|
 
 |
| REVIEW ARTICLE |
|
| Year : 2022 | Volume
: 8
| Issue : 1 | Page : 4 |
|
Progress in clinical application of artificial intelligence in orthopedics
Yiwei Wang1, Ruiqi Li2, Pengfei Zheng1
1 Department of Pediatric Orthopaedics, Children's Hospital of Nanjing Medical University, Nanjing, China
2 School of Medicine, Southeast University, Nanjing, China
| Date of Submission | 04-May-2021 |
| Date of Decision | 21-Jun-2021 |
| Date of Acceptance | 22-Jul-2021 |
| Date of Web Publication | 03-Mar-2022 |
Correspondence Address:
Pengfei Zheng
Department of Pediatric Orthopaedics, Nanjing Children's Hospital Affiliated to Nanjing Medical University, Nanjing
China
 Source of Support: None, Conflict of Interest: None  | Check |
DOI: 10.4103/digm.digm_10_21
Ever since the concept of artificial intelligence (AI) has been suggested, it has undergone years of research and development. Under the current condition of rapid development of information and data technology, AI has shown significant effective value and application capability in multiple fields, especially in medical treatment. AI has become essential for routine medical treatment. This review summarizes the current clinical application of AI in orthopedics, in reference to the basic principle of AI, AI supported in clinical diagnosis, AI supported in clinical decision-making, AI supported clinical surgery, and the combination of AI and telemedicine. At the same time, this review also specifies the advantages, disadvantages, and capability of AI in the current clinical application, to provide some understanding for further research of AI.
Keywords: Artificial intelligence, Deep learning, Machine learning, Orthopedics
How to cite this article:
Wang Y, Li R, Zheng P. Progress in clinical application of artificial intelligence in orthopedics. Digit Med 2022;8:4 |
| Introduction | |  |
Artificial intelligence (AI) was first suggested by Professor McCarthy et al. in 1956 at the Dartmouth Society. Researchers have formed many theories and standards since its beginning, and the field of AI has expanded. It is a new technical science for the study, development and formation of theories, methods, techniques for the simulation, and extension of human intelligence.[1],[2],[3],[4] Today, with the fast advancement of technology and the exponential growth of big data, AI has transformed from a simple theory to an extraordinary field. At present, AI is widely applicable. It has been commonly used in manufacturing, agriculture, logistics, finance, commerce, home, and so on. AI also shows great potential in medical applications.[5],[6] In 1977, Anderson predicted that AI would bring about a “postdoctorate era” in the 21st century.[7] The latest advancements in AI show that the development of the most recent medical treatment includes the promotion and application of new AI treatment models and new means, and the formation of a rapid and perceptive medical system. This system will include exploring the construction of smart hospitals, developing human- computer collaborative surgical robots, intelligent diagnosis and treatment assistants, developing flexible wearable, biocompatible physiological monitoring systems, developing human- computer collaborative clinical intelligent diagnosis and treatment solutions, intelligent image recognition, pathological classification and intelligent multidisciplinary consultation, etc.AI has strong learning capability and data processing ability, showing good application value and application potential in orthopedic clinical practice, and has shown its advantages in identification of images, assisting in diagnostic decision-making, assisting surgery, guiding exercise rehabilitation and other clinical stages, and successfully improving the medical field.
Using “AI” and “orthopaedics” as keywords, we searched for related articles published in PubMed from 1976 to 2021 and found a total of 1540 articles [Figure 1]. We found that the related articles had shown an increasing trend, especially in the past 5 years, the number of related articles increased rapidly. We can see that AI has become a hot topic in orthopedics. This is similar to the results of Guo et al., who searched the Web of Science for articles related to AI published as of December 2019 and found that the growth rate of articles from 2014 to 2019 was 45.15%, much greater than the average of 17.02%.[8] Along with the articles we searched, “machine learning (ML)” and “surgery” are the most common keywords, which indicate that the main research direction of AI in orthopedics is still ML and AI-assisted surgery. | Figure 1: Articles published in PubMed from 1976 to 2021 searched using “artificial intelligence” and “orthopaedics” as keywords.
