About

Lisa Marie Knowlton is an Associate Professor of Surgery (General Surgery) at Stanford University. She is affiliated with the Center for Artificial Intelligence in Medicine & Imaging (AIMI), where her work focuses on integrating artificial intelligence into medical and imaging applications. Her research involves advancing AI methodologies to improve healthcare outcomes, leveraging her expertise in surgery and medical imaging. As a faculty member at Stanford, she contributes to the development of innovative AI-driven solutions in medicine, supporting the center's mission to enhance medical imaging and healthcare through artificial intelligence.

Research topics

• Medicine
• Internal medicine
• Surgery
• Emergency medicine
• Computer Science
• Intensive care medicine
• General surgery
• Family medicine
• Medical education
• Anesthesia
• Pedagogy
• Medical emergency
• Physical therapy
• Demography
• Psychology
• Mathematics education
• Actuarial science

Selected publications

• Empowering surgeons with integrated synthetic data: solutions for mastering complex clinical scenarios

  npj Digital Medicine · 2026-04-30

  articleOpen accessSenior author

  Synthetic data generation across domains can bridge gaps between visual training, skill development, and personalized surgical planning, ultimately transforming how surgeons and artificial intelligence (AI) systems prepare for the complexities of the operating room. In this Perspective, we explore applications of synthetic data to advance surgical education and AI across three key areas: visual data synthesis for training surgeons and AI systems, surgical simulation for skill development and robotics, and digital twins for patient-specific surgical planning and guidance. These domains have largely remained siloed, but their integration has the potential to transform surgical training and AI development across the entire surgical workflow. To fully realize this potential, synthetic data must extend beyond routine surgical events to model atypical anatomy and intraoperative complications-the high-stakes clinical scenarios where enhanced training and AI support are most critical.

  PublisherDOI

• BREAKING DOWN THE EVIDENCE: A MULTICENTER ANALYSIS OF VENOUS THROMBOEMBOLISM AMONG TRAUMA PATIENTS WITH LOWER EXTREMITY FRACTURES

  Journal of Orthopaedic Trauma · 2026-04-08

  article1st authorCorresponding

  OBJECTIVES: To determine whether chemoprophylaxis initiation within 24 hours reduces venous thromboembolism risk among trauma patients with lower extremity long bone fractures. METHODS: Design: This was a retrospective cohort study. SETTING: 17 Level I trauma centers as a part of the Consortium of Leaders in the Study of Traumatic Thromboembolism (CLOTT) study group. PATIENT SELECTION CRITERIA: Patients aged 18-40 years with a diagnosis of lower extremity long bone fracture between January 1, 2018 to December 31, 2020 were included. OUTCOME MEASURES AND COMPARISONS: Primary outcome was VTE during admission. Patients were compared based on VTE chemoprophylaxis initiation within 24 hours (early prophylaxis, E-PROPH) versus not (late or no prophylaxis, L-PROPH) using inverse probability weighted Cox survival analysis. RESULTS: 120 (5.3%) among 2,264 patients with lower extremity fractures developed VTE. 57.5% received E-PROPH and 42.5% received L-PROPH. E-PROPH group had fewer patients with an injury severity score 16 (30.3% vs. 52.4%, p<0.001) and a lower proportion of associated head injury (8.1% vs. 26.7%, p<0.001). VTE incidence was significantly higher in L-PROPH group than in E-PROPH group (8.6% vs. 2.8%, p<0.001). In the adjusted model, E-PROPH was independently associated with nearly a half reduction in VTE incidence (hazard ratio: 0.54, 95% confidence interval: 0.34-0.85). There was no significant difference in the adjusted bleeding complications model (aOR 1.84, 95% CI 0.75, 4.57). CONCLUSION: VTE chemoprophylaxis within the first 24 hours of admission was associated with a marked reduction in VTE incidence among patients with traumatic lower extremity long bone fractures without increase in bleeding risks. LEVEL OF EVIDENCE: Level III: retrospective cohort study.

