A new 3D software for analysis and planning of lower limb and patellofemoral alignment: Reliability and accuracy

Knee landmarks detection via deep learning for automatic imaging evaluation of trochlear dysplasia and patellar height

OBJECTIVES

To develop and validate a deep learning-based approach to automatically measure the patellofemoral instability (PFI) indices related to patellar height and trochlear dysplasia in knee magnetic resonance imaging (MRI) scans.

METHODS

A total of 763 knee MRI slices from 95 patients were included in the study, and 3393 anatomical landmarks were annotated for measuring sulcus angle (SA), trochlear facet asymmetry (TFA), trochlear groove depth (TGD) and lateral trochlear inclination (LTI) to assess trochlear dysplasia, and Insall-Salvati index (ISI), modified Insall-Salvati index (MISI), Caton Deschamps index (CDI) and patellotrochlear index (PTI) to assess patellar height. A U-Net based network was implemented to predict the landmarks' locations. The successful detection rate (SDR) and the mean absolute error (MAE) evaluation metrics were used to evaluate the performance of the network. The intraclass correlation coefficient (ICC) was also used to evaluate the reliability of the proposed framework to measure the mentioned PFI indices.

RESULTS

The developed models achieved good accuracy in predicting the landmarks' locations, with a maximum value for the MAE of 1.38 ± 0.76 mm. The results show that LTI, TGD, ISI, CDI and PTI can be measured with excellent reliability (ICC > 0.9), and SA, TFA and MISI can be measured with good reliability (ICC > 0.75), with the proposed framework.

CONCLUSIONS

This study proposes a reliable approach with promising applicability for automatic patellar height and trochlear dysplasia assessment, assisting the radiologists in their clinical practice.

CLINICAL RELEVANCE STATEMENT

The objective knee landmarks detection on MRI images provided by artificial intelligence may improve the reproducibility and reliability of the imaging evaluation of trochlear anatomy and patellar height, assisting radiologists in their clinical practice in the patellofemoral instability assessment.

KEY POINTS

• Imaging evaluation of patellofemoral instability is subjective and vulnerable to substantial intra and interobserver variability. • Patellar height and trochlear dysplasia are reliably assessed in MRI by means of artificial intelligence (AI). • The developed AI framework provides an objective evaluation of patellar height and trochlear dysplasia enhancing the clinical practice of the radiologists.

Epidemiology of Lateral Patellar Dislocation Including Bone Bruise Incidence: Five Years of Data from a Trauma Center

OBJECTIVE

Systematic summary of the epidemiology of patellar dislocation is rare. This study aims to investigate sex-, age-, type-, injury causing events-, incidence of bone bruise and time from last injury (TFLI)-specific characteristics, and detail the epidemiological characteristics of patellar dislocation.

METHOD

In this descriptive epidemiological study, a total of 743 patients who have a history of lateral patellar dislocation with either first-time patellar dislocation (FPD) or recurrent patellar dislocation (RPD) between August 2017 and June 2022 at our institution met the inclusion criteria and were selected in this study. Patient characteristics including the type, gender, age, events leading to patellar dislocation, incidence of patellar bone bruise, and the time from last injury (TFLI) of patellar dislocation were retrospectively obtained and described. Magnetic resonance imaging scans (MRI) of the knee were reviewed for insuring bone bruise.

RESULTS

Among the 743 patients with patellar dislocation who required surgical reconstruction of the medial retinaculum, 418 (56.2%) had RPD and 325 (43.8%) had FPD. There were more females (65.0%) than males (35.0%) in patellar dislocation patients. Among the female patients, those aged <18 years had higher incidence (31.4%) of patellar dislocation. Among the male patients, those aged <18 and 19-28 years had higher incidence (16.8%) of patellar dislocation. Of all age groups, the prevalence rate of patellar dislocation was high in juvenile population and females, but with no statistical significance. The most common patellar dislocation-causing event was sport accidents (40.1%), followed by life accidents (23.2%). The incidence of left-knee patellar dislocation was slightly higher than that of right-knee patellar dislocation. The incidence of patellar bone bruise of RPD (63.2%) was significantly lower (p < 0.05) than that of FPD (82.2%). Patellar dislocation patients with bone bruise had shorter time from last injury (TFLI) than those without patellar bone bruise (p < 0.05).

CONCLUSIONS

The incidence of bone bruise of RPD was lower than that of FPD, and patients with patellar bone bruise may have a shorter time from last injury than those without bone bruise.

Accuracy, Reliability, and Repeatability of a Novel Artificial Intelligence Algorithm Converting Two-Dimensional Radiographs to Three-Dimensional Bone Models for Total Knee Arthroplasty

BACKGROUND

With the emergence of advanced technology, such as robotics, three-dimensional (3D) imaging is necessary to execute preoperative surgical plans accurately. However, 3D imaging adds cost and potential risk to patients. This study determined the measurement accuracy, reliability, and repeatability of a novel artificial intelligence (AI) algorithm which converts two-dimensional (2D) radiographs to 3D bone models.

METHODS

An AI algorithm was developed to convert 2D radiographs to 3D bone model reconstructions. The accuracy of the AI algorithm was evaluated by comparing mean absolute error in measurements performed on 3D bone reconstructions, 3D computed tomography (CT) scans, and manual measurements on five cadaveric knees. Reliability and repeatability of the AI algorithm were evaluated by assessing the inter-observer and intra-observer agreement between measurements performed on 3D bone reconstructions, respectively.

RESULTS

Accuracy of the AI algorithm was considered excellent with mean absolute errors <2mm in 9 of 12 anatomical parameters compared with measurements performed on CTs and manual calipers. All inter-observer and intra-observer correlation coefficients were greater than 0.90 representing a high level of measurement reliability and repeatability by independent observers and the same observers.

CONCLUSION

This particular AI algorithm demonstrated a high degree of accuracy, reliability, and repeatability for converting 2D radiographs to 3D bone reconstructions similar to a CT-scan. Study results suggest this AI algorithm has the potential for use in preoperative surgical planning due to its efficiencies related to cost and time and reduced radiation exposure without the use of 3D imaging.