
		<paper>
			<loc>https://jjcit.org/paper/305</loc>
			<title>BREAST LESION CLASSIFICATION: LEAKAGE-FREE PATCH-BASED TRANSFER LEARNING WITH CROSS- DATASET EVALUATION ON CBIS-DDSM, MIAS AND INBREAST</title>
			<doi>10.5455/jjcit.71-1783079555</doi>
			<authors>Hitesh Kag,Dinesh Kumar,Anjali Diwan</authors>
			<keywords>Breast lesion classification,Data leakage,Transfer learning,External validation,Mammography</keywords>
			<views>1198</views>
			<downloads>376</downloads>
			<received_date>3-Jul.-2026</received_date>
			<revised_date>  23-Aug.-2026 and 27-Aug.-2026</revised_date>
			<accepted_date>  28-Aug.-2026</accepted_date>
			<abstract>High reported accuracies on small mammographic datasets are frequently inflated by data leakage, most often 
from augmentation applied before the train/test split. Changing only the split order on whole-image Mini-MIAS 
moves a  deep-feature  plus  Support  Vector  Machine  (SVM)  pipeline  from  0.976  accuracy to 0.513, which is 
equivalent to the random chance. The predictive signal survives at the lesion-patch scale. Our contribution is 
protocol-level rather than architectural: a reproducible leakage audit; the first quantification and repair of the 
reported patient overlap in the CBIS-DDSM official split, which we trace to its mass and calcification partitions 
being drawn independently (31 patients, 8.5% of test lesions); an architecture sweep that tests rather than assumes 
the performance ceiling; and a label-free procedure for transferring the decision threshold to a new institution. A 
lightweight EfficientNet patch pipeline trained on CBIS-DDSM under a strict patient-wise protocol obtains ROC-
AUC 0.745 (95% CI 0.703–0.794); without retraining, it attains 0.734 (0.640–0.826) on MIAS and 0.877 (0.804–
0.946) on INbreast and a DeLong test cannot distinguish the MIAS result from the internal one (p = 0.82). Results 
are stable over ten seeds (σ≤ 0.011) with clustered bootstrap intervals. Across 11 configurations spanning six 
architecture families, test AUC spans only 0.707–0.777. For small datasets, leakage control and external 
validation matter more than headline accuracy.</abstract>
		</paper>