Click here to view |
| Machine Learning, Deep Learning, And Artificial Intelligence | | |
AI is a broad field of research aimed at creating computer systems that represent human intelligence. ML is a subspecialty of AI, which creates algorithms that allow computers to deal with a new problem without being reprogrammed. In other words, ML systems solve problems independently by precisely “learning and processing” the data. It is recognized by using a statistical method to perform pattern recognition from the data provided, without the need for human guidance. Most ML algorithms can be viewed as mathematical models that map a set of “predictors” from various types of data or samples into a set of target outcome variables[9],[10],[11] [Figure 2]. | Figure 2: Taxonomy and overview of main machine learning algorithms.(a) Taxonomy of the different methods presented. (b) Overview of ML methods. The spectrum of available methods ranges from simpler and more interpretable to more advanced algorithms with potentially higher performance at the expense of less interpretability. Position of methods on the figure is qualitative and in practice depends on the number of free parameters, model complexity, data type, and the exact definition of interpretability used.8PCA, principal component analysis; SVM, support vector machine; tSNE, t - distributed stochastic neighbor embedding; UMAP, uniform manifold approximation and projection.[9]
Click here to view |
Deep learning is a new system of ML and is also a subspecialty of ML. Deep learning methods are based on artificial neural networks using millions of neuron-like units to learn the complex relationship between image pixel values and their semantic labels.[12] This kind of network relies on the complexity of the system, by modifying the interconnection relationship between a large number of internal nodes, so as to achieve the purpose of handling information, and has the ability of learning, self-organization, generalization, and training.[13],[14] In deep learning, features are often extracted by the model itself. At present, deep learning has been applied to orthopedic surgery and orthopedic imaging examinations such as spinal magnetic resonance imaging,[15] and it has also been shown that deep learning can predict the postoperative complications and mortality of the corresponding surgery, which has a better predictive ability than common prediction models.[16],[17] In the current era of medical information data, deep learning technology is undoubtedly in line with the current demand of the medical industry to deal with the development of medical data.
| Artificial Intelligence-Assisted Diagnosis | | |
Imaging examination to assist clinical diagnosis is one of the departmental characteristics of orthopedics. AI technology can improve the imaging path to a great extent, including the optimization of image acquisition, image reconstruction, and image analysis and interpretation.[18],[19],[20] Therefore, AI can effectively help orthopedic surgeons improve the accuracy of diagnosis, reduce the possibility of misdiagnosis, and reduce the burden on doctors. At present, AI systems have been applied to multiple scenarios, including assessment and identification of fractures, osteoarthritis, bone age assessment, and bone mineral density.[21],[22] Li et al. included 12,528 patients with standardized pelvic anteroposterior X-ray for AI image training, and selected the corresponding model, so as to realize AI automatic screening and calculation of Sharp's angle to assist in the diagnosis of developmental dysplasia of hip.[23] Through the comparison with multiple surgeons, the results showed that the AI model had reasonable diagnostic sensitivity and diagnostic consistency with orthopedic surgeons in diagnosing DHH according to Sharp's angle, but the measurement time required by the AI model was significantly lower than that required by the physicians. The results are consistent with the purpose of applying AI to help clinical decision-making and develop treatment strategies.
The medical system based on AI technology also shows its benefits in the process of orthopedic clinical diagnosis and treatment. As mentioned above, diagnosing bone tumors has many difficulties in diagnosis and prognosis due to their biological behavioral changings and high local recurrence. Relevant auxiliary diagnostic expert systems can be established using AI, thereby improving the accuracy and detection of rate of bone tumors, so as to achieve better prognostic objectives. Using the big data analysis ability and learning ability of AI, the emptiness of manual analysis correlation of hundreds of variables at the same time for diagnostic procedures can be effectively solved. On the basis of the original Scoliosis Research Society–Schwab classification model, Ames et al. incorporated many other demographic and clinical factors, such as past history, BMI, and so on. Through ML, AI was used to simulate human cognitive function to evaluate the clinical classification and prognosis of patients and built a classification model of adult spinal deformity (ASD). The results showed that through ML, by defining the patient group and intervention category on the basis of the original classification, a new ASD classification including 12 subgroups was generated, and a potential theoretical benchmark was proposed to evaluate the safety of surgical treatment, effectively solving the heterogeneity level of ASD clinical manifestations and treatment options.[24] The results suggest that on the one hand, the application of AI can effectively enhance and improve the accuracy and reliability of diagnostic classification and improve the cost-effectiveness of ASD treatment, and on the other hand, previously unrecognized outcomes and associations with potentially associated factors can be defined by immediate analysis of multiple factors. The results also suggest the possibility of re-optimization and re-creation of AI on the basis of the original prediction model.[25],[26] Spampinato et al. tested pretrained convolutional neural networks (CNNs) and found that bone age assessment could be completed after training with deep learning on generic images.[27] This is the first work for automated skeletal bone assessment tested on a public dataset for all possible cases. In fact, there are still many factors affecting bone development, such as parental height and body mass index. Through deep learning, we will be able to establish an expert system that can comprehensively incorporate a variety of factors to make predictions, and form a more accurate, reasonable and scientific bone age prediction model.
