Artificial Intelligence in Breast Surgery: History
View Latest Version
Please note this is an old version of this entry, which may differ significantly from the current revision.
Subjects: Surgery
Contributor:
Ishith Seth
,
Gabriella Bulloch
,
Konrad Joseph
,
David J. Hunter-Smith
,
Warren Matthew Rozen

Breast reconstruction is a pivotal part of the recuperation process following a mastectomy and aims to restore both the physical aesthetic and emotional well-being of breast cancer survivors. In recent years, artificial intelligence (AI) has emerged as a revolutionary technology across numerous medical disciplines. This narrative review of the current literature and evidence analysis explores the role of AI in the domain of breast reconstruction, outlining its potential to refine surgical procedures, enhance outcomes, and streamline decision making. A systematic search on Medline (via PubMed), Cochrane Library, Web of Science, Google Scholar, Clinical Trials, and Embase databases from January 1901 to June 2023 was conducted. By meticulously evaluating a selection of recent studies and engaging with inherent challenges and prospective trajectories, this review spotlights the promising role AI plays in advancing the techniques of breast reconstruction. However, issues concerning data quality, privacy, and ethical considerations pose hurdles to the seamless integration of AI in the medical field. The future research agenda comprises dataset standardization, AI algorithm refinement, and the implementation of prospective clinical trials and fosters cross-disciplinary partnerships. The fusion of AI with other emergent technologies like augmented reality and 3D printing could further propel progress in breast surgery.

  • artificial intelligence
  • AI
  • breast reconstruction
  • breast surgery
  • advancement

1. AI and Detection of Breast Tumors from Imaging

The primary functions of AI in breast cancer screening currently involve object detection and tumor classification as benign or malignant according to the Breast Imaging Reporting and Data System (BIRADS), which was traditionally performed by human experts. These approaches underscore the comprehensive role AI plays in improving disease detection and treatment outcomes (Figure 1). In a multi-center study conducted by Hamyoon et al., ML was utilized to examine 1288 breast lesions based on BIRADS and morphometric features [1]. The model achieved an area under the receiver operating characteristic curve (AUC) of 0.885, outperforming both expert radiologists and radiology residents who achieved AUCs of 0.814 and 0.632, respectively, across all cohorts.
Radiomics is an approach to medical imaging that relies on AI to analyse quantitative information extracted from images [2]. It operates on the premise that these extracted features reflect activities that occur at the genetic and molecular levels. Radiomics is classified as supervised or unsupervised. Supervised ML starts with training AI using pre-existing data archives, whereas unsupervised ML categorizes information without reference to pre-existing data or data derived from the image itself. DL models recognize and classify images by autonomously reducing each to a set of numeric features that are processed using a multilayer neural network. Kallenberg et al. were among the early adopters of DL for breast cancer risk assessment. They used a convolutional sparse autoencoder coupled with a simple classifier, successfully linking the identified features to breast cancer in a significant case–control classification performance [3]. This was achieved by training and testing the model on contralateral mammographic images from two distinct databases. Concurrently, Li et al. effectively differentiated between high-risk groups and healthy controls by applying a pre-trained AlexNet model to mammographic images [4]. In another innovative approach, Gastounioti et al. harnessed convolutional neural networks to merge parenchymal complexity measurements into distinct meta-features for risk prediction [5]. Their method outperformed conventional parenchymal pattern analysis, further underscoring the potential of such groundbreaking techniques to enhance breast cancer risk prediction. These studies present early evidence that full-field digital-mammography-based DL models may prove more accurate than traditional density-based and epidemiology-based models in predicting breast cancer risk.
In 2017, Becker et al. performed a retrospective study using neural network image analysis software for mammography diagnostics. Their model was tested on two datasets (n = 178) and demonstrated comparable performance with radiologists but with better sensitivity [6]. In 2020, a study by McKinney et al. evaluated an AI breast cancer screening system on an international scale [7]. They found that using it for breast cancer screening improved prediction performance by reducing both type 1 and type 2 error rates whilst simultaneously reducing physician workload. In 2021, Raya-Povedano et al. performed a retrospective evaluation of AI-based breast cancer screening strategies. In an examination of 15,987 mammograms, they found that AI system implementation would simultaneously increase cancer detection sensitivity and reduce human workload [8]. These findings support the theoretical benefits of implementing AI in the detection of breast lesions and, therefore, have significant implications for breast reconstruction procedures.

