Hengtao Guo

Google Scholar  /  LinkedIn  /  Research Gate  /  Github

I am a Machine Learning Software Engineer at Google, where I focus on building recommendation systems in YouTube main app. Before joining Google, I have worked as a Research Intern at UII America.

I received my Ph.D. in Biomedical Engineering from Rensselaer Polytechnic Institute in 2022, and B.S. in Computer Science from Wuhan University in 2018.

Research Fields

My research focuses span across machine learning, deep learning, and generative artificial intelligence:

• Computer Vision: Medical image analysis, image registration/segmentation, ultrasound reconstruction
• Recommendation Systems: User modeling, video/audio representation learning
• Natural Language Processing: Large language models, retrival-augmented generation

Publications

We propose a learning-based method that uses medical scans to predict internal organ deformation across various human poses, aiding radiotherapy and similar treatments. This approach builds a patient-specific model encoding the organ's shape and elasticity, allowing for deformation estimation based on the patient's current pose. This innovation offers clinicians precise guidance without additional scans or procedures.

We propose a deep contextual learning network (DCL-Net), which can efficiently exploit the image feature relationship between US frames and reconstruct 3D US volumes without any tracking device. The proposed DCL-Net utilizes 3D convolutions over a US video segment for feature extraction. An embedded self-attention module makes the network focus on the speckle-rich areas for better spatial movement prediction, with a novel case-wise correlation loss to stabilize the training process for improved accuracy.

We propose a deep contextual-contrastive network, utilizing self-attention to focus on the speckle-rich areas to estimate spatial movement and then minimizes a margin ranking loss for contrastive feature learning. We train and validate the model on two independent datasets, evaluated with including the DICE of reconstructed organ segmentation.

This paper introduces a knowledge-based analytical method using deep convolutional neural network (CNN) for all-cause mortality prediction. The underlying approach combines structural image features extracted from CNNs, based on LDCT volume at different scales, and clinical knowledge obtained from quantitative measurements, to predict the mortality risk of lung cancer screening subjects.

We propose a novel US frame-to-volume registration (FVReg) pipeline to bridge the dimensionality gap between 2-D US frames and 3-D US volume. The developed pipeline is implemented using deep neural networks, which are fully automatic without requiring external tracking devices. We validated our method on a clinical dataset with 618 subjects and tested its potential on real-time 2-D-US to 3-D-MR fusion navigation tasks. The proposed FVReg achieved an average target navigation error of 1.93 mm at 5 to 14 fps.

Cancer patients have a higher risk of cardiovascular disease (CVD) mortality than the general population. Low dose computed tomography (LDCT) for lung cancer screening offers an opportunity for simultaneous CVD risk estimation in at-risk patients. Our deep learning CVD risk prediction model, trained with 30,286 LDCTs from the National Lung Cancer Screening Trial, achieves an area under the curve (AUC) of 0.871 on a separate test set of 2,085 subjects and identifies patients with high CVD mortality risks (AUC of 0.768).

Review Service

• Computer Vision and Pattern Recognition (CVPR)
• European Conference on Computer Vision (ECCV)
• Medical Image Computing and Computer Assisted Intervention (MICCAI)
• Machine Learning in Medical Imaging
• Computerized Medical Imaging and Graphics
• Artificial Intelligence with Biased or Scarce Data
• Frontiers in Cardiovascular Medicine
• Neurocomputing