Development of a Deep Learning Model to Assist With Diagnosis of Hepatocellular Carcinoma
Shi Feng1, Xiaotian Yu2, Wenjie Liang3
1Department of Pathology, The First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, China.
Insights
A novel deep learning model accurately diagnoses hepatocellular carcinoma (HCC) from pathological slides, even with noisy annotations. This AI tool enhances diagnostic efficiency and accuracy for this high-mortality cancer.
Area of Science:
- Oncology
- Pathology
- Artificial Intelligence
Background:
- Hepatocellular carcinoma (HCC) diagnosis is challenging, time-consuming, and relies heavily on pathologist expertise.
- Accurate pathological diagnosis is crucial for effective HCC treatment and management.
- Existing diagnostic methods for HCC can be subjective and prone to errors.
Purpose of the Study:
- To develop and validate a deep learning model for automated HCC diagnosis and classification.
- To improve the accuracy and efficiency of pathological diagnosis in HCC.
- To create a model capable of handling noisy annotations in histopathological images.
Main Methods:
- A noise-specific deep learning model was developed using whole-slide images from 592 HCC patients.
- The model employed patch screening and dynamic label smoothing to address noisy annotations.
- Training involved 137 HCC cases, with independent testing on 455 cases and validation on The Cancer Genome Atlas dataset.
Main Results:
- The model achieved 87.81% pixel-level and 98.77% slide-level accuracy on the independent test dataset.
- Validation on The Cancer Genome Atlas dataset yielded 87.90% accuracy.
- The deep learning approach significantly improved pixel-level accuracy compared to standard convolutional neural network (CNN) models.
Conclusions:
- The noise-specific deep learning model is effective for HCC classification, even with noisy annotations.
- The model demonstrates superior performance in detecting well-differentiated HCC and microvascular invasion.
- This AI-driven approach offers a promising tool to enhance pathological diagnosis in HCC.
Background:
An accurate pathological diagnosis of hepatocellular carcinoma (HCC), one of the malignant tumors with the highest mortality rate, is time-consuming and heavily reliant on the experience of a pathologist. In this report, we proposed a deep learning model that required minimal noise reduction or manual annotation by an experienced pathologist for HCC diagnosis and classification.
Methods:
We collected a whole-slide image of hematoxylin and eosin-stained pathological slides from 592 HCC patients at the First Affiliated Hospital, College of Medicine, Zhejiang University between 2015 and 2020. We propose a noise-specific deep learning model. The model was trained initially with 137 cases cropped into multiple-scaled datasets. Patch screening and dynamic label smoothing strategies are adopted to handle the histopathological liver image with noise annotation from the perspective of input and output. The model was then tested in an independent cohort of 455 cases with comparable tumor types and differentiations.
Results:
Exhaustive experiments demonstrated that our two-step method achieved 87.81% pixel-level accuracy and 98.77% slide-level accuracy in the test dataset. Furthermore, the generalization performance of our model was also verified using The Cancer Genome Atlas dataset, which contains 157 HCC pathological slides, and achieved an accuracy of 87.90%.
Conclusions:
The noise-specific histopathological classification model of HCC based on deep learning is effective for the dataset with noisy annotation, and it significantly improved the pixel-level accuracy of the regular convolutional neural network (CNN) model. Moreover, the model also has an advantage in detecting well-differentiated HCC and microvascular invasion.
