Skip to main content

Table 1 Characteristics of studies included in the final analysis

From: Understanding the use of artificial intelligence for implant analysis in total joint arthroplasty: a systematic review

Author

Journal

Sample Size

Number of Implant Designs

Image Source

Imaging

Implant location

AI device

Purpose

Training:

Validation:

Testing Split

Minors Score

Belete et al., 2021 [22]

Informatics in Medicine Unlocked

558

7

Single Institution

AP Radiograph

Knee

CNN (ResNet)

Implant Identification

50:25:25

20

Bonnin et al., 2023 [23]

Journal of Arthroplasty

38,751

4

Single Institution

AP & Lateral Radiograph

Knee

X-TKA, 12 DCNN

Implant Identification

60:20:20

21

Borjali et al., 2020 [24]

Journal of Orthopaedic Research

252

3

Single Institution

AP Radiograph

Hip

DCNN

Implant Identification

80:10:10

20

Borjali et al., 2021 [25]

Medical Physics

402

9

Single Institution

AP Radiograph

Hip

DCNN

Implant Identification

80:10:10

21

Lau et al., 2022 [37]

Journal of Orthopaedic Translation

440

NR

Single Institution

AP Radiograph

Knee

Xception Model

Failure (Loosening)

75:25a

21

Ghose et al., 2020 [26]

ICISS

878

6

Multi-Center & Textbooks

AP & Lateral Radiograph

Knee

DCNN (various)

Implant Identification

80:10:10

19

Gong et al., 2022 [27]

Scientific Reports

714

4

Single Institution

AP Radiograph

Hip (Stem and Cup)

CNN

Implant Identification

60:30:10

20

Jang et al., 2023 [28]

Journal of Arthroplasty

235

NR

Single Institution

AP Radiograph

Knee

U-Net Model

Fixation Zone & Cone Mapping Identification

60:20:20

20

Kang et al., 2020 [29]

Journal of Orthopaedic Translation

170

29

Multi-Center

AP Radiograph

Hip and Knee

YOLOv3 Object Detection

Keras Deep Learning Platform

Implant Identification

75:25a

20

Karnuta et al., 2021 [8]

Journal of Arthroplasty

682

9

Multi-Center

AP Radiograph

Knee

DCNN (Inception V3)

Implant Identification

80:10:10

21

Karnuta et al., 2021 [7]

Journal of Arthroplasty

1972

18

Multi-Center

AP Radiograph

Hip

CNN (InceptionV3)

Implant Identification

80:10:10

20

Karnuta et al., 2022 [21]

Journal of Arthroplasty

2954

8

Multi-Center

AP Radiograph

Hip

CNN

Implant Identification

70:10:20

21

Klemt et al., 2022 [30]

JAAOS

11,204

24 THA, 14 TKA

Single Institution

AP Radiograph

Hip and Knee

CNN

Implant Identification

80:20a

20

Murphy et al., 2022 [36]

HIP International

2,440

8

Single Institution

AP Radiograph

Hip

DCNN

Implant Identification

60:30:10

20

Patel et al., 2021 [31]

Radiology: Artificial Intelligence

922 THA, 427 TKA

8 THA, 4 TKA

Single Institution

AP Radiograph

Hip and Knee

DCNN (various)

Implant Identification

70:20:10

21

Rouzrokh et al., 2022 [32]

Radiology Artificial Intelligence

700

2

Single Institution

AP Radiograph

Hip

U-Net Model

Failure (Subsidence)

70:15:15

21

Schwarz et al., 2022 [33]

KSSTA

1,512

8

Single Institution

Long Leg Radiograph

Knee

IB Lab LAMA

Measurements

 > 15,000:200:1,312

21

Sharma et al., 2021 [34]

Indian Journal of Orthopaedics

1,078

6

Multi-Center

AP & Lateral Radiograph

Knee

DCNN (various)

Implant Identification

75:15:10

20

Tiwari et al., 2022 [20]

Journal of Orthopaedics

521

6

Single Institution & Google Images

AP & Lateral Radiograph

Knee

Transfer Machine Learning

Implant Identification

70:20:10

21

Yi et al., 2020 [35]

The Knee Journal

274

2

Single Institution & Two Public Datasets

AP Radiograph

Knee

CNN

Implant Identification

70:10:20

20

  1. CNN Convolutional Neural Network, DCNN Deep Convolutional Neural Network, ResNet Residual Network, AP Anterior Posterior
  2. aNo Validation