Collaborative AI-for-Health Projects

We collaborate with various colleagues in the UC Davis Health - School of Medicine on applying Internet-of-Things (IoTs), signal processing, and machine learning/artificial intelligence (ML/AI) techniques to critical care risk assessments, disease detection, and pathology image analysis.

AI-Driven Neuropathologic Deep Phenotyping of Alzheimer's Disease

Lead Domain Expert: Dr. Brittany Dugger, Alzheimer's Disease Center/UC Davis Health-Pathology and Laboratory Medicine

Collaborator: Dr. Samsom Cheung, Electrical and Computer Engineering, Univ. of Kentucky

Overview: Alzheimer's disease pathologies have been reported in both grey matter (GM) and white matter (WM) with different density distributions, making the automated separation task of GM and WM necessary to neuropathologic deep phenotyping. Standard segmentation methods typically involve manual annotations, where a trained researcher traces the boundary of GM and WM in ultra-high resolution whole slide images (WSIs). This method can be time-consuming and subjective, preventing the analysis of large amounts of WSIs in a scalable way. This project aims to develop a robust and label-efficient deep-learning based pipeline to automate GM/WM segmentation and plaque detection. While supervised learning techniques using convolutional neural networks (CNNs) achieve promising results, procuring a sufficiently large dataset with annotations is labor intensive and time consuming. We are exploring semi-supervised learning and active-learning approaches to construct a label-efficient learning framework.

Publications

  • Z. Lai, L. Cerny Oliveira, R. Guo, W. Xu, Z. Hu, K. Mifflin, C. DeCarlie, S-C. Cheung, C-N. Chuah, and B. N. Dugger, "BrainSec: Automated Brain Tissue Segmentation Pipeline for Scalable Neuropathological Analysis," IEEE Access, May 2022. [DOI: 10.1109/ACCESS.2022.3171927]
  • Z. Lai, C. Wang, Z. Hu, B. N. Dugger, S-C. Cheung, C-N. Chuah, "A Semi-supervised Learning for Segmentation of Gigapixel Histopathology Images from Brain Tissues," 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Oct 31-Nov 4, 2021. [DOI: 10.1109/EMBC46164.2021.9629715]
  • Z. Lai, C. Wang, L. Cerny Oliveira, B. Dugger, S-C. Cheung, and C.N. Chuah, "Joint Semi-supervised and Active Learning for Segmentation of Gigapixel Pathology Images with Cost-Effective Labeling," ICCV Workshop on Computational Challenges in Digital Pathology (CDpath), Oct 11 2021. [DOI: 10.1109/ICCVW54120.2021.00072]
  • Z. Lai, K. Guo, W. Xu, Z. Hu, B. Dugger, S. Cheung, and C-N. Chuah, "Automated Grey and White Matter Segmentation in Digitized Ab Human Brain Tissue Slide Images," IEEE ICME 2020 Workshop on Multimedia Services and Technologies for Smart Health (MUST-SH), July 2020. [URL]
Abstracts
  • Z. Lai, L. Cerny Oliveira, D. Harvey, K. Nzenkue, L-W. Jin, C. DeCarli, C-N. Chuah, and B. N. Dugger, "Generalizability of Deep Learning Frameworks for Amyloid Beta Deposit Assessment, Evaluation of Pre-analytic Variables," American Association of Neuropathologists (AANP) Annual Meeting, June 2022. (Lai received R13 Grant Travel Award designated for trainee with the best basic or clinical abstract)
Related Technical Publications
  • Z. Lai, C. Wang, H. Gunawan, S-C. Cheung, and C-N. Chuah, "Smoothed Adaptive Weighting for Imbalanced Semi-Supervised Learning: Improve Reliability Against Unknown Distribution Data," The 39th International Conference on Machine Learning (ICML), July 2022. (Lai received ICML 2022 Participation Grant)
  • Z. Lai, C. Wang, S-C. Cheung, and C-N. Chuah, "SaR: Self-Adaptive Refinement on Pseudo Labels for Multiclass-Imbalanced Semi-Supervised Learning," Computer Vision and Pattern Recognition (CVPR) Workshop on Learning with Limited Labeled Data for Image and Video Understanding (L3DIVU), June 20, 2022.

