Label-free multiphoton microscopy and machine learning for recognition of hepatocellular carcinoma

2026-03-10 · Source: Machine learning : nature.com subject feeds · Field: Science & Research — Health & Medical Research, Engineering & Applied Sciences, Mathematics & Computational Sciences · Depth: Advanced, long

Summary

Researchers have developed a label-free multiphoton microscopy (MPM) and machine learning approach for recognizing hepatocellular carcinoma (HCC) and evaluating intraoperative resection margins. The study utilized a multimodal imaging technique, including coherent anti-Stokes Raman scattering, two-photon autofluorescence, and second harmonic generation, on matched HCC and background liver tissue samples from 76 patients. Morphological information from these channels was condensed into 17 texture parameters for classification. A neural network model, trained on approximately 25,000 images from 35 patients, achieved a 97.3% correct rate on a test set of 27,000 images from 38 patients, with 98.2% for liver and 96.5% for tumor. The analysis highlighted autofluorescence's critical role in distinguishing neoplastic from non-neoplastic tissue, demonstrating the potential for real-time, on-site tissue analysis in surgical settings, even with low-lateral-resolution images mimicking endoscope use.

Key takeaway

For AI Scientists developing medical imaging diagnostics, this research demonstrates a robust method for intraoperative tumor recognition. Your focus should be on integrating label-free multiphoton microscopy with machine learning to enhance surgical precision. Consider optimizing neural network architectures for texture parameter classification, particularly leveraging autofluorescence signals, to improve real-time assessment of resection margins and potentially reduce recurrence rates in oncological liver surgery.

Key insights

Label-free multiphoton microscopy combined with machine learning accurately identifies hepatocellular carcinoma.

Principles

• Autofluorescence is key for tissue discrimination.
• Low-resolution MPM images enable accurate tumor recognition.

Method

A neural network was trained on 17 texture parameters derived from multimodal MPM images (CARS, 2P-autofluorescence, SHG) to classify HCC and background liver tissue, achieving high accuracy on a test set.

In practice

• Integrate MPM into endoscopes for real-time analysis.
• Utilize autofluorescence for rapid tissue differentiation.

Topics

• Hepatocellular Carcinoma
• Multiphoton Microscopy
• Machine Learning
• Intraoperative Imaging
• Texture Analysis

Best for: AI Scientist, AI Researcher, Research Scientist, AI Engineer

Related on AIssential

• Label-free pathological subtyping of non-small cell lung cancer using deep classification and virtual immunohistochemical staining — Label-free autofluorescence imaging combined with deep learning accurately subtypes NSCLC and generates virtual IHC stains.
• Explainable Ensemble-Based Machine Learning Models for Detecting the Presence of Cirrhosis in Hepatitis C Patients — Machine learning, particularly Extra Trees models, can accurately detect cirrhosis in Hepatitis C patients, offering crucial early diagnostic capabilities.
• Temporally Consistent Label Interpolation for Robust Surgical Multi-Task Learning under Challenging Conditions — FAROS interpolates sparse surgical annotations into dense, temporally consistent labels using optical flow, balancing multi-task learning.
• Efficient AI-Driven Multi-Section Whole Slide Image Analysis for Biochemical Recurrence Prediction in Prostate Cancer — A novel AI framework predicts prostate cancer biochemical recurrence using multi-section whole slide images, outperforming clinical benchmarks.
• Beyond Morphology: Quantifying the Diagnostic Power of Color Features in Cancer Classification — Pixel intensity alone, without morphology, provides a strong diagnostic signal for cancer classification.
• Parameter-Efficient Subspace Decoupling ViT for Mitigating Multi-Task Negative Transfer in Histological Scoring — Subspace-decoupled ViT with orthogonal adapters mitigates negative transfer in multi-task histological scoring, improving efficiency.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine learning : nature.com subject feeds.