轻量级AI系统实现甲状腺癌分类，准确率高且运行快。
ThyroidEffi 1.0: A Cost-Effective System for High-Performance Multi-Class Thyroid Carcinoma Classification
- 用YOLO检测细胞簇，分阶段融合多尺度图像提升识别效果。
- 外部验证达94.95%（良性）、83.96%（恶性）的准确率，1000例仅需30秒。
- 模型参数仅400万，适合普通设备部署，结果可解释性强。
背景：自动化分析甲状腺细针穿刺活检（FNAB）图像面临数据有限、医生间差异大及计算成本高的挑战。高效且可解释的模型对临床支持至关重要。目标：开发并外部验证一个深度学习系统，将FNAB图像分为三类：良性（Bethesda II）、可疑/不确定（BI, III, IV, V）和恶性（BVI），以指导越南临床治疗，实现高精度与低计算开销。方法：流程包括：(1) 使用YOLOv10检测细胞簇以提取关键区域并降噪；(2) 采用课程学习策略，逐步融合局部裁片与全图，实现多尺度捕捉；(3) 采用自适应轻量级EfficientNetB0（4M参数）平衡性能与效率；(4) 引入受Transformer启发的模块进行多尺度、多区域分析。外部验证使用1,015张独立FNAB图像。结果：ThyroidEffi Basic在内部测试集上达到宏F1 89.19%，AUC分别为0.98（良性）、0.95（可疑/不确定）、0.96（恶性）。外部验证中AUC分别为0.9495（良性）、0.7436（可疑/不确定）、0.8396（恶性）。ThyroidEffi Premium将宏F1提升至89.77%。Grad-CAM可视化显示关键诊断区域，证实可解释性。系统处理1000例仅耗时30秒，证明其可在普及硬件上运行。结论：本研究证明，在极低计算成本下实现高精度、可解释的甲状腺FNAB图像分类是可行的。
原文摘要 · Abstract (English)
Background: Automated classification of thyroid Fine Needle Aspiration Biopsy (FNAB) images faces challenges in limited data, inter-observer variability, and computational cost. Efficient, interpretable models are crucial for clinical support. Objective: To develop and externally validate a deep learning system for multi-class thyroid FNAB image classification into three key categories directly guiding post-biopsy treatment in Vietnam: Benign (Bethesda II), Indeterminate/Suspicious (BI, III, IV, V), and Malignant (BVI), achieving high diagnostic accuracy with low computational overhead. Methods: Our pipeline features: (1) YOLOv10 cell cluster detection for informative sub-region extraction/noise reduction; (2) curriculum learning sequencing localized crops to full images for multi-scale capture; (3) adaptive lightweight EfficientNetB0 (4M parameters) balancing performance/efficiency; and (4) a Transformer-inspired module for multi-scale/multi-region analysis. External validation used 1,015 independent FNAB images. Results: ThyroidEffi Basic achieved macro F1 of 89.19% and AUCs of 0.98 (Benign), 0.95 (Indeterminate/Suspicious), 0.96 (Malignant) on the internal test set. External validation yielded AUCs of 0.9495 (Benign), 0.7436 (Indeterminate/Suspicious), 0.8396 (Malignant). ThyroidEffi Premium improved macro F1 to 89.77%. Grad-CAM highlighted key diagnostic regions, confirming interpretability. The system processed 1000 cases in 30 seconds, demonstrating feasibility on widely accessible hardware. Conclusions: This work demonstrates that high-accuracy, interpretable thyroid FNAB image classification is achievable with minimal computational demands.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。