Ramesh Jain
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Ramesh Jain has a total of 171 co-authors. Here are the top co-authors by number of shared publications:
Laleh Jalali
University of California, Irvine (UCI), Hitachi America Ltd
11 shared publications- Amir M. Rahmani
University of California, Irvine, Irvine, California, USA
9 shared publications - Jessica L. Borelli
Department of Psychology University of California, Irvine Irvine California USA
7 shared publications - Jocelyn Lai
Washington University in St. Louis
7 shared publications - Nikil Dutt
University of California at Irvine
7 shared publications - Sina Labbaf
University of California, Irvine, USA
7 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 34-42
PAMU: Personalized Affective Music Generation Using U-Transformer with EEG-driven Multimodal Control
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 34-42
Open AccessPAMU: Personalized Affective Music Generation Using U-Transformer with EEG-driven Multimodal Control
- Open Access
- 31 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-5
Restoring the Human: Purposive Metadata for Trust in Generative AI
- 31 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-5
Open AccessRestoring the Human: Purposive Metadata for Trust in Generative AI
- Open Access
- 21 July 2026
- Published by MDPI AG in Preprints.org
AI-Enabled Digital Twins in Healthcare: Epistemic Foundations, Principles, Progress, and Future Directions
Digital twins have rapidly entered healthcare discourse, accompanied by ambitious claims regarding personalization, prediction, and decision support. Originating in engineering domains characterized by well-observed, controllable, and mechanistically understood systems, the concept is increasingly being applied to biological, clinical, and organizational settings whose underlying assumptions differ substantially. As a result, the term now encompasses a heterogeneous collection of models, simulations, and artificial intelligence (AI) systems, often without clear specification of what is being represented, what can be inferred, or what actions can legitimately be justified.This review develops a scale-aware framework for evaluating healthcare digital twins. Through a structured analysis spanning cardiovascular disease, diabetes, Alzheimer's disease, maternal–fetal health, addiction, radiology, and hospital operations, we argue that the scientific validity and practical utility of a digital twin depend fundamentally on scale. As applications move from mechanistic physiological systems to individual patients, coupled human systems, sociotechnical settings, and healthcare organizations, the central challenges shift from model fidelity and state estimation to non-identifiability, behavioral adaptation, contested objectives, and governance.Across domains, recurring failure modes arise less from insufficient computational capability than from misalignment among state definitions, uncertainty representation, causal support, decision authority, and governance. We examine the roles of AI in state estimation, prediction, uncertainty quantification, and decision support while highlighting its limits in addressing poorly defined state spaces, non-identifiability, causal ambiguity, and normative questions regarding acceptable actions. We further argue that governance is not an external constraint but a foundational component of healthcare digital twins, shaping accountability, lifecycle oversight, data stewardship, and the legitimacy of decisions informed by digital twin outputs.This review reframes healthcare digital twins not as a single technology but as a family of computational representations whose validity depends on scale, context, purpose, and governance. Future progress will depend less on increasingly sophisticated models than on maintaining alignment among representation, evidence, intervention, and accountability, and on ensuring digital twin ambitions remain commensurate with the epistemic conditions and institutional constraints of the systems they seek to represent. Building on this analysis, the review offers complementary research roadmaps for AI and governance that emphasize explicit state semantics, uncertainty-aware inference, decision-centered evaluation, adaptive oversight, and governance-by-design.
