About

Ivo D. Dinov is a Professor at the University of Michigan School of Nursing, holding the title of Henry Philip Tappan Collegiate Professor and serving as Chair of the Department of Systems, Populations and Leadership. He is also affiliated with the Department of Health Behavior and Clinical Sciences and the Department of Systems, Populations and Leadership, and is the Associate Director for Education and Training at the Michigan Institute for Data Science. Dr. Dinov's research interests include spacekime and predictive healthcare analytics, foundational and generative artificial intelligence modeling, biomedical data science, computational neuroscience, informatics, augmented intelligence, mathematical-physics modeling, and statistical inference. He is the Director of the Statistics Online Computational Resource (SOCR) and is an expert in mathematical modeling, statistical analysis, high-throughput computational processing, and scientific visualization of large datasets (Big Data). His applied research focuses on neuroscience, nursing informatics, multimodal biomedical image analysis, and distributed genomics computing, with specific projects involving longitudinal morphometric studies of development, maturation, and aging, as well as exploring the relationships between genetic traits, clinical phenotypes, and demographics in brain and heart disorders. Dr. Dinov is also developing and disseminating innovative pedagogical approaches for science education and active learning.

Research topics

• Computer science
• Artificial intelligence
• Medicine
• Data science
• Psychology

Selected publications

• Intricacies of human–AI interaction in dynamic decision-making for precision oncology

  Nature Communications · 2025-01-29 · 29 citations

  articleOpen access

  AI decision support systems can assist clinicians in planning adaptive treatment strategies that can dynamically react to individuals' cancer progression for effective personalized care. However, AI's imperfections can lead to suboptimal therapeutics if clinicians over or under rely on AI. To investigate such collaborative decision-making process, we conducted a Human-AI interaction study on response-adaptive radiotherapy for non-small cell lung cancer and hepatocellular carcinoma. We investigated two levels of collaborative behavior: model-agnostic and model-specific; and found that Human-AI interaction is multifactorial and depends on the complex interrelationship between prior knowledge and preferences, patient's state, disease site, treatment modality, model transparency, and AI's learned behavior and biases. In summary, some clinicians may disregard AI recommendations due to skepticism; others will critically analyze AI recommendations on a case-by-case basis; clinicians will adjust their decisions if they find AI recommendations beneficial to patients; and clinician will disregard AI recommendations if deemed harmful or suboptimal and seek alternatives.

  PublisherOA PDFDOI

• Foreground-aware Virtual Staining for Accurate 3D Cell Morphological Profiling.

  PubMed · 2025-07-07

  preprintOpen access

  Microscopy enables direct observation of cellular morphology in 3D, with transmitted-light methods offering low-cost, minimally invasive imaging and fluorescence microscopy providing specificity and contrast. Virtual staining combines these strengths by using machine learning to predict fluorescence images from label-free inputs. However, training of existing methods typically relies on loss functions that treat all pixels equally, thus reproducing background noise and artifacts instead of focusing on biologically meaningful signals. We introduce Spotlight, a simple yet powerful virtual staining approach that guides the model to focus on relevant cellular structures. Spotlight uses histogram-based foreground estimation to mask pixel-wise loss and to calculate a Dice loss on soft-thresholded predictions for shape-aware learning. Applied to a 3D benchmark dataset, Spotlight improves morphological representation while preserving pixel-level accuracy, resulting in virtual stains better suited for downstream tasks such as segmentation and profiling.

  PublisherOA PDF

• Adaptive Coping and Caregiving Self-Efficacy Among Black Family Caregivers of Persons with Dementia

