Towards an AI-based assessment model of surgical difficulty during early phase laparoscopic cholecystectomy

Abbing, Julian Ronald; Voskens, Frank J.; Gerats, Beerend; Egging, Ruby Mae; Milletarì, Fausto; Broeders, I. A. M. J.

doi:10.1080/21681163.2022.2163296

Cited by 3 publications

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In general, semantic segmentation enables the identification of critical structures and the delimitation of safe surgical zones, encourages the development of robotic assistance applications for cooperative work between the surgeon and the robot, facilitates the evaluation of the complexity of operations for an effective assignment of surgeries based on the complexity of the surgical procedure and the experience of the surgeons, and, through preprocessing of visual information, improves the visualization of structures hidden by surgical instruments, among other applications [ 62 , 63 , 64 ].…”

Section: Discussionmentioning

confidence: 99%

Improving Surgical Scene Semantic Segmentation through a Deep Learning Architecture with Attention to Class Imbalance

Urrea,

Garcia-Garcia,

Kern

2024

Biomedicines

View full text Add to dashboard Cite

This article addresses the semantic segmentation of laparoscopic surgery images, placing special emphasis on the segmentation of structures with a smaller number of observations. As a result of this study, adjustment parameters are proposed for deep neural network architectures, enabling a robust segmentation of all structures in the surgical scene. The U-Net architecture with five encoder–decoders (U-Net5ed), SegNet-VGG19, and DeepLabv3+ employing different backbones are implemented. Three main experiments are conducted, working with Rectified Linear Unit (ReLU), Gaussian Error Linear Unit (GELU), and Swish activation functions. The applied loss functions include Cross Entropy (CE), Focal Loss (FL), Tversky Loss (TL), Dice Loss (DiL), Cross Entropy Dice Loss (CEDL), and Cross Entropy Tversky Loss (CETL). The performance of Stochastic Gradient Descent with momentum (SGDM) and Adaptive Moment Estimation (Adam) optimizers is compared. It is qualitatively and quantitatively confirmed that DeepLabv3+ and U-Net5ed architectures yield the best results. The DeepLabv3+ architecture with the ResNet-50 backbone, Swish activation function, and CETL loss function reports a Mean Accuracy (MAcc) of 0.976 and Mean Intersection over Union (MIoU) of 0.977. The semantic segmentation of structures with a smaller number of observations, such as the hepatic vein, cystic duct, Liver Ligament, and blood, verifies that the obtained results are very competitive and promising compared to the consulted literature. The proposed selected parameters were validated in the YOLOv9 architecture, which showed an improvement in semantic segmentation compared to the results obtained with the original architecture.

show abstract

Section: Discussionmentioning

confidence: 99%

Improving Surgical Scene Semantic Segmentation through a Deep Learning Architecture with Attention to Class Imbalance

Urrea,

Garcia-Garcia,

Kern

2024

Biomedicines

View full text Add to dashboard Cite

show abstract

“…Here, widespread adoption of electronic health records and advances in artificial intelligence could be part of the solution. Large language models and other artificial intelligence algorithms may be able to unlock information about patient risk and surgical complexity from the electronic health record, allowing for a more objective assessment of complexity, which could be translated into compensation models once properly evaluated …”

mentioning

confidence: 99%

“…Large language models and other artificial intelligence algorithms may be able to unlock information about patient risk and surgical complexity from the electronic health record, allowing for a more objective assessment of complexity, which could be translated into compensation models once properly evaluated. 4,5 As new risk adjustment tools are developed and implemented, it is key to engender trust and mitigate risks of bias by heeding the authors' suggestion for increased transparency from all parties. While surgeons deserve adequate compensation for complex operations, payers need to be confident that they will only be billed for indicated care, and most importantly, patients deserve certainty that incentives for surgical care are aligned with improving their health.…”

mentioning

confidence: 99%

Modifier 22—Is It All Just a Catch-22?

Abbasi,

Wick

2024

JAMA Surg

View full text Add to dashboard Cite

The study by Childers et al 1 provides insights into the flaws of fee-for-service payment systems when it comes to the surgical care of complex patients. The study found that the magnitude of additional payments associated with the use of modifier 22 was small, and this was almost completely offset by increased claim denials. These results are sure to be disappointing to surgeons who frequently use modifier 22 believing that it will lead to additional compensation related to the time and stress associated with more challenging cases. The findings highlight the ongoing challenge with reliably rewarding health care professionals for delivering appropriate, high-quality care to more complex patients. While models have been proposed and even tested, most have failed because it is difficult to objectively measure surgical complexity and risk.Childers and colleagues outline several alternatives to the generic modifier 22 approach that could be implemented in feefor-service models. But without better objective data assessing complexity, these changes will likely lead to wide variation in use and even potential abuse, which may contribute to payers' reluctance to provide additional payments. For example, the authors suggest coding for complexity levels, akin to the Diagnostic Related Groups system, which can lead to upcoding and creep in billed complexity over time. 2 In contrast to fee-for-service models, value-based bundled payment mod-

show abstract

Transfer learning and Machine Learning Classification for Laparoscopic Video Distortion Detection

Belmokeddem,

khemis,

Loudjedi

2024

2024 8th International Conference on Image and Signal Processing and Their Applications (ISPA)

View full text Add to dashboard Cite

Towards an AI-based assessment model of surgical difficulty during early phase laparoscopic cholecystectomy

Cited by 3 publications

References 13 publications

Improving Surgical Scene Semantic Segmentation through a Deep Learning Architecture with Attention to Class Imbalance

Improving Surgical Scene Semantic Segmentation through a Deep Learning Architecture with Attention to Class Imbalance

Modifier 22—Is It All Just a Catch-22?

Transfer learning and Machine Learning Classification for Laparoscopic Video Distortion Detection

Contact Info

Product

Resources

About