Learning sweep embeddings from blind obstetric ultrasound for gestational age prediction

dc.contributor.advisorDang, Vien Ngoc
dc.contributor.authorHristov, Christin
dc.date.accessioned2026-10-23T12:40:57Z
dc.date.available2026-10-23T12:40:57Z
dc.date.issued2026-06-28
dc.descriptionTreballs finals del Màster de Fonaments de Ciència de Dades, Facultat de matemàtiques, Universitat de Barcelona. Any: 2026. Tutor: Vien Ngoc Dang
dc.description.abstractGestational age (GA) prediction from blind-sweep obstetric ultrasound could help extend fetal dating to settings with limited access to trained sonographers. However, GA labels for training predictive models are harder to obtain than unlabeled sweep videos. This thesis studies whether a frozen self-supervised ultrasound encoder, combined with a lightweight sweep-level aggregator, can support label-efficient GA prediction when only a subset of training sweeps has GA labels. Using 1,314 blind-sweep videos from two acquisition sites, Spain and Kenya, three frame-encoder pretraining conditions and four sweep aggregators are compared. The resulting sweep embeddings are evaluated for GA prediction under full and reduced label budgets. Encoder pretraining quality appears to be the dominant factor among those tested. The best-performing encoder, which combines general ultrasound pretraining with continued pretraining on in-domain sweep data, outperforms a from-scratch encoder by 5.1 days mean absolute error. Aggregator choice and frozen versus trainable aggregation, by contrast, show much smaller effects, spanning only 1.14 days once the encoder is fixed, with frozen aggregators reaching similar performance to aggregators trained jointly with the prediction head. This encoder advantage also appears to carry over to label efficiency: continued pretraining at only 10% of training labels already reaches a weaker encoder’s accuracy at 100% of labels. Aggregator choice and training regime remain secondary across the label budgets tested. A gap of approximately 5 days remains to a fully supervised reference at full label availability, which may be associated with keeping the encoder frozen. For this dataset and fixed frame-selection pipeline, the results suggest prioritizing encoder pretraining over aggregator complexity, since aggregators pretrained without GA labels showed no consistent MAE drop compared with trainable aggregators.
dc.format.extent24 p.
dc.format.mimetypeapplication/pdf
dc.identifier.urihttps://hdl.handle.net/2445/232300
dc.language.isoeng
dc.rightscc-by-nc-nd (c) Hristov, Christin, 2026
dc.rights.accessRightsinfo:eu-repo/semantics/openAccess
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/3.0/es/
dc.subject.classificationEcografia obstètrica
dc.subject.classificationDiagnòstic prenatal
dc.subject.classificationAprenentatge automàtic
dc.subject.classificationTreballs de fi de màster
dc.subject.otherObstetrical ultrasound
dc.subject.otherPrenatal diagnosis
dc.subject.otherMachine learning
dc.subject.otherMaster's thesis
dc.titleLearning sweep embeddings from blind obstetric ultrasound for gestational age prediction
dc.typeinfo:eu-repo/semantics/masterThesis

Fitxers

Paquet original

Mostrant 1 - 1 de 1
Carregant...
Miniatura
Nom:
TFM_Hristov_Christin.pdf
Mida:
3.58 MB
Format:
Adobe Portable Document Format