At the same time, AI has also shown its application value in digital pathology (DP). In the industrial life sciences strategy released in August 2017 by the UK government, it was emphasized that opportunities should be sought after in histopathology to create AI-based systems that can provide tumor grading and prognostic understandings which are currently unavailable by traditional methods. The current development of DP also lays the foundation for the intervention of AI. At present, histopathological sections can be converted into digital images through whole-slide scanners.[28] In recent years, the development of microscopic imaging and software systems for storage and review has led to the development of whole-slide imaging.[29] This technique allows the digitization of the entire slide rather than a single view, allowing it to be examined at a resolution comparable to that of a light microscope.[30] On this basis, the identification and quantification of cells and tissues has been successfully achieved by constructing CNNs with graded pattern detectors using for image-based detection and segmentation tasks.[31] At present, AI has been applied in breast pathology and lung pathology, including tumor heterogeneity and microscopical assessment, tumor grade, subtype assessment, prognostic significance, treatment response, and other aspects.[32],[33],[34],[35],[36],[37] In contrast, the application of AI in bone tumors is still slightly insufficient, but it is expected that AI can fully show its value in the pathological analysis of bone tumors in the future.
In addition, there are still some other problems in AI-aided diagnosis. At present, the types of diseases that can be diagnosed by AI are still very limited, and it is impossible to diagnose most of the common diseases in clinic. The diagnostic accuracy still needs to be improved, especially for rare diseases, which cannot be compared with experienced clinicians. The last problem that needs to be solved is how to combine AI-aided diagnosis with the existing clinical workflow so that it can be applied to clinical diagnosis in a real sense.
| Application Of Artificial Intelligence In Orthopedic Surgery | | |
Surgical treatment unquestionably occupies an important position in many orthopedic treatment methods. Therefore, the application of AI in orthopedic surgery has become an important display form of AI in orthopedic applications and is also an important source of feedback on patients' experience of AI application. At present, the application form of AI in orthopedic surgery can be broadly divided into improving surgical decision-making, assisting surgery, and improving the quality of rehabilitation.[2],[38],[39]
Studies have shown that misjudgment is the second most common cause of preventable injuries in surgical patients.[40] Surgical decisions are mainly made by hypothesis-deductive reasoning and individual judgment, which are often made by orthopedic surgeons under time limitations and numerous uncertainties. Therefore, these reasoning and judgments are often highly variable and not suitable to correct errors that occur. Loftus et al. proposed an AI model based on live-streaming electronic health record data training and combined it with bedside assessment and human intuition to effectively solve the shortcomings of traditional clinical decision support systems and achieve the purpose of enhancing surgical decision-making.[2] It is characterized by the ability to understand and incorporate the patients' subjective will, that is, the patients' values and emotions, in addition to clinically significant procedural influencing factors, and to automate the generation of prognostic data to increase efficiency. However, this AI model may be affected by heuristics and cognitive style, and also does not consider the physiological and risk factors of individual patients, which is prone to errors. How to solve the problem of heterogeneity will also become one of the directions that AI needs to continue to study in the future.
