What Makes Deep Learning Work for Traditional Chinese Medicine Tongue Diagnosis? A Comprehensive Ablation Study

• URL: http://arxiv.org/abs/2607.28148v1
• Date: Thu, 30 Jul 2026 12:55:05 GMT
• Title: What Makes Deep Learning Work for Traditional Chinese Medicine Tongue Diagnosis? A Comprehensive Ablation Study
• Abstract summary: We conducted a systematic ablation study spanning 20+ model versions on TongueDx2 (5,109 images, 976 expert-annotated) and a merged dataset of 11,101 samples.<n>Best 976-sample model achieved weighted-F1 of 0.6625 using ConvNeXt-Tiny with restrained augmentation and weak-group ensemble.<n>Six key design principles emerged: ConvNeXt-Tiny offers optimal parameter efficiency; BCE substantially outperforms Asymmetric Loss; restrained color augmentation is critical; weak-group ensemble replacement is critical.
• Score: 6.004599301889361
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Deep learning has shown promise for automated tongue diagnosis in traditional Chinese medicine (TCM), yet the design space remains underexplored. We conducted a systematic ablation study spanning 20+ model versions under rigorous 5-fold cross-validation on TongueDx2 (5,109 images, 976 expert-annotated) and a merged dataset of 11,101 samples. We compared six backbone architectures, four loss functions, five augmentation strategies, and six training strategies. The best 976-sample model achieved weighted-F1 of 0.6625 using ConvNeXt-Tiny with restrained augmentation and weak-group ensemble, while the best 11,101-sample model reached weighted-F1 of 0.7761. Six key design principles emerged: (1) ConvNeXt-Tiny offers optimal parameter efficiency; (2) BCE substantially outperforms Asymmetric Loss (+2.7%); (3) restrained color augmentation is critical; (4) weak-group ensemble replacement (+2.1%) outperforms probability averaging; (5) data scaling yielded +20.6% improvement; (6) expanding from 13 to 45 label dimensions caused catastrophic collapse (0.78 to 0.22). These principles are generalizable to multi-label medical image classification with class imbalance.

