AI Model Analyzes Cells in Tissue Samples without Need for Trained Pathologist

Fast and precise information regarding the operated tissue is crucial for guiding a surgeon's next steps during cancer surgery. In cases where solid tumors are present in a cancer patient, the surgeon typically sends a biopsy sample to a pathologist for a rapid assessment. The pathologist must determine, among other things, whether the tissue is healthy, the extent of cancer's spread into the organs, etc. The traditional intraoperative diagnostic process is laborious, time-consuming, and resource-intensive. Now, scientists have developed a new technique that can perform a reliable analysis of solid tumors in as little as 30 minutes, without the need for a trained pathologist.

A research team from the Max Planck Institute for the Science of Light (MPL, Erlangen, Germany) has created a novel technique that enables clinicians to analyze cells in tissue samples from cancer patients quickly and precisely, without requiring the expertise of a trained pathologist. The team used artificial intelligence (AI) to evaluate the data generated by their method. For their study, the researchers utilized a tissue grinder to quickly tear apart biopsy samples down to the single-cell level. Subsequently, these single cells were analyzed using real-time deformability cytometry (RT-DC), an approach that is label-free and capable of examining the physical properties of up to 1,000 cells per second. This method is 36,000 times faster than the conventional methods used to evaluate cell deformability.

RT-DC involves pushing single cells at high speed through a microscopic channel, where they undergo deformation due to stress and pressure. Images are captured of each cell, which are then utilized by scientists to ascertain a variety of physical characteristics of the cells, including their size, shape, and deformability. However, solely conducting a physical analysis of cells is insufficient for diagnostic purposes. Physicians must be able to interpret these outcomes independently, without the need for the expertise of a trained pathologist or physicist. Therefore, to accomplish this, the researchers combined the tissue grinder and RT-DC with AI. The AI model evaluates the extensive, complex datasets obtained through RT-DC analysis and rapidly assesses whether a biopsy sample comprises cancerous tissue or not. Furthermore, the use of AI confirmed the significance of cell deformability as a biomarker, as the outcomes were markedly inferior when the AI was not trained with this variable.

Overall, the complete procedure, which includes sample processing and automated data analysis, can be executed in under 30 minutes, making it sufficiently fast to be carried out during surgery. One significant advantage of this method is that it does not require the immediate availability of a pathologist to analyze the sample. This is particularly advantageous since intraoperative consultations may not always be feasible, and in some cases, samples can only be examined after the surgery is completed. Based on the results, patients may need to return to the hospital for further surgery, often days later. Aside from testing for tumor presence, this technique was also utilized to detect tissue inflammation in a model of inflammatory bowel disease (IBD). In the future, this method could assist clinicians in evaluating disease severity or distinguishing between various types of IBD. The team aims to eventually transition their method into a clinical setting to support or perhaps even supplant the traditional pathological analysis.

“This was a proof of concept study - the method could accurately determine the presence of tumor tissue in our samples very quickly,” said Dr. Despina Soteriou, a member of the research team. “The next step will be to continue to work very closely with clinicians to determine how this method can best be translated into the clinic.”

Related Links:
MPL 

Visit expo >

Flocked Swab
HydraFlock and PurFlock Ultra

New

Platinum Supplier

Diagnostic Reader

Acucy System

New

Auto Clinical Chemistry Analyzer

ChemiLab 100

New

Lung Cancer Molecular Diagnostic Test

LENA S1