  Innovation in Aging · 2025-12-01

  articleOpen access

  Abstract Black family caregivers of persons with dementia have an increased risk of poor health, future cognitive impairment, and premature death related to stressors associated with caregiving. Understanding the relationships between adaptive coping, caregiving self-efficacy, and health can facilitate the development of resilience-focused interventions that may improve caregiver health. We surveyed 102 self-identified Black family caregivers of persons with dementia. Survey measures included the coping and adaptation processing scale (CAPS), the revised scale for caregiving self-efficacy, and the PROMIS global health scale. Analysis was completed using regression modeling. Adjusting for age, gender, marital status, employment status, income, education, and presence of chronic disease, we found that adaptive coping was positively associated with caregiver self-efficacy in managing disruptive behaviors (B = 1.57, 95% CI [.96, 2.173], p = 0.000) and controlling upsetting thoughts (B = 1.67, 95% CI [1.13, 2.22], p = 0.000). Adaptive coping was not significantly associated with mental health (i.e., emotional distress, social role satisfaction) (B=.07, 95% CI [-.28, 0.41], p = 0.707), or physical health (i.e., physical function, pain, fatigue) (B=-0.01, 95%CI [-.27, 0.24], p = 0.926). Adaptive coping strategies associated with caregiving self-efficacy and resilience theory included experience, use of planning skills, and humor. Designing caregiver support interventions that facilitate the development of effective planning skills may significantly improve caregiver mental health.

  PublisherOA PDFDOI

• Effects of β-catenin deficiency on adipose tissue physiology

  Molecular Metabolism · 2025-08-05 · 1 citations

  articleOpen access

  OBJECTIVES: Compelling evidence from investigation of preclinical models and humans links canonical Wnt/β-catenin signaling to regulation of many aspects of white adipose tissue development and physiology. Dysregulation of this ancient pathway alters adiposity and metabolic homeostasis. Herein we explore how disruption of adipocyte Wnt/β-catenin signaling affects gene expression and crosstalk between cell types within adipose tissue. METHODS: To investigate mechanisms through which adipose tissue attempts to maintain homeostasis in the absence of β-catenin in adipocytes, we employed standard methods of metabolic phenotyping as well as bulk RNA sequencing, flow cytometry, single-cell RNA sequencing, and isolation of secreted extracellular vesicles. RESULTS: mice is maintained in part by elevated expression of Ctnnb1 mRNA in endothelial cells and in secreted small extracellular vesicles. CONCLUSIONS: Our studies demonstrate the importance of adipocyte Wnt signaling for regulation of lipid and mitochondrial metabolic processes in stromal-vascular cells and adipocytes in adipose tissues. This research provides further support for an intercellular Wnt signaling network with compensatory capability to maintain homeostasis, and underscores importance of Wnt/β-catenin signaling for understanding adipose tissue physiology and pathophysiology.

  PublisherDOI

• Mouse Bio-behavioral Phenotyping Using a Digital Homecage Framework for Long-timescale, High-resolution, and Multi-factor Data Collection and Analytics

  bioRxiv (Cold Spring Harbor Laboratory) · 2025-04-29

  preprintOpen access
  PublisherDOI

• Medical data sharing and synthetic clinical data generation – maximizing biomedical resource utilization and minimizing participant re-identification risks

  npj Digital Medicine · 2025-08-16 · 3 citations

  articleOpen access

  The sensitive nature of electronic health records (EHR) and wearable data presents challenges in sharing biomedical resources while minimizing re-identification risks. This article introduces an end-to-end, titratable pipeline that generates privacy-preserving "digital twin" datasets from complex EHR and wearable-device records (Apple Watch data from 3029 participants) using DataSifter and Synthetic Data Vault (SDV) methods. Various obfuscation levels were applied (DataSifter: small, medium, large; SDV: CTGAN, Gaussian Copula) and benchmarked using utility (statistical fidelity, machine learning performance) and privacy (re-identification risk, detection likelihood) metrics. The highest-obfuscation DataSifter twin delivered the strongest privacy protection (0.83) while preserving key statistical and predictive signals (83.1% confidence interval overlap in regression models), outperforming SDV, particularly for longitudinal data. Despite declining performance in machine learning tasks with higher obfuscation, utility was generally preserved. The study underscores the importance of digital twin datasets and highlights DataSifter's adaptability in privacy-utility trade-offs, advocating its utility for secure data sharing.