  PublisherDOI

• Natural Language Processing for Surgical Quality Enhancement

  Journal of Surgical Research · 2026-04-18

  article
  PublisherDOI

• The Future of Emergency Medicaid

  Annals of Surgery · 2025-07-03

  articleSenior author
  PublisherDOI

• Factors Contributing to Long-Term Medicaid Sustainment Among Burn Patients Enrolled via California’s Hospital Presumptive Eligibility Program

  Journal of Burn Care & Research · 2025-06-05

  articleSenior author

  OBJECTIVE: Hospital presumptive eligibility (HPE) provides uninsured patients temporary Medicaid coverage at the time of hospitalization and offers a pathway to securing long-term Medicaid coverage in California. This is of particular importance for burn survivors who have ongoing healthcare needs and may experience financial hardship due to acute and long-term recovery service utilization. METHODS: Using Medicaid claims eligibility data, the California Department of Health Care Services Management Information Systems and the Decisions Support System records, patients ages 18-64 with a primary diagnosis of burn were identified. Descriptive characteristics and Pearson's χ2 tests were used to evaluate bivariate relationships between those who sustained Medicaid after 6 months and those who did not. Multivariate logistic regression was used to determine association of various factors with Medicaid sustainment. RESULTS: Of, 1382 included patients, 73% sustained Medicaid 6 months after HPE enrollment. There were significant differences in Medicaid sustainment between race/ethnicity groups, primary language, TBSA, length of inpatient stay, and need for mechanical ventilation (P = .008) indicating intensive care unit admission. For those who were discharged, multivariate analyses show Spanish speakers and those who did not disclose their language vs English speakers (P = .020 and P < .001 respectively), those who did not disclose race/ethnicity vs white (P = .017), those with <10% TBSA vs 20+% TBSA (P < .001), and those who were discharged home vs those discharged to services (P = .047) were less likely to sustain Medicare. Similar results were observed for all inpatients, except those without concurrent trauma (P = .042) were also less likely to sustain Medicaid. CONCLUSIONS: Hospital presumptive eligibility enrollment at the time of burn injury hospitalization provides a viable path for patients to obtain long-term Medicaid insurance, but additional support pathways must be identified to support Medicaid sustainment for those who are not English speakers and those who are less likely to require long-term follow-up care (ie, those with lower TBSA and those who are discharged home).

  PublisherDOI

• Understaffed and overworked: The stark reality of acute care surgeon staffing in the United States, an Eastern Association for the Surgery of Trauma multicenter study

  The Journal of Trauma: Injury, Infection, and Critical Care · 2025-07-04 · 4 citations

  article

  OBJECTIVES: Rightsizing the workforce to clinical demand requires a balance of work intensity, productivity, and a definition of clinical full-time equivalent (cFTE). We hypothesized a shortage of acute care surgeons based on a 204-shift per year (average, 17 per month) definition of a 1.0 cFTE established in our prior mixed-methods study (two service weeks plus five calls per month). METHODS: This multicenter study used mixed methods, integrating clinical schedules (CY2022), work relative value units, and qualitative insights from semistructured interviews (July 2023 to June 2024). Schedules were converted to shifts (8-14 hours). Hospitals were short-staffed when shift demand exceeded supply based on each surgeon's cFTE. Interviews explored clinical demand and staffing challenges. Descriptive analysis and a deductive-inductive thematic analysis were performed. RESULTS: Forty Level I/II hospitals representing 412 acute care surgeons (287 cFTEs) from 25 states were included. Seventy-nine percent of hospitals were short-staffed. Compared with well-staffed hospitals, short-staffed hospitals had fewer cFTEs (6.5 [interquartile range (IQR), 3] vs. 8.6 [IQR, 3], p < 0.05), a higher demand for clinical work (1,889 [IQR, 933] vs. 1,388 [IQR, 674] shifts, p = 0.05) and a higher work relative value unit/cFTE (8,779 vs. 7,456, p = 0.12). The aggregate clinical demand exceeded available surgeon capacity by 21% overall. Based on volume, a 1.0 cFTE is needed for every 285 (IQR, 169) trauma admissions. There was a deficit of 75 cFTEs across the centers. Key themes identified were related to the value of acute care surgery and balancing unpredictable demand, intensity, and efficiency. CONCLUSION: There appears to be a shortage of acute care surgeons in the United States when a definition of 204 shifts per year cFTE is applied. Hospitals face significant financial and administrative barriers to workforce expansion despite the overabundance of clinical volume. Future research is needed to ascertain the effects of expanding the existing workforce on both clinical outcomes and surgeon well-being. LEVEL OF EVIDENCE: Prognostic and Epidemiologic; Level III.

