ShifaMind: A Multiplicative Concept Bottleneck for Interpretable ICD-10 Coding

2026-05-08

AI Summary

Automated ICD-10 coding faces dual challenges of long-tailed multi-label classification and clinical interpretability, with conventional Concept Bottleneck Models (CBMs) suffering performance limitations due to information compression. This work proposes ShifaMind, a novel architecture that reimagines the concept bottleneck by replacing narrow-layer projection with a multiplicative gating mechanism. This design preserves human-interpretable, verifiable scalar concept interfaces while mitigating representational loss through gated modulation. By integrating clinical text embeddings with grounded concept representations, ShifaMind achieves F1, AUC, and ranking metrics on the MIMIC-IV top-50 task comparable to the strongest baseline, LAAT, and significantly outperforms six other baselines—including Vanilla CBM—while delivering reliable concept-level explanations.

Abstract

Automated ICD-10 coding from clinical discharge summaries requires models that are both accurate on long-tailed multi-label classification tasks and interpretable to clinicians. Concept Bottleneck Models (CBMs) offer a principled framework for interpretability by routing predictions through human-interpretable concepts, but this transparency often comes at a cost: compressing rich clinical text representations into a narrow concept layer can restrict gradient flow and limit predictive capacity. We present ShifaMind, a concept-grounded architecture built around a Multiplicative Concept Bottleneck (MCB), which changes the form, rather than the width, of the bottleneck. Instead of projecting through a narrow concept layer, ShifaMind uses a learned multiplicative gate over a concept-grounded representation while retaining a scalar concept interface for inspection. On MIMIC-IV top-50 ICD-10 coding, ShifaMind achieves performance competitive with LAAT, the strongest baseline, across F1, AUC, and ranking metrics, while outperforming five additional ICD-coding baselines and providing concept-mediated explanations. Its substantial gains over a capacity-matched Vanilla CBM in both predictive performance and interpretability-oriented metrics highlight the importance of the bottleneck design.

Problem

• ICD-10 coding
• interpretability
• Concept Bottleneck Models
• multi-label classification
• clinical text

Innovation

• Multiplicative Concept Bottleneck
• Interpretable AI
• ICD-10 coding
• Concept Bottleneck Models
• Clinical Text Processing