in News
A close-up of a human eye. (Image credit: abbestock via Getty Images)Share this article 0Join the conversationFollow usAdd us as a preferred source on GoogleSubscribe to our newsletter
Scientists have created a method that utilizes brief recordings of the eye to gauge a person’s red blood cell levels — no hypodermic needles required.
The innovation, detailed in a publication on April 8 in the journal npj Digital Medicine, successfully identified anemia in over 80% of cases during a trial involving 224 subjects.
This technology is not yet prepared to supersede conventional blood draws, the researchers involved in the investigation cautioned. However, they believe it could potentially function as a preliminary screening instrument to identify individuals who might benefit from a comprehensive blood test. This could prove particularly advantageous in less affluent nations where laboratory diagnostic access is often restricted.
“Its potential benefit may lie in facilitating frequent, non-intrusive longitudinal observation or the early detection of patients needing further examination,” stated Dr. Christine Kiire, a consultant ophthalmologist at Oxford Eye Hospital and a visiting researcher in the artificial medical intelligence lab at the University College London Institute of Ophthalmology. If proven and made economically viable, the system could enhance the availability of blood monitoring in settings with limited resources, Kiire, who was not part of the study, informed Live Science via email.
The approach could be valuable in circumstances where repeated blood collection and analysis are inconvenient, remarked Dr. Theodore Leng, an ophthalmologist and vitreoretinal surgeon at Stanford University who was not involved in the research. This might encompass outpatient screening, remote patient monitoring, follow-up visits for dialysis and cancer therapies, or pediatric care, he conveyed in an email.
Nevertheless, the system is not yet ready for widespread clinical use. “It’s excellent research but will require numerous advancements to be clinically applicable,” commented Dr. Peter Campbell, an ophthalmologist at Oregon Health & Science University who was not part of the investigation, in an email.
How the needle-free system works
Noninvasive blood sensors are already in existence. In 2021, the Food and Drug Administration (FDA) authorized a device, the Pronto-7, which ascertains hemoglobin levels in the blood by emitting light through the fingernail. Hemoglobin is the protein responsible for carrying oxygen within blood cells.
Regrettably, Pronto-7’s measurements can be affected by skin pigmentation, indicating reduced accuracy for individuals with darker skin tones. The sclera, or the white portion of the eye, conversely, has minimal pigment and appears largely consistent across diverse populations.
The novel screening technique capitalizes on this characteristic. To develop it, researchers employed a microscope camera with 50x magnification to capture 10-second videos of the whites of study participants’ eyes. A software program named Video-to-Vessels refines the footage — removing blinks, eye movements, and illumination variations — and transforms the videos into time-lapse images of the ocular blood vessels.
Subsequently, an artificial intelligence model called VesselNet, which was trained using blood vessel images paired with laboratory blood count results, forecasts the individual’s hemoglobin concentration and red blood cell count by analyzing blood cell flow patterns.
“This paper is distinctive as it illustrates images of the anterior surface of the eye, rather than the retinal vasculature (at the posterior of the eye),” Campbell remarked. “Thus, in principle, it could be implemented without costly retinal cameras — even a smartphone.”
The investigators evaluated the methodology on 224 individuals, encompassing patients with blood disorders undergoing cancer treatment and healthy volunteers, at Sheba Medical Center in Israel. They contrasted the AI model’s predicted hemoglobin values with actual hemoglobin levels determined through standard blood tests, discovering that the model accurately identified low hemoglobin approximately 83% of the time.
This accuracy falls short of the requirements for practical application, according to Kiire. For perspective, Pronto-7 achieves scores between 80% and 88% for detecting low hemoglobin in men, and 84% to 87% in women.
“In a practical sense, this appears more suited as a preliminary detection tool, rather than a technology ready to guide dosage, transfusion decisions, or definitive hematological management,” Leng opined.
Related stories
- In a 1st, two people receive transfusions of lab-grown blood cells
- Gene mutation helps Andean highlanders thrive at altitude, and ‘living fossil’ fish live deep underwater
- Lab-made universal blood could revolutionize transfusions. Scientists just got one step closer to making it.
Furthermore, while the eye-based method quantifies only two parameters — hemoglobin and red blood cell count — a standard blood test assesses numerous others, Kiire pointed out. The study authors speculate that they might be able to enumerate white blood cells as well, by developing a camera with enhanced resolution and magnification capabilities.
Kiire also highlighted that certain ocular conditions, such as conjunctivitis and dry eye syndrome, along with medications like therapeutic eye drops, could influence the eye’s blood vessels and result in inaccurate readings. From a practical standpoint, achieving reliable results “necessitates careful patient positioning and adequate optical focus,” skills that are not easily mastered by clinicians and could restrict the utility of the new method, she elaborated.
The research team intends to conduct further investigations involving larger and more varied participant groups, including individuals with iron-deficiency anemia, who were underrepresented in the current study. They also aim to perform repeated evaluations of their methodology to confirm and broaden their findings.
