[
  {
    "id": "10.3390/a19111011",
    "type": "article-journal",
    "title": "Breast Ultrasound Classification Under Cross-Dataset Shift: A Multi-Seed Benchmark of Convolutional Networks, a Vision Transformer, and a Self-Supervised Foundation Model",
    "author": [
      {
        "family": "Ramollari",
        "given": "Ervin"
      },
      {
        "family": "Vajjhala",
        "given": "Narasimha Rao"
      },
      {
        "family": "Kumar",
        "given": "Rajesh"
      },
      {
        "family": "Ramchandani",
        "given": "Komal"
      },
      {
        "family": "Mahraj",
        "given": "Ranjeeta"
      },
      {
        "family": "Hussain",
        "given": "Khalid"
      }
    ],
    "container-title": "Algorithms",
    "issued": {
      "date-parts": [
        [
          2026,
          11,
          18
        ]
      ]
    },
    "volume": "19",
    "issue": "11",
    "number": "1011",
    "publisher": "MDPI",
    "ISSN": "1999-4893",
    "DOI": "10.3390/a19111011",
    "URL": "https://www.narasimharao.net/research/breast-ultrasound-classification-cross-dataset-shift-multi-seed-benchmark/",
    "abstract": "Deep-learning classifiers for breast ultrasound are usually evaluated on images from their training institution, and self-supervised foundation models are increasingly proposed as more transferable than supervised networks; whether ranking ability, decision thresholds, and probability calibration survive a change in institution has rarely been tested under leakage-controlled conditions. We fine-tuned ResNet-50, EfficientNet-B0, a supervised Vision Transformer (ViT-B/16), and self-supervised DINOv2 under one shared protocol on the official BUS-BRA five-fold partitions, each fold with three seeds, and on three case-grouped hold-out splits, and evaluated every trained model unchanged on the Breast Ultrasound Images (BUSI) dataset. The supervised models largely retained their discrimination externally. ResNet-50 had the highest mean area under the receiver operating characteristic curve under the official folds, but no criterion separated it from ViT-B/16, whose small hold-out advantage at the decision threshold did not recur. In every official-fold run, both models labeled a larger share of BUSI than of BUS-BRA images malignant, although prevalence was the same—a shift that temperature scaling cannot correct. Top-label calibration error also understated malignant-class miscalibration. Full-frame DINOv2 did not fit its training data in any run, although exploratory lesion cropping produced good discrimination. External evaluations should report operating-point transport and class-wise calibration alongside discrimination.",
    "keyword": "breast ultrasound, foundation models, self-supervised learning, DINOv2, vision transformer, convolutional neural network, external validation, dataset shift, probability calibration, decision threshold, breast ultrasound classification, medical image analysis, deep learning, cross-institution generalization, ResNet-50, EfficientNet-B0, Vision Transformer, temperature scaling, decision thresholds, reproducibility, BUS-BRA dataset, BUSI dataset",
    "language": "en"
  }
]