2. Preoperative Use of AI

Once a breast lesion has been detected, AI algorithms can assist surgeons in preoperative planning by assessing patient-specific factors, such as breast volume, shape, and symmetry. Preoperative imaging studies are essential for assessing the vascular supply of the breast and can determine the reliability of reconstructive techniques [9]. Specific ML technologies, like the Faster-RCNN with Inception-ResNet-v2 deep-learning framework for ultrasound breast images, have further optimized these processes and show potential for use in surgical interventions.
Research highlighting the benefits of AI support in the preoperative period of breast reconstruction has been performed. In 2020, Mavioso et al. evaluated the feasibility of using computer software to support preoperative planning for microsurgical reconstruction using the DIEP flap technique. The researchers developed a convolutional neural network to identify perforators and tested the software on 40 patients. Results showed that the software was able to detect key perforators in 97.7% of cases, further supporting the idea that ML techniques can improve the preoperative planning of breast reconstruction [10].
AI can also be used to simulate results during the preoperative period of breast surgery, providing clinicians and patients alike valuable insights into likely cosmetic outcomes. In 2022, Chartier et al. evaluated a neural network that, after training itself on real clinical images, was able to consistently generate preoperative images that were comparable to real surgical results [11].

3. Intraoperative Use of AI

Within the confines of the operating theater, AI has the potential to support surgeons via real-time decision making and enhance surgical precision, but for the present, these have not been explored with respect to breast reconstruction surgery. In endoscopic surgeries, computer vision algorithms could scrutinize intraoperative imagery, highlighting crucial anatomical structures and offering guidance for meticulous tissue dissection and implant placement. Additionally, in orthopedic surgery, AI-powered robotic systems augmented surgical dexterity, reducing the risk of complications and improving surgical outcomes [12]. Although autonomous systems have achieved clinical application in orthopedic and neurosurgical fields [13], the role of AI-assisted robotics for surgical intervention on deformable breast tissue remains theoretical at present.

4. Postoperative Use of AI

AI-based systems can help optimize postoperative care via the early detection of complications. By carrying out clinical monitoring and the evaluation of medical imagery, AI algorithms can autonomously detect the symptoms of infection, hematoma, or mispositioned implants. This would in turn facilitate timely intervention and decrease the likelihood of enduring complications. In 2021, Myung et al. validated the ability of machine learning models and identified factors that predict breast reconstruction complications in 568 cases. They found that AI technologies could be effectively used for assessing the risk of negative patient outcomes in reconstructive breast surgeries [14].
AI has also been used to facilitate comprehensive patient assessment and help forecast postoperative pain [15]. In 2020, Nair et al. developed an ML system that was able to predict postoperative opioid requirements in patients undergoing ambulatory surgery, further optimizing postoperative care [15]. These systems offer significant improvements over conventional methods, such as clinician-collected questionnaires, and provide advantages within a healthcare system where time and human resources are often constrained. By improving the efficiency and quality of postoperative pain assessments and prediction models, AI has the potential to help prevent debilitating post-mastectomy pain syndromes in patients who undergo breast surgery. In 2020, Juwara et al. tested the ability of AI to predict neuropathic pain after breast cancer surgery. After testing on a cohort of 204 patients, the authors concluded that ML models can identify predictors for neuropathic pain in patients who have undergone breast surgery better than traditional methods [16].
Kenig et al. recently developed an ad hoc neural network to identify key breast features for breast symmetry evaluation. The study used 200 frontal photographs of patients who underwent breast surgery and tested the neural network on 47 frontal images of patients who underwent breast reconstruction after breast cancer. The neural network was successful at localizing key breast features, including breast boundaries and the nipple–areolar complex in 97.7% of cases, with an impressive mean detection time of 0.52 s [17]. By automating the detection of key breast features, AI could revolutionize the evaluation of breast symmetry post-reconstruction [17].
 