The application of AI in orthopedic surgery is currently represented by orthopedic surgical robots, which has been widely tested and applied in orthopedic surgeries, which is also a relatively balanced and well-studied field of AI applications. The traditional orthopedic surgical approach is susceptible to factors such as patient's positioning, accuracy of surgical instrument control, and personal experience and fatigue of doctors, which is difficult to precisely complete the surgical planning and reduces the success rate and reliability of surgery.[41] The application of surgical robots or manipulators under manual operation can improve the stability and accuracy of surgical operation, thereby improving the safety and reliability of surgery, the success rate of surgery, and the prognosis outcomes[42],[43] [Figure 3].With the rapid development of medical information data and the continuous progress of ML and deep learning technology, the surgical form under AI intervention has become a hot spot. The intervention of AI can assist the surgeon to select the optimal surgical approach and surgical angle, ensure the accuracy of surgery, and also innervate the robotic arm through AI to optimize the selection of surgical operation while ensuring the stability of surgical operation, so as to improve the clinical efficacy of surgery, which is also an important direction for the development of orthopedic surgery in the future.[42] At present, many surgical robots have been developed that can be used in clinical practice. For example, the Acrobat robotic system in the UK can be applied to single compartmental knee arthroplasty to obtain a better alignment effect. The study shows that the error between the tibiofemoral object angle and the original planned position after the assisted surgery of Acrobat robotic system is <2°. In contrast, only 40% of the traditional free-hand artificial replacement can achieve the surgical effect of robot-assisted group.[44] Integrated surgical systems company has developed Robodoc robot for assisting total knee arthroplasty (TKA), which is significantly superior to conventional TKA in terms of rotational alignment and condylar axis.[45],[46],[47] The Tianji robot was released in China in 2016. The Tianji robot is a multi-indication orthopedic surgical robot that can be used for spinal devices at all levels and pelvic, acetabular, and limb fracture surgery.[48],[49],[50] The Tianji robot combines the robotic arm with the real-time navigation system to improve the surgical accuracy. And compared with free-hand surgery, the Tianji robot significantly improves the accuracy of surgical equipment placement and improves the clinical treatment effect.[48] Zhao et al. designed a parallel robot (Gough–Stewart platform) for lower limb traction and fracture reduction. After combining the advantages of tandem and parallel mechanisms, hybrid robots can meet the clinical requirements of specific surgeries such as joint fractures,[51] so as to achieve more satisfactory fracture reduction. | Figure 3: Typical orthopedics robots. (a) ROBODOC, (b) BRIGIT, (c) Acrobot, (d) Mako RIO.[43]
Click here to view |
However, due to the lack of large-scale popularity of robots, the cost of robot-assisted surgery is high, and not all patients are conditionally used. In addition, robot-assisted surgery still needs clinicians to operate, so it is highly demanding for clinicians to operate with robots, and a lot of learning and practice are needed before complex robot operations. We also look forward to the emergence of fully automatic surgical robots in the future, completely replacing clinicians for surgery.
| Application Of Smartphones And Wearable Devices In Telemedicine | | |
Smartphones and wearable devices have laid the foundation for the development of telemedicine. These devices represent sensors that can store massive amounts of personal health data. These sensors can track the patients' physical status in a timely manner and solve many problems such as inaccurate and easy loss of follow-up information in traditional models.
Prem et al. developed and validated several machine-learning-based models of primary lower limb arthroplasty for preoperative calculation of patient-specific risk-adjusted value indicators, including distribution of costs, length of hospital stay, and time to discharge, thereby improving preventive management, refining preoperative planning, and addressing potential financial issues. In addition, Ramkumar et al. established a small data registry of human movement around TKA using current quite convenient mobile technologies, allowing remote patient monitoring to evaluate treatment compliance, outcomes, opioid intake (pain killers), mobility, and joint range of motion. They followed 25 patients who underwent TKA for the first time by providing a knee protector paired with a patient's smartphone and showed that no patient had uninterrupted data collection during a 90-day follow-up period, while the system also showed good incentives and attractiveness to patients.[52]
The research of Ramkumar et al suggests that we can combine AI with the application of portable health monitoring equipment, and make use of the advantage of AI in fast and high-dimensional data processing to improve the follow-up effect of patients, so as to further improve the evaluation of the effect of surgical treatment. This certainly brings confidence into the development of health industry and telemedicine in the future.
| Summary And Prospect | | |
With the improvement of economic and living standards, people's demand for medical resources increases rapidly, but the growth rate of medical resources supply with doctors as the core cannot meet this demand. With the rapid development of computer technology, the development of electronic medical information has become an irreversible trend. The continuous development of AI technology provides a new solution to the above needs. Using the unique advantages of ML and deep learning in the processing ability related to data, AI can assist clinicians in diagnostic decision-making and prognosis. Especially for the orthopedics departments that are dependent on imaging, AI can not only play a strong supporting role in the analysis and judgment of pathology but also assist the surgeon or independently find the best surgical path during surgery, playing a role in improving the accuracy and safety of surgery, and finally effectively improve the clinical treatment outcomes in orthopedics.