Funding

    This collaborative work is supported by NSF HDR: TRIPODS grant CCF-1934568, NIH-National Institute On Aging awards #P30AG010129 and #AG062517, and University of California office of the president research grant MRI-19-599956.

AIoT-based Critical Congenital Heart Disease (CCHD) Screening

Lead Domain Expert: Dr. Heather Siefkes, UC Davis Health-Pediatrics

Overview: Critical congenital heart disease (CCHD) is a common group of neonatal life-threatening defects that must be promptly diagnosed to minimize morbidity and mortality. However, despite current screening practices involving oxygen-saturation analysis, timely diagnosis is missed in approximately 900 infants with CCHD annually in the United States (US) and can benefit from increased data processing capabilities. Adding non-invasive perfusion measurements to oxygen-saturation data can improve timeliness and fidelity of CCHD diagnostics. However, real-time monitoring and interpretation of non-invasive perfusion data is currently limited. In this project, we created a hardware and software architecture utilizing a Pi-top for collecting, visualizing, and storing dual oxygen-saturation, perfusion indices, and photoplethysmography data. From these data, we are able to extract non-invasive perfusion features such as perfusion index, radiofemoral delay, and slope of systolic rise or diastolic fall. We use the collected data set to develop proof-of-concept ML algorithm that combines pulse oximetry features to improve detection of CCHD with little impact on false positive rate.

Publications

  • Z. Lai, P. Vadlaputi, D. J. Tancredi, M. Garg, R. I. Koppel, M. Goodman, W. Hogan, N. Cresalia, S. Juergensen, E. Manalo, S. Lashminrusimha, C-N. Chuah, and H. Siefkes, "Enhanced Critical Congenital Cardiac Disease Screening by Combining Interpretable Machine Learning Algorithms," 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Oct 31-Nov 4, 2021.
  • K. Doshi, G. Rehm, P. Vadlaputi, Z. Lai, S. Lakshminrusimha, C-N. Chuah, and H. M Siefkes, "A Novel System to Collect Dual Pulse Oximetry Data for Critical Congenital Heart Disease Screening Research," Journal of Clinical and Translational Science, pp. 1-25, October 2020. [DOI: 10.1017/cts.2020.550]
Abstracts
  • Z. Lai, P. Vadlaputi, D. Tancredi, M. Garg, R. Koppel, M. Goodman, M, W. Hogan, N. Cresalia, S. Juergensen, E. Manalo, S. Lakshminrusimha, C. Chuah, and H. Siefkes, "Machine Learning Algorithm Combining Pulse Oximetry Features for Critical Congenital Heart Disease Screening," Pediatric Academic Society, May 2021.
  • P. Vadlaputi, Z. Lai, M. Garg, R. Koppel, M. Goodman, M, W. Hogan, N. Cresalia, S. Juergensen, E. Manalo, C. Chuah, S. Lakshminrusimha, and H. Siefkes, "Simple Automation Reduces False Positive Rate in Perfusion Index Critical Congenital Heart Disease (CCHD) Screening," Eastern Society for Pediatric Research Annual Meeting, March 2021.
  • H. Siefkes, K. Doshi, G. Rehm, P. Vadlaputi, E. Manalo, W. J. Hogan, M. Garg, R. Koppel, D. Trancedi, S. Lakshminrusimha, C-N. Chuah, "Improving Critical Congenital Heart Disease Screening with Addition of Perfusion Measurements," Translational Science , 2020.

Funding

    This collaborative work is supported by UC Davis AI-for-Health Seed Grant, UCD VEnture Catalyst DIAL Grant, and NIH grant 1R21HD099239-01.