Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Experience-Guided Self-Adaptive Cascaded Agents for Breast Cancer Screening and Diagnosis with Reduced Biopsy Referrals

Pramit Saha, Mohammad Alsharid, Joshua Strong, J. Alison Noble

TL;DR

Evaluation across 10 breast ultrasound datasets shows that the proposed experience-guided workflow reduces diagnostic escalation in BUSD-Agent from 84.95% to 58.72% and overall biopsy referrals from 59.50% to 37.08%, compared to the same architecture without trajectory conditioning.

Abstract

We propose an experience-guided cascaded multi-agent framework for Breast Ultrasound Screening and Diagnosis, called BUSD-Agent, that aims to reduce diagnostic escalation and unnecessary biopsy referrals. Our framework models screening and diagnosis as a two-stage, selective decision-making process. A lightweight `screening clinic' agent, restricted to classification models as tools, selectively filters out benign and normal cases from further diagnostic escalation when malignancy risk and uncertainty are estimated as low. Cases that have higher risks are escalated to the `diagnostic clinic' agent, which integrates richer perception and radiological description tools to make a secondary decision on biopsy referral. To improve agent performance, past records of pathology-confirmed outcomes along with image embeddings, model predictions, and historical agent actions are stored in a memory bank as structured decision trajectories. For each new case, BUSD-Agent retrieves similar past cases based on image, model response and confidence similarity to condition the agent's current decision policy. This enables retrieval-conditioned in-context adaptation that dynamically adjusts model trust and escalation thresholds from prior experiences without parameter updates. Evaluation across 10 breast ultrasound datasets shows that the proposed experience-guided workflow reduces diagnostic escalation in BUSD-Agent from 84.95% to 58.72% and overall biopsy referrals from 59.50% to 37.08%, compared to the same architecture without trajectory conditioning, while improving average screening specificity by 68.48% and diagnostic specificity by 6.33%.

Experience-Guided Self-Adaptive Cascaded Agents for Breast Cancer Screening and Diagnosis with Reduced Biopsy Referrals

TL;DR

Evaluation across 10 breast ultrasound datasets shows that the proposed experience-guided workflow reduces diagnostic escalation in BUSD-Agent from 84.95% to 58.72% and overall biopsy referrals from 59.50% to 37.08%, compared to the same architecture without trajectory conditioning.

Abstract

We propose an experience-guided cascaded multi-agent framework for Breast Ultrasound Screening and Diagnosis, called BUSD-Agent, that aims to reduce diagnostic escalation and unnecessary biopsy referrals. Our framework models screening and diagnosis as a two-stage, selective decision-making process. A lightweight `screening clinic' agent, restricted to classification models as tools, selectively filters out benign and normal cases from further diagnostic escalation when malignancy risk and uncertainty are estimated as low. Cases that have higher risks are escalated to the `diagnostic clinic' agent, which integrates richer perception and radiological description tools to make a secondary decision on biopsy referral. To improve agent performance, past records of pathology-confirmed outcomes along with image embeddings, model predictions, and historical agent actions are stored in a memory bank as structured decision trajectories. For each new case, BUSD-Agent retrieves similar past cases based on image, model response and confidence similarity to condition the agent's current decision policy. This enables retrieval-conditioned in-context adaptation that dynamically adjusts model trust and escalation thresholds from prior experiences without parameter updates. Evaluation across 10 breast ultrasound datasets shows that the proposed experience-guided workflow reduces diagnostic escalation in BUSD-Agent from 84.95% to 58.72% and overall biopsy referrals from 59.50% to 37.08%, compared to the same architecture without trajectory conditioning, while improving average screening specificity by 68.48% and diagnostic specificity by 6.33%.
Paper Structure (16 sections, 7 figures, 1 table)

This paper contains 16 sections, 7 figures, 1 table.

Table of Contents

  1. Introduction
  2. BUSD-Agent
  3. A Novel Cascaded Multi-Agent Framework
  4. Screening Clinic Agent
  5. Diagnostic Clinic Agent
  6. Experience-Guided In-Context Adaptation
  7. Biopsy-Grounded Memory Bank
  8. Experience-Conditioned Policy Adaptation for Screening Stage
  9. Experience-Conditioned Policy Adaptation for Diagnostic Stage
  10. Experiments and Results
  11. Datasets and Implementation Details
  12. Performance of Screening Agent
  13. Performance of Diagnostic Agent
  14. Comparison with SOTA VLMs
  15. Ablation Studies
  16. ...and 1 more sections

Figures (7)

  • Figure 1: BUSD-Agent overview and results. (a) BUSD-Agent Real-time User Interface, (b)# Samples in 10 evaluation datasets, (c) Specificity comparison of baseline vs proposed method for Screening (left) and Diagnostic (right) agents.
  • Figure 2: Comparison of our BUSD-Agent Framework with existing methods
  • Figure 3: Our experience-guided in-context adaptation strategy for BUSD-Agent
  • Figure 4: Screening confusion matrices (%) for 10 datasets comparing baseline (w/o experience-conditioning) and proposed approach (experience-conditioned).
  • Figure 5: Screening Samples. Left 3 cases were correctly predicted by both. Right 3 cases show harder instances where baseline escalated to diagnostic agent, whereas proposed experience-guided agent correctly identified as benign.
  • ...and 2 more figures