pyrkov ◽  
Konstantin Slipensky ◽  
Mikhail Barg ◽  
Alexey Kondrashin ◽  
Boris Zhurov ◽  
...  

2004 ◽  
Vol 16 (1) ◽  
pp. 45-49 ◽  
Author(s):  
P.Y. Lee ◽  
E.M. Khoo

70 patients presented with acute asthma exacerbation requiring nebulised bronchodilator treatment at the emergency department of a teaching hospital in Kuala Lumpur, Malaysia, were interviewed over a two-week period in July 2001. The results showed that 45 (64%) patients had not been educated on the nature of asthma; 30 (43%) had not been advised on preventive measures or avoidance of triggers; 54 (77%) were not advised about the medications used and their side effects; 42 (60%) patients did not know the difference between reliever and preventive medications; 37 (53%) were unable to recognize features of worsening asthma and 68 (97%) were not told about the danger of non-prescribed self-medication or traditional medications. Only six (9%) patients were using peak flow meters and were taught self-management plans. The multiple regression results suggest that patients who were followed up at teaching hospital based clinics were better educated on asthma. In conclusion, asthmatic patients are still not educated well about their disease. Health care providers need to put more emphasis on asthma education so that the number of emergency room visits can be reduced. Asia Pac J Public Health 2004; 16(1): 45-49.


2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
I.D Poveda Pinedo ◽  
I Marco Clement ◽  
O Gonzalez ◽  
I Ponz ◽  
A.M Iniesta ◽  
...  

Abstract Background Previous parameters such as peak VO2, VE/VCO2 slope and OUES have been described to be prognostic in heart failure (HF). The aim of this study was to identify further prognostic factors of cardiopulmonary exercise testing (CPET) in HF patients. Methods A retrospective analysis of HF patients who underwent CPET from January to November 2019 in a single centre was performed. PETCO2 gradient was defined by the difference between final PETCO2 and baseline PETCO2. HF events were defined as decompensated HF requiring hospital admission or IV diuretics, or decompensated HF resulting in death. Results A total of 64 HF patients were assessed by CPET, HF events occurred in 8 (12.5%) patients. Baseline characteristics are shown in table 1. Patients having HF events had a negative PETCO2 gradient while patients not having events showed a positive PETCO2 gradient (−1.5 [IQR −4.8, 2.3] vs 3 [IQR 1, 5] mmHg; p=0.004). A multivariate Cox proportional-hazards regression analysis revealed that PETCO2 gradient was an independent predictor of HF events (HR 0.74, 95% CI [0.61–0.89]; p=0.002). Kaplan-Meier curves showed a significantly higher incidence of HF events in patients having negative gradients, p=0.002 (figure 1). Conclusion PETCO2 gradient was demonstrated to be a prognostic parameter of CPET in HF patients in our study. Patients having negative gradients had worse outcomes by having more HF events. Time to first event, decompensated heart Funding Acknowledgement Type of funding source: None


Sensors ◽  
2021 ◽  
Vol 21 (8) ◽  
pp. 2611
Author(s):  
Andrew Shepley ◽  
Greg Falzon ◽  
Christopher Lawson ◽  
Paul Meek ◽  
Paul Kwan

Image data is one of the primary sources of ecological data used in biodiversity conservation and management worldwide. However, classifying and interpreting large numbers of images is time and resource expensive, particularly in the context of camera trapping. Deep learning models have been used to achieve this task but are often not suited to specific applications due to their inability to generalise to new environments and inconsistent performance. Models need to be developed for specific species cohorts and environments, but the technical skills required to achieve this are a key barrier to the accessibility of this technology to ecologists. Thus, there is a strong need to democratize access to deep learning technologies by providing an easy-to-use software application allowing non-technical users to train custom object detectors. U-Infuse addresses this issue by providing ecologists with the ability to train customised models using publicly available images and/or their own images without specific technical expertise. Auto-annotation and annotation editing functionalities minimize the constraints of manually annotating and pre-processing large numbers of images. U-Infuse is a free and open-source software solution that supports both multiclass and single class training and object detection, allowing ecologists to access deep learning technologies usually only available to computer scientists, on their own device, customised for their application, without sharing intellectual property or sensitive data. It provides ecological practitioners with the ability to (i) easily achieve object detection within a user-friendly GUI, generating a species distribution report, and other useful statistics, (ii) custom train deep learning models using publicly available and custom training data, (iii) achieve supervised auto-annotation of images for further training, with the benefit of editing annotations to ensure quality datasets. Broad adoption of U-Infuse by ecological practitioners will improve ecological image analysis and processing by allowing significantly more image data to be processed with minimal expenditure of time and resources, particularly for camera trap images. Ease of training and use of transfer learning means domain-specific models can be trained rapidly, and frequently updated without the need for computer science expertise, or data sharing, protecting intellectual property and privacy.


Sensors ◽  
2020 ◽  
Vol 21 (1) ◽  
pp. 194
Author(s):  
Sarah Gonzalez ◽  
Paul Stegall ◽  
Harvey Edwards ◽  
Leia Stirling ◽  
Ho Chit Siu

The field of human activity recognition (HAR) often utilizes wearable sensors and machine learning techniques in order to identify the actions of the subject. This paper considers the activity recognition of walking and running while using a support vector machine (SVM) that was trained on principal components derived from wearable sensor data. An ablation analysis is performed in order to select the subset of sensors that yield the highest classification accuracy. The paper also compares principal components across trials to inform the similarity of the trials. Five subjects were instructed to perform standing, walking, running, and sprinting on a self-paced treadmill, and the data were recorded while using surface electromyography sensors (sEMGs), inertial measurement units (IMUs), and force plates. When all of the sensors were included, the SVM had over 90% classification accuracy using only the first three principal components of the data with the classes of stand, walk, and run/sprint (combined run and sprint class). It was found that sensors that were placed only on the lower leg produce higher accuracies than sensors placed on the upper leg. There was a small decrease in accuracy when the force plates are ablated, but the difference may not be operationally relevant. Using only accelerometers without sEMGs was shown to decrease the accuracy of the SVM.


Sign in / Sign up

Export Citation Format

Share Document