Kuang Gong

Research topics

Matched from this investigator's NIH project titles and abstracts.

• Learning and Memory

Active NIH awards

• Deep Learning Methods for Improving Gallium 68-Based PET Imaging

  5R01EB034692-02

  NIBIB · FY 2025 · $425K

  Abstract Neuroendocrine tumors (NETs) are a heterogeneous group of tumors with increasing incidence, which are hard to identify in the early stage and repeatedly misdiagnosed, yielding 20%-50% of patients with distant metastases at initial diagnosis. Prostate cancer (PCa) is the most common solid-organ malignancy in men in the United States. Distinguishing indolent from aggressive PCa and differentiation of localized disease and metastatic spread are essential for PCa treatment selection. For both NETs and PCa, disease-specific overexpression exists in cancer cells. 68Ga-DOTATATE and 68Ga-PSMA-11, the two most widely used Gallium-68 PET tracers, can target the overexpression in cancer cells of NETs and PCa, respectively. They are essential imaging techniques for NETs and PCa management, given their high sensitivity and specificity in detecting primary tumor and metastatic spread. Due to the shorter half-life and larger positron range of Gallium 68 and the lower injection dose limited by generator capacity, Gallium-68 PET has a lower image quality compared with 18F-FDG PET, which significantly compromises its lesion detectability and quantification accuracy. As Gallium-68 PET is increasingly adopted in clinics, there are unmet needs to further optimize it for better disease management. The goal of this project is to improve Gallium 68-based PET image quality through deep learning (DL)-based motion correction, image reconstruction, and kinetic modeling. Aim 1 of this project is to develop a data-driven respiratory motion-correction framework with phase-matched attenuation correction. Aim 2 of this project is to develop a DL-based image reconstruction method to improve static Gallium 68-based PET imaging. Aim 3 of this project is to further develop DL-based kinetic-modeling methods to improve dynamic Gallium 68-based PET imaging. Aim 4 of this study is to perform comprehensive clinical evaluations of the developed methods. We expect that the integrated outcome of the four aims will be novel, effective, and robust motion correction and reconstruction methods for Gallium-68 PET that can improve its lesion detectability and quantification accuracy.

• Reducing Bias in AI Algorithms for Gallium-68 PET: A Bioethical Perspective Using Transfer Learning

  3R01EB034692-02S2

  NIBIB · FY 2025 · $153K

  Abstract This is a supplemental application under NOT-OD-25-015 to conduct bioethics research and advance capacity building related to bias in AI algorithms. This project is specific to both bioethics research and capacity building in bioethics. Bias in AI algorithms can lead to inaccurate predictions, delaying diagnoses, misguiding treatments, and worsening patient outcomes. Furthermore, biases in AI algorithms can disproportionately affect certain demographic groups, amplifying existing healthcare disparities and leading to inequitable outcomes. How to reduce bias, particularly due to insufficient or unrepresentative datasets, is a major concern in healthcare. Transfer learning offers a promising solution to mitigate data bias when datasets are small or unrepresentative. This is especially relevant for PET imaging, such as Gallium 68 PET scans, for which datasets are limited. With support from the parent R01 project, the team has developed a novel transfer learning framework for PET image quality enhancement. The proposed framework aims to perform pre-training using large-scale high-quality 18F- FDG PET datasets and then fine-tune on limited Gallium 68 datasets. In this supplement project, we will comprehensively evaluate the bias-reduction effect of this proposed framework, explore additional bias mitigation strategies, and investigate potential biases that may rise from transfer learning. Furthermore, we will actively disseminate our findings to research, educational and clinical communities. Bias in AI algorithms is a pressing and emerging bioethical issue when adopting AI to clinics. In the era of rapid AI advancements, this study will provide a robust evidence base to inform and guide future bioethical policies for AI/ML practices, particularly as pre-training and transfer learning become foundational in developing large healthcare AI models. Furthermore, this study will also advance AI-bioethics capacity building by: developing transferable frameworks and methodologies applicable to various biomedical applications, and creating educational resources to address bioethical challenges stemming from biases in AI algorithms.