  PublisherOA PDFDOI

• Sex Difference in the Associations of Socioeconomic Status, Cognitive Function, and Brain Volume with Dementia in Old Adults: Findings from the OASIS Study

  Journal of dementia and Alzheimer's disease · 2025-04-03

  articleOpen access

  Background: Sex differences in the association of cognitive function and imaging measures with dementia have not been fully investigated. Understanding sex differences in the dementia-related socioeconomic, cognitive, and imaging measurements is crucial for uncovering sex-related pathways to dementia and facilitating early diagnosis, family planning, and cost control. Methods: We selected data from the Open Access Series of Imaging Studies, with longitudinal measurements of brain volumes, on 150 individuals aged 60 to 96 years. Dementia status was determined using the Clinical Dementia Rating (CDR) scale, and Alzheimer’s disease was diagnosed as a CDR of ≥0.5. Generalized estimating equation models were used to estimate the associations of socioeconomic, cognitive, and imaging factors with dementia in men and women. Results: The study sample consisted of 88 women (58.7%) and 62 men (41.3%), and the average age of the subjects was 75.4 years at the initial visit. A lower socioeconomic status was associated with a reduced estimated total intracranial volume in men, but not in women. Ageing and lower MMSE scores were associated with a reduced nWBV in both men and women. Lower education affected dementia more in women than in men. Age, education, Mini-Mental State Examination (MMSE), and normalized whole-brain volume (nWBV) were associated with dementia in women, while only MMSE and nWBV were associated with dementia in men. Conclusions: The association between education and the prevalence of dementia differs in men and women. Women may have more risk factors for dementia than men.

  PublisherOA PDFDOI

• Back to the future – from nuclear energy to AI

  AI Perspectives & Advances · 2025-05-26 · 2 citations

  reviewOpen access1st authorCorresponding

  There are striking parallels between the historic realization that the atomic forces may be harnessed to generate nuclear energy and the recent explosion of research activities and wide proliferation of artificial intelligence (AI). This perspective piece compares and contrasts developments of controlled nuclear fission (CNF) and AI. It also delineates some of the homologies between nuclear energy and AI in terms of the underlying drivers of technological progress, social, economic and environmental challenges, and unimaginable opportunities to transform the course of human civilization. It appears that the course of contemporary AI parallels the trajectory of classical energy generation by nuclear fission; disruptive and unsettling. However, the future prospects of a long-term sustainable NextGen AI may more closely resemble the quest for clean, renewable, and limitless nuclear fusion energy; always pseudorealistic and never decisively resolved.

  PublisherOA PDFDOI

• Prospective Prediction of Adult Substance Use Disorder Outcomes in a High Risk Sample of Youth Using Machine Learning and Predictors From Multiple Domains

  Drug and Alcohol Dependence · 2025-02-01

  article
  PublisherDOI

• Intricacies of Human-AI Interaction in Dynamic Decision-Making for Precision Oncology: A Case Study in Response-Adaptive Radiotherapy

