Work across clinical outcomes, cardiovascular science, medical education, imaging, and emerging tools for care.
Full record on Google ScholarEvaluated whether large language models such as ChatGPT can generate and accurately answer anesthesiology board-style multiple choice questions. Found that LLMs show promise as supplementary study and question-generation tools, while identifying key limitations for clinical education.
* Co-first authors
A structured review of the evidence for virtual reality as a training modality for invasive procedures. Synthesized existing literature to argue that VR simulation can meaningfully improve procedural competence before trainees perform skills on real patients.
* Co-first authors
An interventional feasibility study incorporating a VR module into the standard central venous catheter training curriculum for novice trainees. Demonstrated that VR can be practically integrated into residency education and showed measurable improvements in procedural performance.
Developed and validated a low-cost 3D-printed ultrasound probe simulator and cardiac phantom for teaching transthoracic echocardiography to novice learners. Demonstrated that physical simulation using 3D-printed models can provide accessible, hands-on ultrasound training outside the clinical environment.
A comprehensive review of deep learning methods for diagnosing, grading, and managing brain tumors. Surveyed how convolutional neural networks and other deep learning architectures can interpret MRI and histopathology data to assist neuro-oncology decision-making, and identified key barriers to clinical translation.
An invited editorial commenting on the application of machine learning to cardiac surgery decision-making and outcomes prediction. Explores the promise and limitations of algorithmic tools entering the cardiothoracic surgical workflow.
Designed and evaluated a structured curriculum to improve resident-physician communication skills with patients at an academic medical centre. Addressed a recognized gap in formal communication training within residency programs and measured its impact on patient-reported outcomes.
Explored how communication quality between resident physicians and patients varies across demographic and social groups. Used qualitative methods to capture perspectives from patients, residents, and attending physicians, uncovering systemic disparities in counseling and information-sharing.
Assessed the current state of patient-resident communication through in-depth qualitative interviews, identifying structural challenges and proposing practical solutions for residency training programs and hospital systems.
Examined how the timing of red blood cell transfusions during and after coronary artery bypass surgery relates to 30-day patient outcomes. Identified critical perioperative windows in which transfusion timing may influence morbidity and mortality, with implications for blood management protocols in cardiac surgery.
Investigated whether peripheral nerve blocks — a multimodal pain strategy — improve recovery metrics in cardiac surgery patients. The retrospective analysis assessed opioid consumption, length of stay, and pain scores, contributing evidence for regional anesthesia as part of enhanced recovery after cardiac surgery (ERAS) protocols.
Investigated the 'obesity paradox' in peripheral arterial disease by comparing body surface area against body mass index as predictors of surgical outcomes. Found important distinctions in how these anthropometric measures associate with morbidity and mortality, suggesting BSA may be a more precise risk-stratification tool.
Analyzed risk factors associated with in-hospital mortality among COVID-19 patients hospitalized across five distinct pandemic waves in Pakistan. Identified clinical, demographic, and temporal predictors of death, offering insights into how disease severity and patient profiles evolved over successive waves.
Characterized a rare congenital mitral valve anomaly — the double-orifice mitral valve — using three-dimensional transesophageal echocardiography. Demonstrated how 3D TEE provides superior delineation of valve morphology compared to conventional 2D imaging, enabling more precise surgical and interventional planning.
Proposed and validated a standardized approach using physical 3D markers to reduce spatial orientation errors during intraoperative 3D echocardiography. Addresses a recognized source of image misinterpretation that can lead to procedural errors in structural heart interventions.
Critically examined the limitations of 2D echocardiography in accurately identifying the true aortic annulus, a measurement central to transcatheter aortic valve replacement sizing. Used 3D imaging to demonstrate systematic measurement errors, supporting the adoption of 3D echo as the standard for annular assessment.
Identified CD39/CD73 purinergic signaling as a sex-specific mediator of cardiac fibrosis and inflammation in women undergoing cardiac surgery. Translational study linking molecular pathway differences to clinically observed disparities in postoperative outcomes between male and female patients.
Demonstrated impaired mitochondrial biogenesis signaling via the PGC-1α–pAMPK axis in postmenopausal women during cardiac surgery, and showed that nicotinamide supplementation can partially reverse this deficit in a murine model. Provides mechanistic insight into the heightened cardiac stress response seen in postmenopausal female surgical patients.
* Co-first authorship is marked alongside author names.
View full publication list