Although AI has shown many advantages, we should also recognize that there are still some shortcomings in the application of AI in the current context. On the one hand, the advantages of AI are based on its ability to process big data information. But now, the development of health information collecting medical device industry is still insignificant. The current situation determines the characteristics of high cost and high cost of AI. On the other hand, the principle of AI ML is based on the learning of large sample data, which means that AI is more suitable for the application of common diseases than rare diseases. In addition, AI also needs to face many problems such as whether it will cause fatigue and apraxia of doctors, as well as the relevant regulatory ethics when making decision judgments.
Overall, AI has shown great application potential and application value in orthopedic clinical practice. In the future, the application of AI in medical treatment presents an integrated and daily usage tool. More deep learning frameworks and deep learning systems are further developed, optimized, deployed, and applied to continuously realize innovation and breakthroughs. The future orthopedics will certainly be intelligent and highly trained orthopedics.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
| References | | |
| 1. | Tjardes T, Heller RA, Pförringer D, Lohmann R, Back DA; AG Digitalisierung der DGOU. Artificial intelligence in orthopedics and trauma surgery. Chirurg 2020;91:201-5.  |
| 2. | Loftus TJ, Tighe PJ, Filiberto AC, Efron PA, Brakenridge SC, Mohr AM, et al. Artificial intelligence and surgical decision-making. JAMA Surg 2020;155:148-58. |
| 3. | Kulkarni S, Seneviratne N, Baig MS, Khan AH. Artificial intelligence in medicine: Where are we now? Acad Radiol 2020;27:62-70. |
| 4. | Chen M, Decary M. Artificial intelligence in healthcare: An essential guide for health leaders. Healthc Manage Forum 2020;33:10-8. |
| 5. | Hamet P, Tremblay J. Artificial intelligence in medicine. Metabolism 2017;69S: S36-40. |
| 6. | Yu KH, Beam AL, Kohane IS. Artificial intelligence in healthcare. Nat Biomed Eng 2018;2:719-31. |
| 7. | Anderson PC. The post-physician Era: medicine in the 21 st century. JAMA 1977;237:2336-7. |
| 8. | Guo Y, Hao Z, Zhao S, Gong J, Yang F. Artificial intelligence in health care: Bibliometric analysis. J Med Internet Res 2020;22:e18228. |
| 9. | Badillo S, Banfai B, Birzele F, Davydov II, Hutchinson L, Kam-Thong T, et al. An introduction to machine learning. Clin Pharmacol Ther 2020;107:871-85. |
| 10. | Handelman GS, Kok HK, Chandra RV, Razavi AH, Lee MJ, Asadi H. eDoctor: Machine learning and the future of medicine. J Intern Med 2018;284:603-19. |
| 11. | Bi Q, Goodman KE, Kaminsky J, Lessler J. What is machine learning? A primer for the epidemiologist. Am J Epidemiol 2019;188:2222-39. |
| 12. | Esteva A, Robicquet A, Ramsundar B, Kuleshov V, DePristo M, Chou K, et al. A guide to deep learning in healthcare. Nat Med 2019;25:24-9. |
| 13. | Wang XS. Medical era in the future-artificial intelligence diagnosis and treatment. Chin J Colorectal Dis (Electronic Edition) 2017;6:349-52. |
| 14. | Chan HP, Samala RK, Hadjiiski LM, Zhou C. Deep learning in medical image analysis. Adv Exp Med Biol 2020;1213:3-21. |
| 15. | Jamaludin A, Lootus M, Kadir T, Zisserman A, Urban J, Battié MC, et al. ISSLS PRIZE IN BIOENGINEERING SCIENCE 2017: Automation of reading of radiological features from magnetic resonance images (MRIs) of the lumbar spine without human intervention is comparable with an expert radiologist. Eur Spine J 2017;26:1374-83. |
