Research Portfolio Cases

Each case is presented as a proposal for a three-year internal grant of Rs 15 lakh. The institution can fund only one at full value. Score each case on the evaluation rubric, then decide at your table which to fund. The cases are fictional composites written for this exercise; names, places and figures are invented and are not drawn from any real project.

Case 1 (Health Sciences): Salivary Biomarkers for Early Periodontal Risk in Adults With Type 2 Diabetes

Proposing unit. Department of Periodontology, with the Department of Medicine and the hospital's endocrinology clinic.

Background. The team has published four papers over five years on salivary inflammatory markers in periodontal disease, two of them in journals in the top quartile of their field. The lead investigator has guided six MDS theses on related topics. The endocrinology clinic sees about 900 adults with type 2 diabetes each year, most of whom have never had a periodontal examination.

Proposal. A prospective cohort of 400 clinic patients, screened at baseline and at 12 and 24 months for periodontal status and a panel of three salivary markers, to develop a low-cost risk score that a non-dental clinician could apply during a routine diabetes visit. Outputs: a validated risk score, three papers, one MDS and one PhD thesis, an anonymised dataset. Intended outcome: adoption of the score in the hospital's diabetes clinic and a proposal to the state health department for a pilot in five district hospitals. Partners: the hospital's endocrinology clinic (confirmed); a state health department official who has expressed interest but has not committed.

Evidence of capacity. Ethics approval from the institutional committee obtained for the pilot phase. A biostatistician is available for four hours a week. The salivary assays are run in a laboratory that has done 200 samples for the earlier studies; the proposal needs about 1,200.

Risks. Loss to follow-up in a two-year cohort; the assay laboratory's capacity; the state department's interest is verbal only.

Budget request. Rs 15 lakh: assays Rs 9 lakh, research assistant Rs 4.5 lakh, patient travel reimbursement Rs 1.5 lakh.

Case 2 (Engineering): Deep-Learning Classification of Retinal Images Using a Public Dataset

Proposing unit. Department of Computer Science and Engineering, single department.

Background. The team has published eleven conference papers and three journal papers in three years on convolutional network architectures applied to publicly available image datasets. Citations are modest. Two PhD scholars are in their second year. The lead investigator has no clinical collaborator and has not visited an ophthalmology clinic.

Proposal. To benchmark six network architectures on a public retinal image dataset from another country, propose a new attention mechanism, and report improvements in classification accuracy. Outputs: six to eight papers, two PhD theses, a code repository. Intended outcome: a possible future collaboration with a hospital, not yet identified. The proposal cites diabetic retinopathy as a national health priority.

Evidence of capacity. The department has GPU workstations sufficient for the work. The team's prior results are reproducible. No ethics approval is needed for the public dataset.

Risks. Results on a foreign dataset may not transfer to Indian populations or to images from low-cost cameras used in Indian screening camps; the improvement over existing methods is likely to be incremental.

Budget request. Rs 15 lakh: GPU upgrade Rs 8 lakh, conference travel Rs 5 lakh, publication charges Rs 2 lakh.

Case 3 (Social Sciences and Management): Why Enrolled Households Do Not Use Their Health Insurance

Proposing unit. School of Management with the School of Behavioural Sciences.

Background. A young faculty member with two published papers, one in a well-regarded Indian journal, has been approached by a district administration that has noticed low utilisation of a public health-insurance scheme among enrolled households in two blocks. The faculty member has a mentor in the School of Management and a co-investigator in Behavioural Sciences.

Proposal. A mixed-methods study: household survey of 1,200 enrolled families across two blocks, interviews with hospital billing staff and scheme officials, and a behavioural intervention (simplified information card and a phone reminder) tested in a randomised design across village clusters. Outputs: two papers, a policy brief for the district administration, a working paper, an anonymised dataset. Intended outcome: the district adopts the intervention if it works; the state programme office reviews the findings. Partners: the district administration (letter of support received), a primary health centre network (confirmed).

Evidence of capacity. Survey capacity through a trained student cohort in the School of Management. The randomised design has been reviewed by the mentor but not yet by a statistician. Ethics approval is pending.

Risks. The lead investigator is early-career and has never run a field study of this size; the district's attention may move elsewhere after an election; the intervention effect may be small.

Budget request. Rs 15 lakh: field survey Rs 8 lakh, intervention materials and phone reminders Rs 3 lakh, research assistant Rs 3 lakh, dissemination Rs 1 lakh.