This entry is adapted from the peer-reviewed paper 10.3390/jcm12155143

References

  1. Hamyoon, H.; Chan, W.Y.; Mohammadi, A.; Kuzan, T.Y.; Mirza-Aghazadeh-Attari, M.; Leong, W.L.; Altintoprak, K.M.; Vijayananthan, A.; Rahmat, K.; Mumin, N.A.; et al. Artificial intelligence, BI-RADS evaluation and morphometry: A novel combination to diagnose breast cancer using ultrasonography, results from multi-center cohorts. Eur. J. Radiol. 2022, 157, 110591.
  2. van Timmeren, J.; Cester, D.; Tanadini-Lang, S.; Alkadhi, H.; Baessler, B. Radiomics in medical imaging—“How-to” guide and critical reflection. Insights Imaging 2020, 11, 91.
  3. Kallenberg, M.; Petersen, K.; Nielsen, M.; Ng, A.Y.; Diao, P.; Igel, C.; Vachon, C.M.; Holland, K.; Winkel, R.R.; Karssemeijer, N.; et al. Unsupervised Deep Learning Applied to Breast Density Segmentation and Mammographic Risk Scoring. IEEE Trans. Med. Imaging 2016, 35, 1322–1331.
  4. Li, S.; Wei, J.; Chan, H.P.; Helvie, M.A.; Roubidoux, M.A.; Lu, Y.; Zhou, C.; Hadjiiski, L.M.; Samala, R.K. Computer-aided assessment of breast density: Comparison of supervised deep learning and feature-based statistical learning. Phys. Med. Biol. 2018, 63, 025005.
  5. Gastounioti, A.; Pantalone, L.; Scott, C.G.; Cohen, E.A.; Wu, F.F.; Winham, S.J.; Jensen, M.R.; Maidment, A.D.A.; Vachon, C.M.; Conant, E.F.; et al. Fully Automated Volumetric Breast Density Estimation from Digital Breast Tomosynthesis. Radiology 2021, 301, 561–568.
  6. Becker, A.S.; Marcon, M.; Ghafoor, S.; Wurnig, M.C.; Frauenfelder, T.; Boss, A. Deep Learning in Mammography: Diagnostic Accuracy of a Multipurpose Image Analysis Software in the Detection of Breast Cancer. Investig. Radiol. 2017, 52, 434–440.
  7. McKinney, S.M.; Sieniek, M.; Godbole, V.; Godwin, J.; Antropova, N.; Ashrafian, H.; Back, T.; Chesus, M.; Corrado, G.S.; Darzi, A.; et al. International evaluation of an AI system for breast cancer screening. Nature 2020, 577, 89–94.
  8. Raya-Povedano, J.L.; Romero-Martín, S.; Elías-Cabot, E.; Gubern-Mérida, A.; Rodríguez-Ruiz, A.; Álvarez-Benito, M. AI-based strategies to reduce workload in breast cancer screening with mammography and tomosynthesis: A retrospective evaluation. Radiology 2021, 300, 57–65.
  9. Vegas, M.R.; Martina, L.; Segovia-Gonzalez, M.; Garcia-Garcia, J.F.; Gonzalez-Gonzalez, A.; Mendieta-Baro, A.; Nieto-Gongora, C.; Benito-Duque, P. Vascular anatomy of the breast and its implications in the breast-sharing reconstruction technique. J. Plast. Reconstr. Aesthet. Surg. 2023, 76, 180–188.
  10. Mavioso, C.; Araújo, R.J.; Oliveira, H.P.; Anacleto, J.C.; Vasconcelos, M.A.; Pinto, D.; Gouveia, P.F.; Alves, C.; Cardoso, F.; Cardoso, J.S.; et al. Automatic detection of perforators for microsurgical reconstruction. Breast 2020, 50, 19–24.
  11. Chartier, C.; Watt, A.; Lin, O.; Chandawarkar, A. BreastGAN: Artificial Intelligence-Enabled Breast Augmentation Simulation. Aesthetic Surg. J. 2022, 4, ojab052.
  12. Saeidi, H.; Le, H.; Opfermann, J.; Leonard, S.; Kim, A.; Hsieh, M.H.; Kang, J.U.; Krieger, A. Autonomous Laparoscopic Robotic Suturing with a Novel Actuated Suturing Tool and 3D Endoscope. In Proceedings of the 2019 International Conference on Robotics and Automation (ICRA), Montreal, QC, Canada, 20–24 May 2019.
  13. Taylor, R.H.; Mittelstadt, B.D.; Paul, H.A.; Hanson, W.; Kazanzides, P.; Zuhars, J.F.; Williamson, B.; Musits, B.L.; Glassman, E.; Bargar, W.L.; et al. An image-directed robotic system for precise orthopaedic surgery. IEEE Trans. Robot. Autom. 1994, 10, 261–275.
  14. Myung, Y.; Jeon, S.; Heo, C.; Kim, E.-K.; Kang, E.; Shin, H.-C.; Yang, E.-J.; Jeong, J.H. Validating machine learning approaches for prediction of donor related complication in microsurgical breast reconstruction: A retrospective cohort study. Sci. Rep. 2021, 11, 5615.
  15. Nair, A.A.; Velagapudi, M.A.; Lang, J.A.; Behara, L.; Venigandla, R.; Velagapudi, N.; Fong, C.T.; Horibe, M.; Lang, J.D.; Nair, B.G. Machine learning approach to predict postoperative opioid requirements in ambulatory surgery patients. PLoS ONE 2020, 15, e0236833.
  16. Juwara, L.; Arora, N.; Gornitsky, M.; Saha-Chaudhuri, P.; Velly, A.M. Identifying predictive factors for neuropathic pain after breast cancer surgery using machine learning. Int. J. Med. Inform. 2020, 141, 104170.
  17. Kenig, N.; Monton Echeverria, J.; De la Ossa, L. Identification of key breast features using a neural network: Applications of machine learning in the clinical setting of Plastic Surgery. Plast. Reconstr. Surg. 2023.
More
© Text is available under the terms and conditions of the Creative Commons Attribution (CC BY) license; additional terms may apply. By using this site, you agree to the Terms and Conditions and Privacy Policy.
This entry is offline, you can click here to edit this entry!
Video Production Service
Feedback