Mohammed Ali Alvi

I lead the clinical and analytic translation of large-scale real-world oncology data into regulatory-grade evidence. I work at the intersection of medicine, data science, and biostatistics, developing clinical heuristics—including treatment-line definitions and inclusion/exclusion frameworks—to curate trial-emulating data cuts and support external control arm analyses. I collaborate closely with biostatistics and machine-learning teams to apply predictive models to real-world datasets and to generate publication-ready analyses, figures, and survival outputs.

As a Scientist in the Fehlings Lab for Neural Repair and Regeneration, I lead multidisciplinary research programs at the intersection of evidence-based medicine, advanced neuroimaging, machine learning, and clinical outcomes research in traumatic and non-traumatic spinal cord disorders. My work spans guideline development, imaging biomarker discovery, and AI-enabled patient phenotyping, with a central focus on improving prognostication and decision-making in spinal cord injury (SCI) and degenerative cervical myelopathy (DCM).

A major component of my role involves leadership in evidence-based medicine (EBM) and clinical practice guideline development. I served as a contributing member of the AO Spine Clinical Practice Guidelines for Spinal Cord Injury, where I led the systematic evaluation of intraoperative neuromonitoring as a strategy to prevent secondary iatrogenic spinal cord injury during surgery. This work required advanced diagnostic test accuracy (DTA) meta-analyses, rigorous risk-of-bias assessment, and GRADE-based evidence synthesis to translate heterogeneous neuromonitoring data into actionable clinical recommendations. In parallel, I contributed to the North American Spine Society (NASS) Sacroiliac Joint Pain Guideline, supporting protocol development and evidence synthesis to inform diagnostic and interventional care pathways.

In parallel, I lead the lab’s AI-facilitated imaging biomarker program for SCI and DCM. Acting as a data scientist, imaging scientist, and clinical investigator, I oversee the end-to-end processing of spinal MRI data, including image curation, quality control, vertebral labeling, and spinal cord segmentation using semi-automated and deep-learning–enabled pipelines. From these data, I extract high-dimensional morphometric and radiomic features that quantify lesion characteristics, tissue integrity, and compression-related changes. These quantitative imaging biomarkers are then integrated with clinical and functional outcomes to develop prognostic models that aim to move beyond qualitative MRI interpretation toward objective, reproducible biomarkers of recovery potential following surgery.

I also lead machine learning–based clinical phenomics initiatives, applying unsupervised and supervised AI pipelines to identify latent patient subtypes within SCI and DCM populations. This work focuses on clustering patients based on multimodal clinical, imaging, and outcome data to uncover biologically and clinically meaningful phenotypes that may explain heterogeneity in recovery trajectories and treatment response.

In addition to my research leadership, I play an active role in mentorship and training, supervising bioengineering students, medical students, undergraduate researchers, and clinical fellows. Through structured mentorship and summer research programs, I guide trainees in study design, imaging analytics, machine learning workflows, statistical analysis, and scientific communication, contributing to workforce development at the interface of neuroscience, engineering, and clinical research.

At Mayo Clinic, I held dual roles as a Research Scientist and Lead Clinical Fellow/Analyst, leading large-scale registry-based outcomes research across neurosurgery, oncology, and spine care. I served as the lead analyst and scientific consultant for multiple national registries within the NeuroPoint Alliance, including the Stereotactic Radiosurgery (SRS) Registry, where I led analyses identifying prognostic radiosurgical parameters associated with local control, intracranial progression, and patient-reported quality-of-life outcomes in patients with brain metastases. This work resulted in inaugural and subsequent peer-reviewed publications in high-impact neurosurgical journals.

In parallel, I played a central role in the development and launch of the Quality Outcomes Database (QOD) Tumor Registry, the first national surgical brain tumor registry, overseeing registry design, variable definition, data quality audits, analytic strategy, and trainee supervision. I also coordinated multicenter data collection and analysis for spine-focused QOD initiatives, including studies on cervical spondylotic myelopathy and lumbar spondylolisthesis, culminating in over 20 publications examining comparative effectiveness, surgical outcomes, and healthcare disparities.

Beyond registry science, my work at Mayo Clinic spanned advanced outcomes research, bioinformatics, and translational neuroscience. I led and contributed to projects involving systematic reviews and meta-analyses, next-generation sequencing–based biomarker discovery using the Mayo Clinic Biobank, healthcare analytics using claims data, and development of predictive models for surgical risk and outcomes. I also authored regulatory documents, grant applications, and IND/IDE-related materials, secured competitive intramural and extramural funding, mentored medical students and research fellows, and contributed to patented medical device innovation.