Automated Quantification of Blood Loss versus Visual Estimation in 274 Vaginal Deliveries

Rubenstein, Andrew F.; Zamudio, Stacy; Douglas, Claudia; Sledge, Sharon; Thurer, Robert L.

doi:10.1055/s-0040-1701507

Cited by 12 publications

(8 citation statements)

References 18 publications

(23 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This post-hoc cutoff of 500 mL represents a potential clinical marker of need for transfusion. The tendency for human experts to underestimate blood loss is well documented 34 – 37 , corroborated by our findings, and may result in delayed recognition of life-threatening hemorrhage.…”

Section: Discussionsupporting

confidence: 84%

Expert surgeons and deep learning models can predict the outcome of surgical hemorrhage from 1 min of video

Pangal

Kugener

Zhu

et al. 2022

Sci Rep

View full text Add to dashboard Cite

Major vascular injury resulting in uncontrolled bleeding is a catastrophic and often fatal complication of minimally invasive surgery. At the outset of these events, surgeons do not know how much blood will be lost or whether they will successfully control the hemorrhage (achieve hemostasis). We evaluate the ability of a deep learning neural network (DNN) to predict hemostasis control ability using the first minute of surgical video and compare model performance with human experts viewing the same video. The publicly available SOCAL dataset contains 147 videos of attending and resident surgeons managing hemorrhage in a validated, high-fidelity cadaveric simulator. Videos are labeled with outcome and blood loss (mL). The first minute of 20 videos was shown to four, blinded, fellowship trained skull-base neurosurgery instructors, and to SOCALNet (a DNN trained on SOCAL videos). SOCALNet architecture included a convolutional network (ResNet) identifying spatial features and a recurrent network identifying temporal features (LSTM). Experts independently assessed surgeon skill, predicted outcome and blood loss (mL). Outcome and blood loss predictions were compared with SOCALNet. Expert inter-rater reliability was 0.95. Experts correctly predicted 14/20 trials (Sensitivity: 82%, Specificity: 55%, Positive Predictive Value (PPV): 69%, Negative Predictive Value (NPV): 71%). SOCALNet correctly predicted 17/20 trials (Sensitivity 100%, Specificity 66%, PPV 79%, NPV 100%) and correctly identified all successful attempts. Expert predictions of the highest and lowest skill surgeons and expert predictions reported with maximum confidence were more accurate. Experts systematically underestimated blood loss (mean error − 131 mL, RMSE 350 mL, R2 0.70) and fewer than half of expert predictions identified blood loss > 500 mL (47.5%, 19/40). SOCALNet had superior performance (mean error − 57 mL, RMSE 295 mL, R2 0.74) and detected most episodes of blood loss > 500 mL (80%, 8/10). In validation experiments, SOCALNet evaluation of a critical on-screen surgical maneuver and high/low-skill composite videos were concordant with expert evaluation. Using only the first minute of video, experts and SOCALNet can predict outcome and blood loss during surgical hemorrhage. Experts systematically underestimated blood loss, and SOCALNet had no false negatives. DNNs can provide accurate, meaningful assessments of surgical video. We call for the creation of datasets of surgical adverse events for quality improvement research.

show abstract

Section: Discussionsupporting

confidence: 84%

Expert surgeons and deep learning models can predict the outcome of surgical hemorrhage from 1 min of video

Pangal

Kugener

Zhu

et al. 2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…This post-hoc cutoff of 500mL represents a potential clinical marker of need for transfusion. The tendency for human experts to underestimate blood loss is well documented, [33][34][35][36] corroborated by our findings, and may result in delayed recognition of life-threatening hemorrhage.…”

Section: Benchmark Performance Of Human Expertssupporting

confidence: 84%

Expert Surgeons and Deep Learning Models Can Predict the Outcome of Surgical Hemorrhage from One Minute of Video

