2026-10-14T14:38:01Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34015555 2026-09-28T17:23:03Z category_7 category_12 category_21 category_873 category_734 category_811 category_64 portal_5 item_type_3 month_year_09_2026

<p>Model training hyperparameter setting summary.</p> Nianbo Xu (25132495) Andy Chun Hang Chen (25132498) Hanzhang Ruan (25132501) Donglin Yang (490406) Renjie Liao (9957992) Xiaojuan Qi (11976451) Sze Wan Fong (25132504) Dandan Cao (102811) Lu Yu (74016) Yuanhua Huang (663816) William Shu Biu Yeung (25132507) Ernest Hung Yu Ng (8672868) Yin Lau Lee (25132510) Medicine Cell Biology Biotechnology Chemical Sciences not elsewhere classified Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Cancer live birth outcomes frozen feature extractor examine whether concentrating div >< p cumulative live birth smaller training set repurposed foundation models powered deep learning findings support histology 7 versus non integrated model ’ 90 ± 0 84 ± 0 80 ± 0 74 ± 0 71 ± 0 cycle endometrial biopsies oriented endometrial assessment oriented learning endometrial histology smaller contributions reproducible assessment foundation model 7 phases endometrial thickness endometrial receptivity vitro </ validation ensembles promising bases observer variability negligible contributions near random multimodal analyses mean pooling luminal epithelium internal outcome internal held fold cross fertilization cycle external phase ensembled aurocs clinical metadata cam review based testing <div><p>Objective and reproducible assessment of endometrial receptivity is essential for optimizing <i>in vitro</i> fertilization (IVF) success, yet traditional histological dating suffers from observer variability. This study investigated whether deep learning of hematoxylin and eosin histology images could support cumulative live birth prediction in IVF. An end-to-end ResNet-18 was compared with a UNI2-h-based pipeline, using UNI2-h as a frozen feature extractor. Ten-fold cross-validation ensembles were developed from natural-cycle endometrial biopsies and evaluated in an internal held-out cohort with live birth outcomes. Additional phase-based evaluation was performed, which tested the performance in distinguishing LH + 7 versus non-LH + 7 phases. A luminal epithelium (LE)-focused design was also assessed to examine whether concentrating on maternal-embryo interface could improve fertility-oriented learning. The whole-slide ResNet-18 model performed well in internal outcome-based testing but near random in external phase-based testing. In contrast, the UNI2-h model with mean pooling and a multilayer perceptron classifier showed less divergent performance, with ensembled AUROCs of 0.74 ± 0.03 in internal outcome-based evaluation, and 0.90 ± 0.05 in external phase-based testing. Despite a smaller training set, LE-focused models retained comparable performance. In internal outcome-based testing, LE-focused ResNet-18 and UNI2-h models achieved AUROCs of 0.71 ± 0.03 and 0.74 ± 0.09 respectively. External phase-based testing yielded AUROCs of 0.80 ± 0.16 for ResNet-18, and 0.84 ± 0.06 for UNI2-h. Grad-CAM review of LE-focused ResNet-18 models showed attention commonly on LE-alone or mixed with adjacent stroma. In multimodal analyses, integrated models incorporating histology significantly outperformed the clinical metadata-only model. Integrated model’s feature weighting showed dominant histology outputs, smaller contributions from estradiol and maternal age, and negligible contributions from progesterone, endometrial thickness, and BMI. These findings support histology-based AI for fertility-oriented endometrial assessment and highlight biologically informed design and repurposed foundation models as promising bases for clinically meaningful prediction.</p></div> 2026-09-28T17:22:52Z Dataset Dataset 10.1371/journal.pdig.0001744.s008 https://figshare.com/articles/dataset/_p_Model_training_hyperparameter_setting_summary_p_/34015555 CC BY 4.0