India AI Impact Summit Insights
All sessions

CATCH Grant Awards 2026: Recognizing Innovation in AI Cancer Care

Contents

Executive Summary

This summit event announced the CATCH (Cancer care with AI Technology and Care Help) grant awards—a major government initiative recognizing 10 winning AI solutions for cancer care in India. The event showcased a government-industry-academic partnership framework designed to move AI solutions from prototype to scaled deployment, emphasizing that AI adoption in oncology requires governance, validation, end-user focus, and multi-sector collaboration rather than technology alone.

Key Takeaways

  1. AI in cancer care works best when framed as a workforce multiplier and clinical decision-support tool, not a replacement—addressing India's specialist shortage while augmenting expertise, not substituting it.

  2. Validation and governance infrastructure is as important as the technology itself. The ICMR-led consortium partnership for standardized evaluation is a critical enabler of scaled deployment and vendor confidence.

  3. India is establishing a template for responsible, scaled AI in healthcare that can be adapted globally, combining government coordination, multi-sector partnerships, problem-first design, and ethical rigor—a model for the global south.

  4. Collaboration over competition (NCG model) creates the ecosystem for innovation. Hospitals and institutions must shift from individual competitiveness to shared goals (e.g., reducing mortality) to enable systemic adoption of AI tools.

  5. Success means access to quality healthcare on a smartphone—AI solutions should democratize expert-level diagnosis and treatment planning across geographies and income levels, not just optimize elite centers.

Key Topics Covered

  • Cancer burden in India: Rising incidence (16+ lakh cases annually) and mortality patterns; expansion of cancer care infrastructure from tertiary centers to district hospitals
  • AI applications in cancer care: Diagnostic radiology, radiotherapy planning, pathology, treatment adherence, and screening
  • AI adoption challenges: Shortage of specialists; need for validation, regulatory approval, and market navigation
  • Government policy and infrastructure: Ministry of Health role; state vs. national governance; creation of 39 state cancer institutes
  • The CATCH grant: Challenge structure (299 applicants → 35 shortlisted → 10 award winners); thematic areas (screening, diagnostics, data curation, R&D, treatment adherence)
  • Evaluation and validation framework: New partnership announced for standardized assessment of AI solutions before deployment
  • Template for global south: India positioning itself as a model for scaled, responsible AI implementation in healthcare
  • Human-centered AI philosophy: AI as a tool augmenting clinicians, not replacing them; user and problem-first approaches
  • Pitfalls and risks: Verticalization (redundant competing solutions), validation shortcuts, lack of localization, dataset bias

Key Points & Insights

  1. Scale crisis requires AI: India faces a 1.4–2.6× multiplier gap in required healthcare workforce; conventional hiring cannot meet cancer care demand. AI is positioned as a force multiplier, not a replacement technology.

  2. Validation is the gating factor: Solutions exist, but vendor confidence and system adoption depend on rigorous, independent evaluation before deployment. ICMR-led consortium partnership announced to provide this capability.

  3. Governance precedes technology: Effective scaling requires multi-sector coordination (government health ministry, research bodies, private hospitals, academic institutions, regulatory bodies). India AI mission acts as convener.

  4. National Cancer Grid (NCG) model of collaboration: 400+ cancer organizations working toward shared goals (narrowing mortality-to-incidence ratio) rather than institutional competition. This creates the ecosystem for AI adoption.

  5. Problem-first, not solution-first: Panelists emphasized identifying healthcare pain points first, then selecting appropriate tools (AI or otherwise). Vertical duplication of similar solutions is a key risk.

  6. Ethical and responsible AI is non-negotiable: Dataset bias (e.g., pulse oximetry failing on darker skin tones) demonstrates that AI solutions must be validated across diverse populations and settings—especially critical for global south applicability.

  7. Template and localization matter: Copy-paste of solutions fails; successful scaling requires governance templates that are then formatted to local contexts and user needs.

  8. End-user and outcome focus: Moving beyond proof-of-concept and proof-of-implementation to proof-of-outcome (real clinical and mortality impact) requires keeping patients and clinicians at the center of design and deployment decisions.

  9. Timeline compression in evaluation: Traditional RCTs are too slow; accelerated but rigorous evaluation pathways needed to maintain safety while enabling faster iteration and feedback.

  10. Systematic institutionalization: AI adoption must be embedded in institutional practice standards and professional body guidelines (e.g., NCG, ASCO) to ensure consistency and prevent siloing.

Notable Quotes or Statements

  • Ministry of Health speaker: "AI at this juncture is not going to replace radiologists or radiotherapists. It's only going to become a tool in their hands for doing more scans and doing accurate scans and not missing out the tumors."

  • Electronics Ministry Secretary: "Humans in the loop or humans at the center of this process is what at least India is attempting to advocate as the key element of what we want to do."

  • Dr. Mona (ICMR): "Before we deploy any solutions, the validation has to be done in such a manner that a vendor feels comfortable taking it on... until you have the right kind of people doing the validation, whatever you might do won't work."

  • Samir Pachi (WHO): "The users decide [what scales], not the developers. You have to make the technology so capable that it can become that value for the users... governance is key... copy, paste, and format for local settings."

