Development of a Machine Learning Model for Sonographic Assessment of Gestational Age.

Importance: Fetal ultrasonography is essential for confirmation of gestational age (GA), and accurate GA assessment is important for providing appropriate care throughout pregnancy and for identifying complications, including fetal growth disorders. Derivation of GA from manual fetal biometry measurements (ie, head, abdomen, and femur) is operator dependent and time-consuming. Objective: To develop artificial intelligence (AI) models to estimate GA with higher accuracy and reliability, leveraging standard biometry images and fly-to ultrasonography videos. Design, Setting, and Participants: To improve GA estimates, this diagnostic study used AI to interpret standard plane ultrasonography images and fly-to ultrasonography videos, which are 5- to 10-second videos that can be automatically recorded as part of the standard of care before the still image is captured. Three AI models were developed and validated: (1) an image model using standard plane images, (2) a video model using fly-to videos, and (3) an ensemble model (combining both image and video models). The models were trained and evaluated on data from the Fetal Age Machine Learning Initiative (FAMLI) cohort, which included participants from 2 study sites at Chapel Hill, North Carolina (US), and Lusaka, Zambia. Participants were eligible to be part of this study if they received routine antenatal care at 1 of these sites, were aged 18 years or older, had a viable intrauterine singleton pregnancy, and could provide written consent. They were not eligible if they had known uterine or fetal abnormality, or had any other conditions that would make participation unsafe or complicate interpretation. Data analysis was performed from January to July 2022. Main Outcomes and Measures: The primary analysis outcome for GA was the mean difference in absolute error between the GA model estimate and the clinical standard estimate, with the ground truth GA extrapolated from the initial GA estimated at an initial examination. Results: Of the total cohort of 3842 participants, data were calculated for a test set of 404 participants with a mean (SD) age of 28.8 (5.6) years at enrollment. All models were statistically superior to standard fetal biometry-based GA estimates derived from images captured by expert sonographers. The ensemble model had the lowest mean absolute error compared with the clinical standard fetal biometry (mean [SD] difference, -1.51 [3.96] days; 95% CI, -1.90 to -1.10 days). All 3 models outperformed standard biometry by a more substantial margin on fetuses that were predicted to be small for their GA. Conclusions and Relevance: These findings suggest that AI models have the potential to empower trained operators to estimate GA with higher accuracy.

A mobile-optimized artificial intelligence system for gestational age and fetal malpresentation assessment.

Background: Fetal ultrasound is an important component of antenatal care, but shortage of adequately trained healthcare workers has limited its adoption in low-to-middle-income countries. This study investigated the use of artificial intelligence for fetal ultrasound in under-resourced settings. Methods: Blind sweep ultrasounds, consisting of six freehand ultrasound sweeps, were collected by sonographers in the USA and Zambia, and novice operators in Zambia. We developed artificial intelligence (AI) models that used blind sweeps to predict gestational age (GA) and fetal malpresentation. AI GA estimates and standard fetal biometry estimates were compared to a previously established ground truth, and evaluated for difference in absolute error. Fetal malpresentation (non-cephalic vs cephalic) was compared to sonographer assessment. On-device AI model run-times were benchmarked on Android mobile phones. Results: Here we show that GA estimation accuracy of the AI model is non-inferior to standard fetal biometry estimates (error difference -1.4 ± 4.5 days, 95% CI -1.8, -0.9, n = 406). Non-inferiority is maintained when blind sweeps are acquired by novice operators performing only two of six sweep motion types. Fetal malpresentation AUC-ROC is 0.977 (95% CI, 0.949, 1.00, n = 613), sonographers and novices have similar AUC-ROC. Software run-times on mobile phones for both diagnostic models are less than 3 s after completion of a sweep. Conclusions: The gestational age model is non-inferior to the clinical standard and the fetal malpresentation model has high AUC-ROCs across operators and devices. Our AI models are able to run on-device, without internet connectivity, and provide feedback scores to assist in upleveling the capabilities of lightly trained ultrasound operators in low resource settings.

Nuclear localization signal sequence is required for VACM-1/CUL5-dependent regulation of cellular growth.

VACM-1/CUL5 is a member of the cullin family of proteins involved in the E3 ligase-dependent degradation of diverse proteins that regulate cellular proliferation. The ability of VACM-1/CUL5 to inhibit cellular growth is affected by its posttranslational modifications and its localization to the nucleus. Since the mechanism of VACM-1/CUL5 translocation to the nucleus is not clear, the goal of this project was to determine the role that the putative nuclear localization signal (NLS) we identified in the VACM-1/CUL5 (640PKLKRQ646) plays in the cellular localization of VACM-1/CUL5 and its effect on cellular growth. We used site-directed mutagenesis to change Lys642 and Lys644 to Gly and the mutated cDNA constructs were transfected into COS-1 cells. Mutation of the NLS in VACM-1/CUL5 significantly reduced its localization to the nucleus and compromised its effect on cellular growth. We have shown previously that the antiproliferative effect of VACM-1/CUL5 could be reversed by mutation of PKA-specific phosphorylation sequence (S730AVACM-1/CUL5), which was associated with its increased nuclear localization and modification by NEDD8. Thus, we examined whether these properties can be controlled by the NLS. The mutation of NLS in S730AVACM-1/CUL5 cDNA compromised its proliferative effect and reduced its localization to the nucleus. The immunocytochemistry results showed that, in cells transfected with the mutant cDNAs, the nuclear NEDD8 signal was decreased. Western blot analysis of total cell lysates, however, showed that VACM-1/CUL5 neddylation was not affected. Together, these results suggest that the presence of the NLS, both in VACM-1/CUL5 and in S730AVACM-1/CUL5 sequences, is critical for their control of cell proliferation.

Mutational analysis of VACM-1/cul5 exons in cancer cell lines.

VACM-1, a cul-5 gene product, functions via an E3 ligase complex and when overexpressed, has an antiproliferative effect in many cell types. Overexpression of VACM-/cul5 cDNA mutated at the PKA-specific phosphorylation site at Ser730 reversed this phenotype. These effects are associated with the appearance of larger M(r) species subsequently identified as a Nedd8-modified VACM-1/cul5. Although decreased levels of VACM-1 mRNA detected in several cancers and cancer cell lines may explain the progression of cell growth, possible genetic and epigenetic changes in its sequence have not been analyzed. We hypothesized that in rapidly proliferating cells, VACM-1/cul5 may be mutated at either the PKA-specific phosphorylation site or the consensus neddylation site. We used RT-PCR and PCR, to amplify and to sequence mRNA and genomic DNA, respectively. To date we have sequenced all 19 coding exons of the VACM-1/cul5 gene in T47D breast cancer cells, U138MG glioma cells, ACHN renal cancer cells, and OVCAR-3 ovarian cancer cells. Our results indicate that in those cells VACM-1/cul5 is not mutated at the putative phosphorylation or the neddylation site. We have found one silent mutation in the genomic DNA isolated from U138MG, ACHN, and OVCAR-3 cell lines, but not from T47D cells. Our work suggests that in T47D breast cancer cells biologic activity of VACM-1/cul5 may be regulated by posttranslational modifications.