Back to projects
M.Tech ThesisIIT Bombay · IEOR · 2024–2026

Parameter-Specific Literature Retrieval for Kidney Diseases

A two-phase biomedical NLP project that takes a clinician from the flood of kidney-disease literature to an evidence-backed clinical answer. Phase 1 generalizes a parameter-specific retrieval system across 24 diseases; Phase 2 re-targets the same extraction machinery into a provenance-tracked knowledge graph with a question-answering engine that cites the papers behind every claim.

Supervisor: Prof. Balamurugan Palaniappan · Roll No. 24M1520

0.964
NephroSearch NDCG (vs 0.898 Google, 0.867 PubMed)
0.83
Ensemble NER micro-F1 across 11 entity types
88%
QA answer rate (V2 engine, up from 70%)
100%
Provenance coverage — every edge cites its source

The Two-Phase Arc

Phase 1Retrieval

Generalized NER & Structured Extraction

Generalized the project's extraction machinery from a single disease and four parameters to 24 kidney-related diseases and 10 clinical parameters.

  • Fetched ~160,000 PubMed abstracts (~7K per disease) and built an annotated corpus with Label Studio.
  • Fine-tuned a PubMedBERT NER model (spaCy, 9 iterative rounds) to extract Age, GFR, Proteinuria, Creatinine, BUN, BP, Hematuria, Weight, Calcium & Cholesterol.
  • Fine-tuned a BART Seq2Tree model that converts free-text measurements into structured XML (magnitude, unit, comparator).
  • Powers the NephroSearch web platform and the NephroTagger Chrome extension.
Phase 2Reasoning · ClinicalMind KG

Knowledge-Graph Question Answering

Re-targeted the same extraction stack from a ranked list to a Neo4j knowledge graph queried with natural-language clinical questions.

  • Composable, hot-swappable pipeline: document parsing → ensemble NER → normalization → relation extraction → KG writer.
  • Ensemble NER (BioBERT + clinical DeBERTa + zero-shot GLiNER) covering 13 KG entity types; UMLS-linked normalization to collapse synonyms.
  • REBEL relation extraction with a context-aware re-mapping stage that recovers clinical edges from structural labels.
  • Two-generation QA engine — entity resolution, type-aware routing, LLM-generated Cypher fallback, and an explainable step log — served through a Streamlit app.

Results & Evaluation

Rather than one opaque score, the system is measured stage by stage — NER, relation extraction, graph construction, and question answering — and then end to end. The Phase 2 graph and QA results below are reported on a curated 500-abstract demonstration graph spanning all 24 diseases.

Phase 1 — Ranking Quality (NDCG)

A parameter-aware ranker places the genuinely relevant trials higher than keyword engines can. The NER model reaches high-0.9 per-parameter F1 and the Seq2Tree model scores in the high-0.9 range across BLEU, ROUGE and METEOR.

SystemMean NDCG
NephroSearch (ours)0.9639
Google0.8978
PubMed0.8669

Phase 2 — Ensemble NER

Per-type precision / recall / F1 on a labelled sample. Specialised types score highest; zero-shot types pay the expected price.

Entity typePRF1
Disease0.910.930.92
Drug0.890.900.89
LabParameter0.830.790.81
Symptom0.800.820.81
Biomarker0.780.740.76
SideEffect0.720.690.70
MechanismOfAction0.680.600.64
Micro-average0.820.840.83

Phase 2 — Context-Aware Re-Mapping

REBEL reliably finds the related pair but labels it with Wikidata-style terms. Using known entity types recovers more than half of the otherwise-discarded clinical edges — no retraining.

Triplets reviewed150
Entity pair correct78%
Relation label correct (raw REBEL)41%
Structural labels recovered27 / 38
Relation label correct (re-mapped)63%

Phase 2 — Knowledge Graph

Built from 500 abstracts with UMLS normalization. Every factual edge carries the identifiers of the papers that support it.

Nodes / Edges1268 / 1312
Most common edge (TREATS)612
Provenance coverage100%
UMLS dedup (Disease nodes)142 → 119 (−16%)

Phase 2 — QA Engine (V1 vs V2)

On a 60-question test set. V2 adds entity resolution against the graph, type-aware routing and a fallback chain — lifting the answer rate by 18 points.

MeasureV1V2
Answer rate70%88%
Intent classification90%
Entity resolution85%

Phase 2 — Grounded QA vs. a Raw LLM

On 30 questions (5-point rubric). The decisive difference is evidence: the graph engine cites real, traceable papers on essentially every answer; the raw model rarely cites anything, and sometimes cites papers that do not exist.

CriterionKG-QA (V2)Raw LLM
Relevance4.24.0
Specificity4.03.4
Citable evidence present4.61.2

Scope note: the Phase 2 graph is a demonstration-scale graph (500 abstracts) and the answer-quality rubric was scored by a single annotator; formal clinical validation and full-corpus ingestion are identified as future work in the thesis.

Stack

PyTorchHuggingFacePubMedBERTBART / Seq2TreeGLiNERREBELspaCy / scispaCyNeo4j / CypherUMLSOllama / OpenAI APIStreamlitFastAPI