Clinical Research Directory
Browse clinical research sites, groups, and studies.
AI-Assisted Fluorescence Microscopy of Leukocytes in Fresh Whole Blood
Sponsor: MIPO Clinic
Summary
The goal of this observational study is to learn whether a microscope-based method with artificial intelligence (AI) can detect and classify white blood cells in fresh, unstained blood and measure the color of fluorescent light inside each cell, in healthy adult volunteers. The fluorescent dye is thought to reflect mitochondria (the parts of a cell that produce energy). The main questions it aims to answer are: 1. How accurately can the AI program detect white blood cells in fresh blood in the tile images? 2. Does the pattern of green, yellow, and red fluorescence inside individual white blood cells differ between dyes? Researchers will take microscope images of the same areas of each blood smear in regular light and in fluorescent light. The AI program outlines each white blood cell in the regular-light image, and a trained operator accepts, corrects, or rejects every outline. The accepted outlines are then used to measure the fluorescent colors in the same cells. As a reference, results are checked against standard stained blood smears (Romanowsky-Giemsa) read by laboratory staff. Participants will: Give a small blood sample by venipuncture (a needle in a vein of the arm) Repeat the blood draw about once a week for 3 to 6 weeks This study does not test any treatment, and participants will not receive any intervention.
Official title: Fluorescence Microscopy With Artificial Intelligence for Automated Single-Cell Analysis of Leukocytes in Native Whole Blood
Key Details
Gender
MALE
Age Range
18 Years - 45 Years
Study Type
OBSERVATIONAL
Enrollment
5
Start Date
2026-05-01
Completion Date
2026-09-23
Last Updated
2026-10-07
Healthy Volunteers
Yes
Interventions
AI-Assisted bright-field and fluorescence microscopy
Automated scanning of fresh, unfixed blood smears stained with 10-N-nonyl acridine orange and a second dye (Janus) on a motorized-stage microscope (100x or 60x objective, 45 positions per scan). Bright-field and fluorescence images of the same positions are taken. An AI detector (YOLO11) outlines white blood cells in the bright-field image; a trained operator accepts, corrects, or rejects each outline. Accepted outlines are copied to the fluorescence image, and green, yellow, and red signal is measured inside each cell. Smears with other dyes (e.g., pyronin Y) are scanned the same way for comparison. Research use only; no result is used for patient care.
Romanowsky-Giemsa stained blood smear (reference)
Standard Romanowsky-Giemsa stained blood smear (monolayer, CLSI H20) prepared from the same blood samples. White blood cell differential is counted by laboratory staff and used as the reference for checking the AI program's detection and classification.
Locations (1)
LMB Clinic
Almaty, Almaty Region, Kazakhstan