  PublisherDOI

• Hospital setting of endovascular repair influences procedural outcomes in blunt traumatic aortic injury

  Journal of Vascular Surgery · 2025-06-01

  articleOpen access
  PublisherOA PDFDOI

• Awaiting insurance coverage: Medicaid enrollment and post-acute care use after traumatic injury

  The Journal of Trauma: Injury, Infection, and Critical Care · 2025-01-30 · 3 citations

  article

  BACKGROUND: Lack of insurance after traumatic injury is associated with decreased use of postacute care and poor outcomes. Insurance linkage programs enroll eligible patients in Medicaid at the time of an unplanned admission. We hypothesized that Medicaid enrollment would be associated with increased use of postacute care, but also with prolonged hospital length of stay (LOS) while awaiting insurance authorization. METHODS: We linked trauma registry and EMR data to identify patients ages 18 years to 64 years admitted from 2017 to 2021 to a Level I trauma center. Patients admitted without insurance and retroactively insured (RI) during hospitalization were compared with patients with established Medicaid (MI) and those remaining uninsured (UI). We measured postacute care use including home health care, rehabilitation, and skilled nursing facilities. We tested the association between insurance status and discharge disposition and LOS (primary outcome) using multivariable negative binomial regression. Direct costs were compared between groups. RESULTS: We compared 494 RI patients to 1706 MI and 148 UI patients. Retroactively insured patients had longer hospitalization (median LOS [interquartile range], 4 days [2-9 days]) than other groups (MI, 4 [2-8] and UI 2 [1-3]), p < 0.001). Retroactively insured patients were more likely to be discharged with home health care and to inpatient rehabilitation than UI patients ( p < 0.001). After adjusting for injury and management characteristics, RI was associated with longer LOS compared with MI for patients discharged to inpatient facilities ( p < 0.001). Median costs for RI patients discharged to a facility were $10,284 higher than MI patients, ranging from $8,582 for Injury Severity Score <9 to $51,883 for Injury Severity Score ≥25. CONCLUSION: Enrollment in Medicaid after traumatic injury is associated with postacute care use, but the current enrollment process may delay discharge. Streamlining insurance enrollment and permitting discharge with pending application status could reduce unnecessary hospital days, saving costs and improving improve patient experience. LEVEL OF EVIDENCE: Prognostic and Epidemiological; Level IV.

  PublisherDOI

• Futility Thresholds for Emergency General Surgery in the Post–Cardiac Intensive Care Unit

  Journal of Surgical Research · 2025-07-15 · 1 citations

  article
  PublisherDOI

• Understanding Financial Hardship for Liver Transplant Recipients: A National Evaluation of Pre-Transplant Costs and Insurance Instability

  American Journal of Transplantation · 2025-08-01

  articleSenior author
  PublisherDOI

Frequent coauthors

• David A. Spain

  Stanford University

  114 shared

• Kristan Staudenmayer

  Kuwait Petroleum Corporation (Kuwait)

  67 shared

• Ryan P. Dumas

  The University of Texas Southwestern Medical Center

  55 shared

• Shannon Bichard

  Texas Tech University

  49 shared

• Bayli Davis

  Texas Tech University

  49 shared

• Trent Seltzer

  Texas Tech University Health Sciences Center

  49 shared

• Michael A. Vella

  General University Hospital of Patras

  49 shared

• Barbie Chambers

  The University of Texas Southwestern Medical Center

  49 shared