Deep-learning nomogram reproduced 3D treated volume for prostate HDR brachytherapy QA
A 3D deep-learning nomogram predicted the clinical prescription isodose with high spatial agreement, adding geometry-aware verification to post-plan QA.
A 3D deep-learning nomogram predicted the clinical prescription isodose with high spatial agreement, adding geometry-aware verification to post-plan QA.
In 22 patients, salvage HDR brachytherapy produced 40.5% five-year relapse-free survival but 18% developed grade ≥3 GU toxicity.
Nearly three in ten surveyed cervical cancer survivors reported partial-to-probable PTSD symptoms several years after HDR brachytherapy.
Across 286 HDR monotherapy patients, higher EQD2 D90 was not associated with biochemical control or overall survival.
In 607 predominantly high-risk patients, 15 Gy HDR boost plus EBRT produced durable six-year disease control with low severe late toxicity.
Motorized probe rotation improved sagittal 3D ultrasound consistency and eliminated significant inter-operator variability seen with manual acquisition.
In 223 previously irradiated patients, 3×10 Gy and 2×13 Gy salvage HDR brachytherapy produced similar disease control and late toxicity.
Iridium-192 has the strongest long-term breast brachytherapy evidence; cobalt-60 is dosimetrically comparable, while ytterbium-169 remains clinically unvalidated.
Short-course EBRT plus endorectal HDR brachytherapy produced 80% cCR/ncCR, but grade 3 late rectal toxicity occurred in 27%.
Compressing three HDR brachytherapy fractions into one week shortened overall treatment time and increased early complete response without improving survival.
Routine CT-based adaptive brachytherapy achieved 84.4% five-year local control despite 56.9% of patients having stage III–IV cervical cancer.
Five-year PSA progression-free survival was 24.8% after two-fraction salvage HDR brachytherapy without androgen deprivation, despite low severe toxicity.
The standardized workflow generated clinically comparable interstitial brachytherapy plans in a median of 10 minutes versus 28 minutes for manual planning.
Template-guided implantation achieved a median HR-CTV D90 of 92.9 Gy with low severe toxicity across 555 brachytherapy fractions.
Brief virtual-reality training increased self-reported procedural confidence, but knowledge, technical skill, retention, and clinical performance were not tested.