  • Dr. Promesh (Tata Memorial, NCG): "We have problems in healthcare and cancer care—let's find whatever solution it might be. AI might be one of those tools... The only way we can [address pathology shortages in the global south] is to be disruptive... but do so in an ethical and responsible way."

  • Dr. Harit Chhatri (Max Healthcare): "Access to quality healthcare... will improve significantly... all this will become accessible on a smartphone... we use these modern AI models as coaches for ourselves and we learn in the process."

Speakers & Organizations Mentioned

Government Officials:

  • Secretary, Department of Health and Family Welfare (opening remarks on cancer burden and policy)
  • Secretary, Ministry of Electronics and IT (strategic context on India AI mission)
  • Dr. Rajiv Bal, Secretary, Department of Health Research and Director-General, ICMR (validation framework announcement)
  • Kavita (CCO, India Mission; instrumental in grant mechanism development)

Clinical & Research Leaders:

  • Dr. Promesh (Director/Lead, Tata Memorial Hospital; National Cancer Grid)
  • Dr. Harit Chhatri (Max Healthcare; NCG member)
  • Dr. Mona Dubal (ICMR)
  • Dr. Kish (Director, Tata Memorial Hospital)

International Perspective:

  • Samir Pachi (WHO representative)

Award-Winning Teams (10 CATCH Challenge Winners):

  • Augmento (partnership: DEEP, Tata Memorial)
  • N Dimension (partnership: Tata Memorial Hospital)
  • Maniaa AI (partnership: KCDH, Ashoka University, Tata Memorial)
  • Cure.ai (partnership: Rajiv Gandhi Cancer Institute, Delhi)
  • Memo Q (partnership: KCDH, IIT Mumbai, Max Healthcare)
  • Bandhu Care (technical: Revan AI; clinical: Christian Medical College)
  • UML Serving AI
  • Vitics (partnership: Malabar Cancer Center)
  • Plus 2 additional award categories (Applications Development Initiative, AI-based Authentication Challenge)

Key Institutional Partners:

  • India AI Mission
  • National Cancer Grid (NCG)
  • ICMR (Indian Council of Medical Research)
  • IIT Mumbai
  • IIT Bombay
  • Ashoka University
  • KCDH (Quota Centers for Digital Health Research; referenced as "KCDH")
  • Tata Memorial Hospital
  • Max Healthcare
  • Rajiv Gandhi Cancer Institute
  • Christian Medical College
  • Ministry of Health and Family Welfare
  • Department of Health Research
  • WHO

Technical Concepts & Resources

AI Application Areas in Oncology:

  • Diagnostic radiology (image analysis, reducing false positives/negatives, increasing radiologist throughput)
  • Radiotherapy planning (treatment plan generation, reducing planning time from hours to minutes)
  • Pathology (endocrinopathology, histopathology automation)
  • Mammography screening
  • Treatment adherence monitoring
  • Tumor board support (integration of multi-disciplinary inputs)

Evaluation & Validation Concepts:

  • Randomized Controlled Trials (RCTs) — noted as slow; accelerated pathways being developed
  • Operational evaluation (how solutions function in real health system contexts)
  • Economic evaluation (cost-benefit and value proposition)
  • Ethical evaluation (dataset diversity, bias mitigation, localization for different populations)
  • Multi-stage technical and clinical evaluation (used in CATCH challenge)

Data & Infrastructure:

  • Dataset bias concerns (e.g., pulse oximetry validation across skin tone variations; dataset skew toward high-income country images)
  • Need for data curation as a standalone thematic area in grants
  • Localization and formatting of solutions for different contexts and settings

Key Metrics & Epidemiology:

  • India's cancer burden: 16+ lakh (1.6+ million) cases annually; 10+ lakh (1 million) deaths annually
  • Mortality-to-incidence ratio in India is higher than other high-burden countries (China, US), indicating late diagnosis and limited access
  • Two-thirds of patients die before treatment initiation
  • Estimated workforce gap: 1.4–2.6× multiplier needed to meet current demand (EY report referenced)
  • Global pathology shortage: 1 pathologist per 2.3 million population in sub-Saharan Africa vs. 1 per 15–20,000 in developed countries

Policy/Regulatory Framework:

  • Compendium of 10 solutions launched (details on technical specs, partnerships, outcomes)
  • Multi-stage evaluation framework with ICMR, IIT Bombay, Ashoka University, KCDH, and India AI as partners
  • CDSCO (Central Drugs Standards Control Organization) mentioned as regulatory partner
  • Phased deployment model: structured pilots → scaleup support → state-level implementation

Collaborative Models:

  • National Cancer Grid: 400+ cancer organizations aligned on common metrics
  • Applications Development Initiative (parallel challenge with 1,000+ applications across governance, climate/disaster management, health, learning disabilities, agriculture)
  • South-South partnership framework (India positioning itself as template-setter for global south)

Document Quality Note: The transcript contains significant repetition and audio artifacts (stuttering, overlapping speech marked with "Heat," "Heat up"), suggesting this is a raw automated transcription. Key substantive content has been extracted, but some speaker attributions and exact quotes may be approximate. The final section is notably compressed due to time constraints noted by moderators before the Prime Minister's arrival.