Related papers

• DiffEEG: A Self-Supervised Denoising Diffusion Model for Learning EEG Generic Representations [1.901880164206253]
  We present DiffEEG, a self-supervised foundation model for deep learning for EEG-based seizure detection.<n>Pre-trained on 1.3M unlabeled segments from the Temple University Hospital Seizure Corpus (TUHSZ), DiffEEG learns generic neural representations via a 1D U-Net with multi-head self-attention.<n>Under strict patient-wise evaluation, DiffEEG achieves 61% accuracy and 59% F1 for 4-class seizure subtyping, and 81% accuracy with 85% weighted F1 for binary detection.
  arXiv  Detail & Related papers  (2026-07-13T13:59:45Z)
• MorphDistill: Distilling Unified Morphological Knowledge from Pathology Foundation Models for Colorectal Cancer Survival Prediction [3.2844015242965927]
  We propose MorphDistill, a framework that distills complementary knowledge from multiple pathology foundation models into a CRC-specific encoder.<n>In Stage I, a student encoder is trained using dimension-agnostic multi-teacher distillation with supervised contrastive regularization on large-scale colorectal datasets.<n>In Stage II, the encoder extracts patch-level features from whole-slide images, which are aggregated via attention-based multiple instance learning to predict five-year survival.
  arXiv  Detail & Related papers  (2026-04-07T19:21:18Z)
• XAI-Driven Skin Disease Classification: Leveraging GANs to Augment ResNet-50 Performance [2.7930955543692817]
  This study proposes a trustworthy and highly accurate Computer-Aided Diagnosis (CAD) system to overcome limitations.<n>The approach utilizes Deep Convolutional Generative Adversarial Networks (DCGANs) for per class data augmentation to resolve the critical class imbalance problem.<n>The system achieved a high overall Accuracy of 92.50 % and a Macro-AUC of 98.82 %, successfully outperforming various prior benchmarked architectures.
  arXiv  Detail & Related papers  (2025-11-29T20:46:30Z)
• Transparent Early ICU Mortality Prediction with Clinical Transformer and Per-Case Modality Attribution [42.85462513661566]
  We present a lightweight, transparent multimodal ensemble that fuses physiological time-series measurements with unstructured clinical notes from the first 48 hours of an ICU stay.<n>A logistic regression model combines predictions from two modality-specific models: a bidirectional LSTM for vitals and a finetuned ClinicalModernBERT transformer for notes.<n>On the MIMIC-III benchmark, our late-fusion ensemble improves discrimination over the best single model while maintaining well-calibrated predictions.
  arXiv  Detail & Related papers  (2025-11-19T20:11:49Z)
• MeCaMIL: Causality-Aware Multiple Instance Learning for Fair and Interpretable Whole Slide Image Diagnosis [40.3028468133626]
  Multiple instance learning (MIL) has emerged as the dominant paradigm for whole slide image (WSI) analysis in computational pathology.<n>textbfMeCaMIL, a causality-aware MIL framework, explicitly models demographic confounders through structured causal graphs.<n>MeCaMIL achieves superior fairness -- demographic disparity variance drops by over 65% relative reduction on average across attributes.
  arXiv  Detail & Related papers  (2025-11-14T06:47:21Z)
• A Disease-Centric Vision-Language Foundation Model for Precision Oncology in Kidney Cancer [54.58205672910646]
  RenalCLIP is a visual-language foundation model for characterization, diagnosis and prognosis of renal mass.<n>It achieved better performance and superior generalizability across 10 core tasks spanning the full clinical workflow of kidney cancer.
  arXiv  Detail & Related papers  (2025-08-22T17:48:19Z)
• Addressing High Class Imbalance in Multi-Class Diabetic Retinopathy Severity Grading with Augmentation and Transfer Learning [1.5939351525664014]
  This paper presents a robust deep learning framework for both binary and five-class Diabetic retinopathy (DR) classification.<n>For binary classification, our proposed model achieves a state-of-the-art accuracy of 98.9%, with a precision of 98.6%, recall of 99.3%, F1-score of 98.9%, and an AUC of 99.4%.<n>In the more challenging five-class severity classification task, our model obtains a competitive accuracy of 84.6% and an AUC of 94.1%, outperforming several existing approaches.
  arXiv  Detail & Related papers  (2025-07-23T01:52:27Z)
• Predicting Length of Stay in Neurological ICU Patients Using Classical Machine Learning and Neural Network Models: A Benchmark Study on MIMIC-IV [49.1574468325115]
  This study explores multiple ML approaches for predicting LOS in ICU specifically for the patients with neurological diseases based on the MIMIC-IV dataset.<n>The evaluated models include classic ML algorithms (K-Nearest Neighbors, Random Forest, XGBoost and CatBoost) and Neural Networks (LSTM, BERT and Temporal Fusion Transformer)
  arXiv  Detail & Related papers  (2025-05-23T14:06:42Z)
• Advancing Tabular Stroke Modelling Through a Novel Hybrid Architecture and Feature-Selection Synergy [0.9999629695552196]
  The present work develops and validates a data-driven and interpretable machine-learning framework designed to predict strokes.<n>Ten routinely gathered demographic, lifestyle, and clinical variables were sourced from a public cohort of 4,981 records.<n>The proposed model achieved an accuracy rate of 97.2% and an F1-score of 97.15%, indicating a significant enhancement compared to the leading individual model.
  arXiv  Detail & Related papers  (2025-05-18T21:46:45Z)
• Correcting Class Imbalances with Self-Training for Improved Universal Lesion Detection and Tagging [43.06199185109424]
  Universal lesion detection and tagging (ULDT) in CT studies is critical for tumor burden assessment and tracking the progression of lesion status (growth/shrinkage) over time.<n>Prior work used the DeepLesion dataset (4,427 patients, 10,594 studies, 32,120 CT slices, 32,735 lesions, 8 body part labels) for algorithmic development, but this dataset is not completely annotated and contains class imbalances.<n>We developed a self-training pipeline for ULDT using a limited 11.5% subset of DeepLesion.
  arXiv  Detail & Related papers  (2025-04-07T15:57:03Z)
• Benchmarking Reasoning Robustness in Large Language Models [76.79744000300363]
  We find significant performance degradation on novel or incomplete data.<n>These findings highlight the reliance on recall over rigorous logical inference.<n>This paper introduces a novel benchmark, termed as Math-RoB, that exploits hallucinations triggered by missing information to expose reasoning gaps.
  arXiv  Detail & Related papers  (2025-03-06T15:36:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.