- 21 July 2026
- Published by MDPI AG in Preprints.org
Open AccessAI-Enabled Digital Twins in Healthcare: Epistemic Foundations, Principles, Progress, and Future Directions
Digital twins have rapidly entered healthcare discourse, accompanied by ambitious claims regarding personalization, prediction, and decision support. Originating in engineering domains characterized by well-observed, controllable, and mechanistically understood systems, the concept is increasingly being applied to biological, clinical, and organizational settings whose underlying assumptions differ substantially. As a result, the term now encompasses a heterogeneous collection of models, simulations, and artificial intelligence (AI) systems, often without clear specification of what is being represented, what can be inferred, or what actions can legitimately be justified.This review develops a scale-aware framework for evaluating healthcare digital twins. Through a structured analysis spanning cardiovascular disease, diabetes, Alzheimer's disease, maternal–fetal health, addiction, radiology, and hospital operations, we argue that the scientific validity and practical utility of a digital twin depend fundamentally on scale. As applications move from mechanistic physiological systems to individual patients, coupled human systems, sociotechnical settings, and healthcare organizations, the central challenges shift from model fidelity and state estimation to non-identifiability, behavioral adaptation, contested objectives, and governance.Across domains, recurring failure modes arise less from insufficient computational capability than from misalignment among state definitions, uncertainty representation, causal support, decision authority, and governance. We examine the roles of AI in state estimation, prediction, uncertainty quantification, and decision support while highlighting its limits in addressing poorly defined state spaces, non-identifiability, causal ambiguity, and normative questions regarding acceptable actions. We further argue that governance is not an external constraint but a foundational component of healthcare digital twins, shaping accountability, lifecycle oversight, data stewardship, and the legitimacy of decisions informed by digital twin outputs.This review reframes healthcare digital twins not as a single technology but as a family of computational representations whose validity depends on scale, context, purpose, and governance. Future progress will depend less on increasingly sophisticated models than on maintaining alignment among representation, evidence, intervention, and accountability, and on ensuring digital twin ambitions remain commensurate with the epistemic conditions and institutional constraints of the systems they seek to represent. Building on this analysis, the review offers complementary research roadmaps for AI and governance that emphasize explicit state semantics, uncertainty-aware inference, decision-centered evaluation, adaptive oversight, and governance-by-design.
- 7 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 33 (1) , 4-6
The Dawn of Multimodal Assistive Intelligence
Intelligence is the ability to solve problems in the context of specific individual needs and capabilities. This reveals why current AI falls short: language processing, however sophisticated, cannot capture multimodal reality or enable personalized action. We’re at a convergence moment where large language models can merge with deep multimodal sensing to create multimodal assistive intelligence (MAI). MAI integrates information across modalities, combines general knowledge with personal context, and enables appropriate action in dynamic situations. The technology components exist. This column introduces MAI and predicts that by 2030, all effective AI will be multimodal, personal, and assistive.
- 7 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 33 (1) , 4-6
The Dawn of Multimodal Assistive Intelligence
Intelligence is the ability to solve problems in the context of specific individual needs and capabilities. This reveals why current AI falls short: language processing, however sophisticated, cannot capture multimodal reality or enable personalized action. We’re at a convergence moment where large language models can merge with deep multimodal sensing to create multimodal assistive intelligence (MAI). MAI integrates information across modalities, combines general knowledge with personal context, and enables appropriate action in dynamic situations. The technology components exist. This column introduces MAI and predicts that by 2030, all effective AI will be multimodal, personal, and assistive.
- 1 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 33 (2) , 4-7
Beyond the Query Box
- 1 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 33 (2) , 4-7
Beyond the Query Box
- 3 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 32 (4) , 8-12
A Conversation Between the Founding and the Current EICs on 30 Years of Multimedia Computing
- 3 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 32 (4) , 8-12
A Conversation Between the Founding and the Current EICs on 30 Years of Multimedia Computing
- Open Access
- 26 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 17-26
Extending FKG.in: Towards A Food Claim Traceability Network
- 26 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 17-26
Open AccessExtending FKG.in: Towards A Food Claim Traceability Network
- Open Access
- 26 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 96-98
Panel - Multimodal Food Computing: Bridging Cultural Heritage, Artificial Intelligence, and Sustainable Food Futures
- 26 October 2025
- Published by Association for Computing Machinery (ACM)
- p. 96-98
Open AccessPanel - Multimodal Food Computing: Bridging Cultural Heritage, Artificial Intelligence, and Sustainable Food Futures
- 1 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 32 (4) , 4-7
Thirty Years of Multimedia Computing: From Academic Vision to Ubiquitous Intelligence
- 1 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE MultiMedia
- Vol. 32 (4) , 4-7
Thirty Years of Multimedia Computing: From Academic Vision to Ubiquitous Intelligence
- 27 May 2025
- Published by Springer Nature
- p. 76-90
Enhancing FKG.in: Automating Indian Food Composition Analysis
- 27 May 2025
- Published by Springer Nature
- p. 76-90
Enhancing FKG.in: Automating Indian Food Composition Analysis