Non-Invasive Detection of Anaemia Using Digital Photographs of the Conjunctiva
Background and Aims Anaemia is a major health burden worldwide. Although the finding of conjunctival pallor on clinical examination is associated with anaemia, inter-observer variability is high, and definitive diagnosis of anaemia requires a blood sample. We aimed to detect anaemia by quantifying conjunctival pallor using digital photographs taken with a consumer camera and a popular smartphone. Our goal was to develop a non-invasive screening test for anaemia. Patients and Methods The conjunctivae of haemato-oncology in- and outpatients were photographed in ambient lighting using a digital camera (Panasonic DMC-LX5), and the internal rear-facing camera of a smartphone (Apple iPhone 5S) alongside an in-frame calibration card. Following image calibration, conjunctival erythema index (EI) was calculated and correlated with laboratory-measured haemoglobin concentration. Three clinicians independently evaluated each image for conjunctival pallor. Results Conjunctival EI was reproducible between images (average coefficient of variation 2.96%). EI of the palpebral conjunctiva correlated more strongly with haemoglobin concentration than that of the forniceal conjunctiva. Using the compact camera, palpebral conjunctival EI had a sensitivity of 93% and 57% and specificity of 78% and 83% for detection of anaemia (haemoglobin < 110 g/L) in training and internal validation sets, respectively. Similar results were found using the iPhone camera, though the EI cut-off value differed. Conjunctival EI analysis compared favourably with clinician assessment, with a higher positive likelihood ratio for prediction of anaemia. Conclusions Erythema index of the palpebral conjunctiva calculated from images taken with a compact camera or mobile phone correlates with haemoglobin and compares favourably to clinician assessment for prediction of anaemia. If confirmed in further series, this technique may be useful for the non-invasive screening for anaemia.
Code (0)
등록된 구현이 없습니다.
Tasks
SpecificitySimilar Papers 제목 키워드 기반
A smartphone based multi input workflow for non-invasive estimation of haemoglobin levels using machine learning techniques
We suggest a low cost, non invasive healthcare system that measures haemoglobin levels in patients and can be used as a preliminary diagnostic test for anaemia. A combination of image processing, machine learning and dee…
Active LearningDiagnosticPortable system for the prediction of anemia based on the ocular conjunctiva using Artificial Intelligence
Anemia is a major health burden worldwide. Examining the hemoglobin level of blood is an important way to achieve the diagnosis of anemia, but it requires blood drawing and a blood test. In this work we propose a non-inv…
AnemiaVision: Non-Invasive Anemia Detection via Smartphone Imagery Using EfficientNet-B3 with TrivialAugmentWide, Mixup Augmentation, and Persistent Patient History Management
Anemia affects over one billion people globally and remains severely under-diagnosed in low-resource regions where laboratory blood tests are inaccessible. This paper presents AnemiaVision, an end-to-end web-based system…
Image AugmentationDigital Image Tamper Detection Techniques - A Comprehensive Study
Photographs are considered to be the most powerful and trustworthy media of expression. For a long time, those were accepted as proves of evidences in varied fields such as journalism, forensic investigations, military i…
INTERACT-CMIL: Multi-Task Shared Learning and Inter-Task Consistency for Conjunctival Melanocytic Intraepithelial Lesion Grading
Accurate grading of Conjunctival Melanocytic Intraepithelial Lesions (CMIL) is essential for treatment and melanoma prediction but remains difficult due to subtle morphological cues and interrelated diagnostic criteria. …