  medRxiv · 2024-04-30 · 2 citations

  preprintOpen access

  Abstract Background Adaptive treatment strategies that can dynamically react to individual cancer progression can provide effective personalized care. Longitudinal multi-omics information, paired with an artificially intelligent clinical decision support system (AI-CDSS) can assist clinicians in determining optimal therapeutic options and treatment adaptations. However, AI-CDSS is not perfectly accurate, as such, clinicians’ over/under reliance on AI may lead to unintended consequences, ultimately failing to develop optimal strategies. To investigate such collaborative decision-making process, we conducted a Human-AI interaction case study on response-adaptive radiotherapy (RT). Methods We designed and conducted a two-phase study for two disease sites and two treatment modalities—adaptive RT for non-small cell lung cancer (NSCLC) and adaptive stereotactic body RT for hepatocellular carcinoma (HCC)—in which clinicians were asked to consider mid-treatment modification of the dose per fraction for a number of retrospective cancer patients without AI-support (Unassisted Phase) and with AI-assistance (AI-assisted Phase). The AI-CDSS graphically presented trade-offs in tumor control and the likelihood of toxicity to organs at risk, provided an optimal recommendation, and associated model uncertainties. In addition, we asked for clinicians’ decision confidence level and trust level in individual AI recommendations and encouraged them to provide written remarks. We enrolled 13 evaluators (radiation oncology physicians and residents) from two medical institutions located in two different states, out of which, 4 evaluators volunteered in both NSCLC and HCC studies, resulting in a total of 17 completed evaluations (9 NSCLC, and 8 HCC). To limit the evaluation time to under an hour, we selected 8 treated patients for NSCLC and 9 for HCC, resulting in a total of 144 sets of evaluations (72 from NSCLC and 72 from HCC). Evaluation for each patient consisted of 8 required inputs and 2 optional remarks, resulting in up to a total of 1440 data points. Results AI-assistance did not homogeneously influence all experts and clinical decisions. From NSCLC cohort, 41 (57%) decisions and from HCC cohort, 34 (47%) decisions were adjusted after AI assistance. Two evaluations (12%) from the NSCLC cohort had zero decision adjustments, while the remaining 15 (88%) evaluations resulted in at least two decision adjustments. Decision adjustment level positively correlated with dissimilarity in decision-making with AI [NSCLC: ρ = 0.53 ( p < 0.001); HCC: ρ = 0.60 ( p < 0.001)] indicating that evaluators adjusted their decision closer towards AI recommendation. Agreement with AI-recommendation positively correlated with AI Trust Level [NSCLC: ρ = 0.59 ( p < 0.001); HCC: ρ = 0.7 ( p < 0.001)] indicating that evaluators followed AI’s recommendation if they agreed with that recommendation. The correlation between decision confidence changes and decision adjustment level showed an opposite trend [NSCLC: ρ = −0.24 ( p = 0.045), HCC: ρ = 0.28 ( p = 0.017)] reflecting the difference in behavior due to underlying differences in disease type and treatment modality. Decision confidence positively correlated with the closeness of decisions to the standard of care (NSCLC: 2 Gy/fx; HCC: 10 Gy/fx) indicating that evaluators were generally more confident in prescribing dose fractionations more similar to those used in standard clinical practice. Inter-evaluator agreement increased with AI-assistance indicating that AI-assistance can decrease inter-physician variability. The majority of decisions were adjusted to achieve higher tumor control in NSCLC and lower normal tissue complications in HCC. Analysis of evaluators’ remarks indicated concerns for organs at risk and RT outcome estimates as important decision-making factors. Conclusions Human-AI interaction depends on the complex interrelationship between expert’s prior knowledge and preferences, patient’s state, disease site, treatment modality, model transparency, and AI’s learned behavior and biases. The collaborative decision-making process can be summarized as follows: (i) some clinicians may not believe in an AI system, completely disregarding its recommendation, (ii) some clinicians may believe in the AI system but will critically analyze its recommendations on a case-by-case basis; (iii) when a clinician finds that the AI recommendation indicates the possibility for better outcomes they will adjust their decisions accordingly; and (iv) When a clinician finds that the AI recommendation indicate a worse possible outcome they will disregard it and seek their own alternative approach.

  PublisherOA PDFDOI

Recent grants

• Project I, Circuit Mechanisms of Attentional-Motor Interface Dysfunction in PD Falls

  NIH · $20.1M · 2020

• Collaborative Research: Distributome--An Interactive Web-based Resource for Probability Distributions

  NSF · $88k · 2010–2014

• Statistics Online Computational Resource for Education

  NSF · $473k · 2007–2011

• Collaborative Research: Distributome--An Interactive Web-based Resource for Probability Distributions

  NSF · $61k · 2013–2014

• Biomedical Informatics and Data Science Training Program (BIDS-TP)

  NIH · $2.1M · 2021–2026

Frequent coauthors

• Arthur W. Toga

  144 shared

• Paul M. Thompson

  University of Southern California

  54 shared

• Pau Medrano−Gracia

  University of Auckland

  36 shared

• João A.C. Lima

  Johns Hopkins Medicine

  36 shared

• Alistair A. Young

  King's College London

  36 shared

• Michael Backhaus

  University of Auckland

  36 shared

• Wenchao Tao

  University of California, Los Angeles

  36 shared

• Brett R. Cowan

  Institute of Environmental Science and Research

  36 shared

Education

• Postdoc, Neuroscience

  University of California Los Angeles

  2000

• MS, Statistics

  Florida State University

  1998

• PhD, Mathematics

  Florida State University

  1998

• MS, Mathematics

  Michigan Technological University

  1993

Awards & honors

• World Wide Web Awards™ "Gold" Award, July 2007
• IEEE Mathematical Methods in Biomedical Image Analysis (MMBI…
• Runner up, ASA Hands-On Statistics Activity Competition, 201…