Kuofu Liu
Here is my CV.
Research Experience
- Discharge Decision-making in Cardiac Rehabilitation (CR) System with Heterogeneous Patient Adherence Pattern
Supervised by Prof. Mariel Lavieri | UMich- Collaborated with Michigan Medicine clinicians to understand current clinical workflows, discharge practices, and patient flow constraints in CR settings.
- Conducted literature review on cardiac rehabilitation operations, adherence prediction, and patient flow optimization problems to understand problem challenges and research gaps.
- Defined key research objectives, including modeling patient adherence, predicting dropouts, and optimizing discharge timing to optimize patient admission and discharge decisions in CR centers under uncertainty.
- Collaborated with Michigan Medicine clinicians to understand current clinical workflows, discharge practices, and patient flow constraints in CR settings.
- Center-Level Variation in the Use of HCV+ Livers: Implications for Waitlist and Post-Transplant Outcomes
Supervised by Prof. Mariel Lavieri | UMich- Investigated the impact of listing for hepatitis C virus-positive (HCV+) liver organs on transplant access and outcomes using Kaplan-Meier survival curves, Cox proportional hazard models, and Fine-Gray competing risks models based on national UNOS registry data from 2015–2023.
- Identified demographic and clinical differences between candidates listed for HCV+ vs. HCV– organs (e.g., liver disease etiology, insurance status, and ethnicity).
- Analyzed center-level variability in HCV+ organ use and HCV+ organ listing, accounting for transplant volume and deceased cardiac donor (DCD) utilization rates.
- Investigated the impact of listing for hepatitis C virus-positive (HCV+) liver organs on transplant access and outcomes using Kaplan-Meier survival curves, Cox proportional hazard models, and Fine-Gray competing risks models based on national UNOS registry data from 2015–2023.
- Synchronizing the Treatment of Multiple Chronic Conditions Based on Maximum Safe Treatment Intervals
Supervised by Prof. Mariel Lavieri | UMich- Conducted a literature review on the synchronization of treatment scheduling, treat-and-extend policy, and maximum safe treatment interval (MSTI) for chronic conditions.
- Conducted 10-year simulations under three treatment policies (π₁, π₂, πᵢ) to quantify overtreatment risk relative to clinic visit reductions in chronic conditions treatment.
- Defined and implemented the Extra Injections per Saved Visit (EISV) index to identify MSTI patterns where synchronization policies yield minimal overtreatment and maximal visit reductions.
- Provided clinical decision-making suggestions based on simulated policy performance across 45 MSTI combinations.
- Conducted a literature review on the synchronization of treatment scheduling, treat-and-extend policy, and maximum safe treatment interval (MSTI) for chronic conditions.
- Parameters Estimation and Global Sensitivity Analysis of Time-Varying SEIRD Compartmental Model Based on State-level Covid-19 Data in the U.S.
Supervised by Prof. Randolph Hall | USC- Fitted 8 parameters (4 shape parameters of 2 sigmoid functions) for each state using Covid-19 case and death data to an extended SEIRD model in which transmission rate and fatality rate changing chronologically, with the model reaching RRMSEs of 1.54% for cases and 1.20% for deaths.
- Conducted sensitivity analysis with a self-developed Monte Carlo simulation algorithm and investigated scenarios of 8 parameters following uniform, normal, lognormal, gamma, and truncated normal distribution, inspecting 245 days in California and New York State starting from March 13, 2020.
- Established country-level sensitivity analysis with 410 parameters for the SERID model considering the transportation effect between states on COVID-19 transmission among all 50 states.
- Fitted 8 parameters (4 shape parameters of 2 sigmoid functions) for each state using Covid-19 case and death data to an extended SEIRD model in which transmission rate and fatality rate changing chronologically, with the model reaching RRMSEs of 1.54% for cases and 1.20% for deaths.