| 16. | Ravi D, Wong C, Deligianni F, Berthelot M, Andreu-Perez J, Lo B, et al. Deep learning for health informatics. IEEE J Biomed Health Inform 2017;21:4-21. |
| 17. | Hyung WJ, Son T, Park M, Lee H, Kim YN, Kim HI, et al. Superior prognosis prediction performance of deep learning for gastric cancercompared to Yonsei prognosis prediction model using Cox regression. J Clin Oncol 2017;35 Suppl 4:164. |
| 18. | Currie G, Hawk KE, Rohren E, Vial A, Klein R. Machine learning and deep learning in medical imaging: Intelligent imaging. J Med Imaging Radiat Sci 2019;50:477-87. |
| 19. | Wagner JB. Artificial intelligence in medical imaging. Radiol Technol 2019;90:489-501. |
| 20. | Giger ML. Machine learning in medical imaging. J Am Coll Radiol 2018;15:512-20. |
| 21. | Gyftopoulos S, Lin D, Knoll F, Doshi AM, Rodrigues TC, Recht MP. Artificial intelligence in musculoskeletal imaging: Current status and future directions. Am J Roentgenol 2019;213:506-13. |
| 22. | Olczak J, Fahlberg N, Maki A, Razavian AS, Jilert A, Stark A, et al. Artificial intelligence for analyzing orthopedic trauma radiographs. Acta Orthop 2017;88:581-6. |
| 23. | Li Q, Zhong L, Huang HN. Auxiliary diagnosis of developmental dysplasia of the hip by automated detection of Sharp's angle on standardized anteroposterior pelvic radiographs. Medicine 2019;98:52. |
| 24. | Ames CP, Smith JS, Pellisé F, Kelly M, Alanay A, Acaroğlu E, et al. Artificial intelligence based hierarchical clustering of patient types and intervention categories in adult spinal deformity surgery: Towards a new classification scheme that predicts quality and value. Spine (Phila Pa 1976) 2019;44:915-26. |
| 25. | Scheer JK, Osorio JA, Smith JS, Schwab F, Hart RA, Hostin R, et al. Development of a preoperative predictive model for reaching the Oswestry Disability Index minimal clinically important difference for adult spinal deformity patients. Spine Deform 2018;6:593-9. |
| 26. | Scheer JK, Smith JS, Schwab F, Lafage V, Shaffrey CI, Bess S, et al. Development of a preoperative predictive model for major complications following adult spinal deformity surgery. J Neurosurg Spine 2017;26:736-43. |
| 27. | Spampinato C, Palazzo S, Giordano D, Aldinucci M, Leonardi R. Deep learning for automated skeletal bone age assessment in X-ray images. Med Image Anal 2017;36:41-51. |
| 28. | Niazi MK, Parwani AV, Gurcan MN. Digital pathology and artificial intelligence. Lancet Oncol. 2019;20:e253-61. |
| 29. | Aeffner F, Zarella MD, Buchbinder N, Bui MM, Goodman MR, Hartman DJ, et al. Introduction to digital image analysis in whole-slide imaging: A white paper from the digital pathology association. J Pathol Inform 2019;10:9. [ PUBMED] [Full text] |
| 30. | Ibraahim A, Gamble P, Jaroensr R, Abdelsamea MM, Mermel CH, Chen PC, et al. Artificial intelligence in digital breast pathology: Techniques and applications. Breast 2020;49:267-73. |
| 31. | Garcia E, Hermoza R, Castanon CB, Cano C, Castillo M, Castanñeda C. Automatic lymphocyte detection on gastric cancer IHC images using deep learning. In: 2017 IEEE 30th International Symposium on Computer- Based Medical Systems (CBMS). IEEE 2017; pp:200-4. |
| 32. | Wang SD, Yang DH, Rong RC. Artificial Intelligence in lung cancer pathology image analysis. Cancers 2019;11:1673. |
| 33. | Lu C, Xu H, Xu J, Gilmore H, Mandal M, Madabhushi A. Multi-pass adaptive voting for nuclei detection in histopathological images. Sci Rep 2016;6:33985. |
| 34. | Whitney J, Corredor G, Janowczyk A, Ganesan S, Doyle S, Tomaszewski J, et al. Quantitative nuclear histomorphometry predicts oncotype DX risk categories for early stage ERþ breast cancer. BMC Cancer 2018;18:610. |
