Page 35 - Read Online

Page 303 Brenac et al. Art Int Surg 2024;4:296-315 https://dx.doi.org/10.20517/ais.2024.49

Table 1. Benefits and challenges of AI in pre- and postoperative assessments
Plastic surgery
Phase AI-based solutions Benefits of AI Challenges of AI
current limitations
Preoperative - Imaging and diagnostic 3D imaging and modeling - High precision and - High cost
- Surgical planning AI algorithms process patient data to create customization - Requires extensive training
- Personalized detailed 3D models for surgical planning - Enhanced visualization - Data privacy concerns
visualization - Improved patient-
- Predictive outcomes surgeon communication
AR - Real-time guidance - Technical limitations
AR overlays digital information onto the - Supports minimally - Integration challenges
real-world surgical field, enhancing invasive techniques - High cost
visualization and precision - Valuable educational
tool
Predictive analytics - Early identification of - Dependent on data quality
ML models analyze patient data to predict risks - Potential for algorithmic bias
potential complications and outcomes - Personalized surgical to training data
plans - Integration complexity
- Optimized resource
allocation
Breast, facial, hand, and wound healing - Reduced planning time - Accuracy issues with small
(Skin) assessments - Enhanced monitoring vessels
AI aids in selecting reconstructive methods, - Improved satisfaction - Discrepancies in vertical
preoperative planning, and evaluating and reduced component estimation for
imaging data readmissions breast reconstruction
Postoperative - Managing surgical Telemedicine and remote monitoring - Continuous support - Technology barriers
complications and patient AI-driven platforms monitor patient - Increased accessibility - Data security concerns
information recovery remotely, ensuring continuous - Improved adherence to - Limited physical examination
- Variability in outcomes communication and support protocols
- Subjective results and Predictive analytics - Timely intervention - Data dependency
evaluation
AI models continue to predict - Reduced morbidity and - Algorithmic bias to training
- Patient satisfaction
complications based on ongoing patient mortality data
data, facilitating early interventions - Personalized care - Complexity in clinical
workflow integration
AI-enhanced readability of patient - Increased patient - Need for further refinement
education materials understanding - Ensuring personalized advice
AI tools simplify medical information to - Better adherence to - Patient trust and reliability
improve patient comprehension and recovery protocols issues
adherence to postoperative care - Improved outcomes
instructions
Breast, facial, hand, and wound healing - Early detection of - Technical limitations
(skin) monitoring complications - Dependence on image
Smartphone apps and ML algorithms - Improved monitoring quality and data input
monitor status and predict complications - Higher predictive
like infections and functional recovery accuracy

AI: Artificial intelligence; AR: augmented reality; ML: machine learning.

algorithm was able to accurately predict complications [area under the curve (AUC): 0.89] and could
[36]
further be used as a reference for assessing the individual risk associated with abdominal flaps .

Additionally, ML has been applied to minimize postoperative infection following implant-based
reconstruction, including the development of internal algorithms to guide clinical decisions such as the
need for reoperation or introduction of antibiotics . Hassan et al. developed an algorithm using ML to
[1]
predict periprosthetic infection and explantation . The study demonstrated that ML models can provide a
[37]
higher predictive accuracy compared to multivariate logistic regression for periprosthetic infection and
explantation [37,38] . Therefore, ML could help reduce postoperative burden and promote better outcomes in
breast reconstruction.

Facial surgery
In the context of facial surgery, photographic data provides a means to assess surgical success or failure, and