Detection and classification of hepatocytes and hepatoma cells using atomic force microscopy and machine learning...

Area of Science:

• Biophysics
• Nanotechnology
• Machine Learning in Medicine

Background:

• Hepatocellular carcinoma (HCC) is a high-risk malignancy originating from transformed hepatocytes.
• HCC cells possess unique surface nanofeatures distinct from normal hepatocytes.
• Accurate early diagnosis of HCC is crucial for patient outcomes.

Purpose of the Study:

• To develop a novel method for distinguishing between normal hepatocytes and hepatoma cells.
• To leverage atomic force microscopy (AFM) and machine learning (ML) for objective cell classification.
• To provide an improved diagnostic tool for early detection of hepatocellular carcinoma.

Main Methods:

• Atomic force microscopy (AFM) was employed to capture 3D morphology and mechanical properties (elastic modulus, viscoelasticity) of human hepatocytes and hepatoma cells.
• Extracted nanofeature data were used to train machine learning algorithms.
• Performance was evaluated against other ML models like support vector machine and logistic regression.

Main Results:

• The ML model achieved a classification accuracy of 94.54% and an AUC of 0.99 in distinguishing hepatocytes from hepatoma cells.
• The combined analysis of cell morphology and mechanics provided a robust classification basis.
• The developed method demonstrated superior classification performance compared to using single nano-parameters.

Conclusions:

• AFM combined with ML offers an objective and accurate method for identifying hepatocellular carcinoma cells based on their nanofeatures.
• This approach overcomes limitations of subjective microscopic analysis and aids in reducing diagnostic errors.
• The findings support the potential of this technique for the early and reliable diagnosis of HCC.