Pangal

Kugener

Zhu

et al. 2022

Preprint

View full text Add to dashboard Cite

Background: Major vascular injury resulting in uncontrolled bleeding is a catastrophic and often fatal complication of minimally invasive surgery. At the outset of these events, surgeons do not know how much blood will be lost or whether they will successfully control the hemorrhage (achieve hemostasis). We evaluate the ability of a deep learning neural network (DNN) to predict hemostasis control ability using the first minute of surgical video and compare model performance with human experts viewing the same video. Methods: The publicly available SOCAL dataset contains 147 videos of attending and resident surgeons managing hemorrhage in a validated, high-fidelity cadaveric simulator. Videos are labeled with outcome and blood loss (mL). The first minute of 20 videos was shown to four, blinded, fellowship trained skull-base neurosurgery instructors, and to SOCALNet (a DNN trained on SOCAL videos). SOCALNet architecture included a convolutional network (ResNet) identifying spatial features and a recurrent network identifying temporal features (LSTM). Experts independently assessed surgeon skill, predicted outcome and blood loss (mL). Outcome and blood loss predictions were compared with SOCALNet. Results: Expert inter-rater reliability was 0.95. Experts correctly predicted 14/20 trials (Sensitivity: 82%, Specificity: 55%, Positive Predictive Value (PPV): 69%, Negative Predictive Value (NPV): 71%). SOCALNet correctly predicted 17/20 trials (Sensitivity 100%, Specificity 66%, PPV 79%, NPV 100%) and correctly identified all successful attempts. Expert predictions of the highest and lowest skill surgeons and expert predictions reported with maximum confidence were more accurate. Experts systematically underestimated blood loss (mean error -131 mL, RMSE 350 mL, R2 0.70) and fewer than half of expert predictions identified blood loss > 500mL (47.5%, 19/40). SOCALNet had superior performance (mean error -57 mL, RMSE 295mL, R2 0.74) and detected most episodes of blood loss > 500mL (80%, 8/10). In validation experiments, SOCALNet evaluation of a critical on-screen surgical maneuver and high/low-skill composite videos were concordant with expert evaluation. Conclusion: Using only the first minute of video, experts and SOCALNet can predict outcome and blood loss during surgical hemorrhage. Experts systematically underestimated blood loss, and SOCALNet had no false negatives. DNNs can provide accurate, meaningful assessments of surgical video. We call for the creation of datasets of surgical adverse events for quality improvement research.

show abstract

“…A direct comparison of EBL and Triton QBL at 274 vaginal deliveries demonstrated slightly lower blood loss values with EBL compared with QBL (median [IQR] 300 [200-350] mL vs. 339 [217-515] mL and a higher proportion of PPH defined as blood loss >1 L (4.3% vs. 0%) or >500 mL (26.6 vs. 5.1%, respectively). 48 While there was no verification of accuracy, this study suggests that QBL may assist in the earlier recognition of PPH. With a retrospective interrupted time series design, introduction of the vaginal delivery QBL system, including a calibrated V-drape and gravimetry, resulted in the increased odds of PPH detection, from 11.5% to 26.8% incidence (defined as QBL ≥500 mL) after vaginal delivery.…”

Section: New Advances: Hybrid Systems and Colorimetrymentioning

confidence: 61%

Can measuring blood loss at delivery reduce hemorrhage-related morbidity?

Katz

Farber

2021

International Journal of Obstetric Anesthesia

View full text Add to dashboard Cite

Automated Quantification of Blood Loss versus Visual Estimation in 274 Vaginal Deliveries

Abstract: Key Points

Cited by 12 publications

References 18 publications

Expert surgeons and deep learning models can predict the outcome of surgical hemorrhage from 1 min of video

Expert surgeons and deep learning models can predict the outcome of surgical hemorrhage from 1 min of video

Expert Surgeons and Deep Learning Models Can Predict the Outcome of Surgical Hemorrhage from One Minute of Video

Can measuring blood loss at delivery reduce hemorrhage-related morbidity?

Contact Info

Product

Resources

About