MULTICOLOUR FLOW CYTOMETRY (MFC)
Multicolour Flow Cytometry (MFC) is a technology for analyzing cell characteristic with high sensitivity: it can detect up to one divergent cell amongst one million normal cells.
01
What is Multicolour Flow Cytometry (MFC)
Multicolour Flow Cytometry (MFC) is a technology for analyzing cell characteristic with high sensitivity: it can detect up to one divergent cell amongst one million normal cells.
Multicolour Flow cytometry is a powerful tool, it has seen dramatic advances over the last 30 years, allowing unprecedented detail in studies of the immune system and other areas of cell biology. Advancements in this technique have led to the possibility of measuring dozens of characteristics per individual cell, providing large volumes of raw data.
02
What happens to collected data?
The crowded data output is currently manually interpreted by the expert for diagnosis and monitoring. Translating the data into relevant information requires lots of expertise, time, and experience. Moreover, the process is restricted to the availability of clinical expert and presents risks of human biases and errors.
03
Clinical Challenge
The fast growing advancements in MFC have led to the possibility of measuring dozens of characteristics per individual cell, providing large volumes of raw data. The crowded data output is currently manually interpreted by the expert for diagnosis and monitoring. Translating the data into relevant information requires lots of expertise, time, and experience. Moreover, the process is restricted to the availability of clinical expert and presents risks of human biases and errors.
The Clinical Challenge
Complexity of analysis
- multiple markers
- multiple lasers
- manual decision steps
- experienced based job
- human biases and errors
Time Consuming
- pre-processing
- gating
- manual identification of cell populations of interest
- operator dependent
- high dimensionality of data
Large Datasets
- manual analysis of homogeneous cell population
- subjective but critical to correct findings
- technological evolution FC devices
Complexity of analysis
- multiple markers
- multiple lasers
- manual decision steps
- experienced based job
- human biases and errors
Large Datasets
- manual analysis of homogeneous cell population
- subjective but critical to correct findings
- technological evolution FC devices
Time Consuming
- pre-processing
- gating
- manual identification of cell populations of interest
- operator dependent
- high dimensionality of data
OUR CLINICAL CHALLENGE
Asudes is the answer.
To tackle this clinical challenge, researchers of the Radboud University and the University Medical Centre Utrecht developed an AI algorithm that can analyze the large volumes of raw data produced by the MFC and represents the findings as a simplified 2D-image.
The algorithm filters out all the normal cells, leading to a non-crowded representation of the abnormal cells only. Consequently, interpretation requires less expertise, less time and is trained to detect early indications of e.g. relapse.
Eclipse Algorithm
The novel method ECLIPSE (Elimination of Cells Lying in Patterns Similar to Endogeneity) identifies and characterize aberrant cells present in individuals out of homeostasis. ECLIPSE combines dimensionality reduction by Simultaneous Component Analysis with Kernel Density Estimates.
1. STEP
PCA model on the control cells + projection of cells from patients
2. STEP
Elimination of healthy-like cells in patients:
“equal to normal” as principle
3. STEP
PCA model on abnormal / disease specific cells