Conectar
by Dilara Tank, Bianca G. S. Schor, Lisa M. Trommelen, Judith A. F. Huirne, Iacer Calixto, Robert A. de Leeuw
PurposeTransvaginal ultrasound (TVUS) is pivotal for diagnosing reproductive pathologies in individuals assigned female at birth, often serving as the primary imaging method for gynecologic evaluation. Despite recent advancements in AI-driven segmentation, its application to gynecological ultrasound still needs further attention. Our study aims to bridge this gap by training and evaluating two state-of-the-art deep learning (DL) segmentation models on TVUS data.
Materials and methodsAn experienced gynecological expert manually segmented the uterus in our TVUS dataset of 124 patients with adenomyosis, comprising still images (n = 122), video screenshots (n = 472), and 3D volume screenshots (n = 452). Two popular DL segmentation models, U-Net and nnU-Net, were trained on the entire dataset, and each imaging type was trained separately. Optimization for U-Net included varying batch size, image resolution, pre-processing, and augmentation. Model performance was measured using the Dice score (DSC).
ResultsU-Net and nnU-Net had good mean segmentation performances on the TVUS uterus segmentation dataset (0.75 to 0.97 DSC). We observed that training on specific imaging types (still images, video screenshots, 3D volume screenshots) tended to yield better segmentation performance than training on the complete dataset for both models. Furthermore, nnU-Net outperformed the U-Net across all imaging types. Lastly, we report the best results using the U-Net model with limited pre-processing and augmentations.
ConclusionsTVUS datasets are well-suited for DL-based segmentation. nnU-Net training was faster and yielded higher segmentation performance; thus, it is recommended over manual U-Net tuning. We also recommend creating TVUS datasets that include only one imaging type and are as clutter-free as possible. The nnU-Net strongly benefited from being trained on 3D volume screenshots in our dataset, likely due to their lack of clutter. Further validation is needed to confirm the robustness of these models on TVUS datasets. Our code is available on https://github.com/dilaratank/UtiSeg.
To examine the impact of the COVID-19 pandemic on the substitution of surgical procedures in benign gynaecology in the Netherlands.
Quantitative longitudinal study evaluating the effects of the COVID-19 pandemic.
Nationwide healthcare delivery was analysed across six benign gynaecological pathways from 2016 to 2022 using Vektis and Dutch Hospital Data (DHD), accessed via Statistics Netherlands (Centraal Bureau voor de Statistiek).
The study focused on six benign gynaecological pathways classified using Dutch Diagnosis Treatment Combinations (DTCs): heavy menstrual blood loss (G11), uterine fibroids (G15), endometriosis (G17), prolapse (G25), infertility treatment (F11) and first trimester pregnancy complications (Z12). All patients receiving care within these pathways between 1 January 2016 and 31 December 2022 were included. Exclusions applied to all patients under 18 years old and, only within the menstrual disorder pathway, patients over 51 years old to exclude most postmenopausal blood loss cases where no alternative treatment applies.
Cohorts from the initial pandemic year (2020) were compared with four prepandemic cohorts (2016–2019) and late-pandemic (2021) and postpandemic (2022) cohorts.
The primary outcome was the trend in the total number of patients in surgical and non-surgical procedure groups across cohort periods. Secondary outcomes included trends within individual pathways.
The analysis identified a significant reduction in benign gynaecological care during 2020, with an 18.3% (p
The COVID-19 pandemic significantly disrupted both surgical and non-surgical procedures within benign gynaecological pathways. Reduced care uptake during the pandemic waves was not recovered but instead forgone. The reduction in surgical procedures did not correspond with increased use of non-surgical alternatives. Future research should prioritise evaluating the long-term impacts of this disruption on patients and society.
FreshRSS 