Artificial intelligence model predicts M2 macrophage levels and HCC prognosis with only globally labeled pathological images

Background and aims

The levels of M2 macrophages are significantly associated with the prognosis of hepatocellular carcinoma (HCC), however, current detection methods in clinical settings remain challenging. Our study aims to develop a weakly supervised artificial intelligence model using globally labeled histological images, to predict M2 macrophage levels and forecast the prognosis of HCC patients by integrating clinical features.

Methods

CIBERSORTx was used to calculate M2 macrophage abundance. We developed a slide-level, weakly-supervised clustering method for Whole Slide Images (WSIs) by integrating Masked Autoencoders (MAE) with ResNet-32t to predict M2 macrophage abundance.

Results

We developed an MAE-ResNet model to predict M2 macrophage levels using WSIs. In the testing dataset, the area under the curve (AUC) (95% CI) was 0.73 (0.59-0.87). We constructed a Cox regression model showing that the predicted probabilities of M2 macrophage abundance were negatively associated with the prognosis of HCC (HR=1.89, p=0.031). Furthermore, we incorporated clinical data, screened variables using Lasso regression, and built the comprehensive prediction model that better predicted prognosis. (HR=2.359, p=0.001).

Conclusion

Our models effectively predicted M2 macrophage levels and HCC prognosis. The findings suggest that our models offer a novel method for determining biomarker levels and forecasting prognosis, eliminating additional clinical tests, thereby delivering substantial clinical benefits.

MAE-Based Self-Supervised Pretraining Algorithm for Heart Rate Estimation of Radar Signals

Noncontact heart rate monitoring techniques based on millimeter-wave radar have advantages in unique medical scenarios. However, the accuracy of the existing noncontact heart rate estimation methods is still limited by interference, such as DC offsets, respiratory harmonics, and environmental noise. Additionally, these methods still require longer observation times. Most deep learning methods related to heart rate estimation still need to collect more heart rate marker data for training. To address the above problems, this paper introduces a radar signal-based heart rate estimation network named the "masked phase autoencoders with a vision transformer network" (MVN). This network is grounded on masked autoencoders (MAEs) for self-supervised pretraining and a vision transformer (ViT) for transfer learning. During the phase preprocessing stage, phase differencing and interpolation smoothing are performed on the input phase signal. In the self-supervised pretraining step, masked self-supervised training is performed on the phase signal using the MAE network. In the transfer learning stage, the encoder segment of the MAE network is integrated with the ViT network to enable transfer learning using labeled heart rate data. The innovative MVN offers a dual advantage-it not only reduces the cost associated with heart rate data acquisition but also adeptly addresses the issue of respiratory harmonic interference, which is an improvement over conventional signal processing methods. The experimental results show that the process in this paper improves the accuracy of heart rate estimation while reducing the requisite observation time.