Skip to content

Non-Invasive Estimation of Hemoglobin Levels in Blood Using Combined Deep Learning Methods

Non-Invasive Estimation of Hemoglobin Levels in Blood Using Combined Deep Learning Methods

Status
Completed
Phases
Unknown
Study type
Observational
Source
ClinicalTrials.gov
Registry ID
NCT04865224
Enrollment
353
Registered
2021-04-29
Start date
2020-07-15
Completion date
2020-12-30
Last updated
2021-04-29

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Hemoglobin Level Estimation, Hemoglobin Level Measurement

Keywords

non-invasive, deep learning, hemoglobin level estimation

Brief summary

Hemoglobin is an important protein that gives the blood its red color and helps transport oxygen to organs. There are quite a lot of people in the world who need to have their hemoglobin levels checked regularly, especially anemia patients. It is important to keep the hemoglobin levels in the blood of these individuals under control. For this, it is necessary to measure in frequent periods. A large group, from adults to children, needs to take these measurements. Most of these are invasive measurements made with blood that require devices and trained people. These painful methods are quite uncomfortable. In this study, non-invasive hemoglobin level measurement was aimed with a combined DL model created by using nail images with age, height, weight, BMI, and gender information of the volunteers.

Interventions

DIAGNOSTIC_TESThemoglobin estimation

In this method, in which different deep learning models are used together, the age range is intended to be as wide as possible to provide a quick solution with a high accuracy and low error values.

Sponsors

Karabuk University
Lead SponsorOTHER

Study design

Observational model
OTHER
Time perspective
CROSS_SECTIONAL

Eligibility

Sex/Gender
ALL
Healthy volunteers
Yes

Inclusion criteria

* Individuals need to be healthy. * Individuals should not have any active hand or nail disease. * Individuals should know their hemoglobin level.

Exclusion criteria

-In case of any condition that hides the natural nail color, such as henna, nail polish, protective or trauma-induced dents, bruises on the nails of individuals, it is removed from the study.

Design outcomes

Primary

MeasureTime frameDescription
Bland-Altman analysis1 dayBland-Altman analysis was performed to make the results simpler and more understandable. This analysis is especially used in the simple and graphical expression of measurements in which two methods are compared.
Estimation accuracy analysis1 dayIt is an analysis that compares the estimated hemoglobin values with the measured hemoglobin values.

Countries

Turkey (Türkiye)

Outcome results

None listed

Source: ClinicalTrials.gov · Data processed: Feb 4, 2026