| 35. | Couture HD, Williams LA, Geradts J, Nyante SJ, Butler EN, Marron JS, et al. Image analysis with deep learning to predict breast cancer grade, ER status, histologic subtype, and intrinsic subtype. NPJ Breast Cancer 2018;4:30. |
| 36. | Robertson S, Azizpour H, Smith K, Hartman J. Digital image analysis in breast pathology-from image processing techniques to artificial intelligence. Transl Res 2018;194:19-35. |
| 37. | Ibrahim A, Gamble P, Jaroensri R, Abdelsamea MM, Mermel CH, Chen PC, et al. Artificial intelligence in digital breast pathology: Techniques and applications. Breast 2020;49:267-73. |
| 38. | Maffulli N, Rodriguez HC, Stone IW, Nam A, Song A, Gupta M, et al. Artificial intelligence and machine learning in orthopedic surgery: A systematic review protocol. J Orthop Surg Res 2020;15:478. |
| 39. | Panchmatia JR, Visenio MR, Panch T. The role of artificial intelligence in orthopaedic surgery. Br J Hosp Med (Lond) 2018;79:676-81. |
| 40. | Healey MA, Shackford SR, Osler TM, Rogers FB, Burns E. Complications in surgical patients. Arch Surg 2002;137:611-7. |
| 41. | Shao ZY, Xu WF, Liao XL, Xie WY. Application and prospects of orthopedic robots. Military Med 2016;40:1003-8. |
| 42. | Putzer D, Klug S, Haselbacher M, Mayr E, Nogler M. Retracting soft tissue in minimally invasive hip arthroplasty using a robotic arm: A comparison between a semiactive retractor holder and human assistants in a cadaver study. Surg Innov 2015;22:500-7. |
| 43. | Kuang S, Tang Y, Lin A, Yu S, Sun L. Intelligent control for human-robot cooperation in orthopedics surgery. Adv Exp Med Biol 2018;1093:245-62. |
| 44. | Cobb J, Henckel J, Gomes P, Harris S, Jakopec M, Rodriguez F, et al. Hands-on robotic unicompart-mental knee replacement: A prospective, randomised controlled study of the acrobot system. Bone Joint Surg Br 2006;88:188-97. |
| 45. | Moon YW, Ha CW, Do KH, Kim CY, Han JH, Na SE, et al. Comparison of robot-assisted and conventional total knee arthroplasty: A control cadaver study using multiparameter quantitative three-dimensional CT assessment of the alignment. Comput Aided Surg 2012;17:86-95. |
| 46. | Cao XH, Bai ZX, Sun CY, Li YL, Sun WD. Application of robot in orthopedic surgery: reliability and room for improvement. Chin J Tissue Eng Res 2020;24:1416-21. |
| 47. | Bautista M, Manrique J, Hozack WJ. Robotics in total knee arthroplasty. J Knee Surg 2019;32:600-6. |
| 48. | Han X, Tian W, Liu Y, Liu B, He D, Sun Y, et al. Safety and accuracy of robot-assisted versus fluoroscopy-assisted pedicle screw insertionin thoracolumbar spinal surgery: A prospective randomized controlled trial. Neurosurg Spine 2019;8:1-8. |
| 49. | Tian W. Robot-assisted posterior c1-2 transarticular screw fixation for atlantoaxial instability: A case report. Spine (Phila Pa 1976) 2016;41 Suppl 19:B2-5. |
| 50. | Han XG, Tian W. Artificial intelligence in orthopedic surgery: Current state and future perspective. Chin Med J (Engl) 2019;132:2521-3. |
| 51. | Zhao JX, Li CS, Ren HL, Hao M, Zhang LC, Tang PF. Evolution and current applications of robot-assisted fracture reduction: A comprehensive review. Ann Biomed Eng 2020;48:203-24. |
| 52. | RamkumarPN, Haeberle HS, BloomfieldMR, SchafferJL, KamathAF, Patterson BM, et al. Artificial intelligence and arthroplasty at a single institution: Real-world applications of machine learning to big data, value- based care, mobile health, and remote patient monitoring. J Arthroplasty 2019;34:2204-9. |
[Figure 1], [Figure 2], [Figure 3]
|
Search Pubmed for