Radiographic evaluation for patients with scoliosis using Cobb method is the current gold standard, but radiography has radiation hazards. Several groups have recently demonstrated the feasibility of using 3D ultrasound for the evaluation of scoliosis. Ultrasound imaging is radiation-free, comparatively more accessible, and inexpensive. However, a reliable and valid 3D ultrasound system ready for clinical scoliosis assessment has not yet been reported. Scolioscan is a newly developed system targeted for scoliosis assessment in clinics by using coronal images of spine generated by a 3D ultrasound volume projection imaging method. The aim of this study is to test the reliability of spine deformity measurement of Scolioscan and its validity compared to the gold standard Cobb angle measurements from radiography in adolescent idiopathic scoliosis (AIS) patients. Prospective study divided into two stages: 1) Investigation of intra- and inter- reliability between two operators for acquiring images using Scolioscan and among three raters for measuring spinal curves from those images; 2) Correlation between the Cobb angle obtained from radiography by a medical doctor and the spine curve angle obtained using Scolioscan (Scolioscan angle). The raters for ultrasound images and the doctors for evaluating radiographic images were mutually blinded. The two stages of tests involved 20 (80 % females, total of 26 angles, age of 16.4 ± 2.7 years, and Cobb angle of 27.6 ± 11.8°) and 49 (69 % female, 73 angles, 15.8 ± 2.7 years and 24.8 ± 9.7°) AIS patients, respectively. Intra-class correlation coefficients (ICC) and Bland-Altman plots and root-mean-square differences (RMS) were employed to determine correlations, which interpreted based on defined criteria. We demonstrated a very good intra-rater and intra-operator reliability for Scolioscan angle measurement with ICC larger than 0.94 and 0.88, respectively. Very good inter-rater and inter-operator reliability was also demonstrated, with both ICC larger than 0.87. For the thoracic deformity measurement, the RMS were 2.5 and 3.3° in the intra- and inter-operator tests, and 1.5 and 3.6° in the intra- and inter-rater tests, respectively. The RMS differences were 3.1, 3.1, 1.6, 3.7° in the intra- and inter-operator and intra- and inter-rater tests, respectively, for the lumbar angle measurement. Moderate to strong correlations (R2 > 0.72) were observed between the Scolioscan angles and Cobb angles for both the thoracic and lumbar regions. It was noted that the Scolioscan angle slightly underestimated the spinal deformity in comparison with Cobb angle, and an overall regression equation y = 1.1797x (R2 = 0.76) could be used to translate the Scolioscan angle (x) to Cobb angle (y) for this group of patients. The RMS difference between Scolioscan angle and Cobb angle was 4.7 and 6.2°, with and without the correlation using the overall regression equation. We showed that Scolioscan is reliable for measuring coronal deformity for patients with AIS and appears promising in screening large numbers of patients, for progress monitoring, and evaluation of treatment outcomes. Due to it being radiation-free and relatively low-cost, Scolioscan has potential to be widely implemented and may contribute to reducing radiation dose during serial monitoring.

The application of ultrasound imaging for spine evaluation could minimize radiation exposure for patients with adolescence idiopathic scoliosis (AIS). A customized three-dimensional (3D) ultrasound imaging system has been demonstrated to provide reliable and valid coronal curvature measurements. However, these measurements were using the spinous processes as anatomical reference, leading to a predictable underestimation of the traditionally used Cobb angles. An alternative 3D ultrasound image reconstruction method was applied to create coronal images with more lateral features for angle measurement. The objective of this study was to test the reliability and the validity of this angle, the ultrasound curve angle (UCA), and compare the UCA with the Cobb angles on X-ray images of patients with AIS. This study was divided into: 1) Investigation of intra- and inter-reliability between two raters for measuring the UCA and two operators for acquiring ultrasound images; 2) Investigation of the validity between the radiographic Cobb angle and the UCA. Fifty patients and 164 patients with AIS, were included in the two stages, respectively. Patients underwent bi-planar X-ray and 3D ultrasound scanning on the same day. The proposed UCA was used to measure the coronal curvature from the ultrasound coronal images, which were formed using a newly customized volume projection imaging (VPI) method. The intra-rater/operator and inter-rater and operator reliability of the UCA were tested by intra-class correlation coefficient (ICC) (3,1) and (2,1), respectively. The validity of UCA measurements as compared to radiographic Cobb angles was tested by inter-method ICC (2,1), mean absolute difference (MAD), standard error of measurement (SEM), Pearson correlation coefficient and Bland–Altman statistics. The level of significance was set as 0.05. Excellent intra-rater and intra-operator (ICC (3,1)≥0.973) and excellent inter-rater and inter-operator reliability (ICC (2,1)≥0.925) for UCA measurement, with overall MAD and SEM no more than 3.5° and 1.7° were demonstrated for both main thoracic and (thoraco)lumbar curvatures. Very good correlations were observed between UCA and Cobb angle for main thoracic (R2=0.893) and (thoraco)lumbar (R2=0.884) curves. The mean (SD) measurements in terms of radiographic Cobb and UCA were 27.2 ​± ​11.6° and 26.3 ​± ​11.4° for main thoracic curves; and 26.2 ​± ​11.4° and 24.8 ​± ​9.7° for (thoraco)lumbar curve respectively. One hundred sixty-four subjects (33 male and 131 female subjects; 11–18 years of age, mean of 15.1 ​± ​1.9 years) were included for the validity session. Excellent inter-method variations (ICC (2,K) ≥0.933) with overall MAD and SEM no more than 3.0° and 1.5° were demonstrated for both main thoracic and (thoraco)lumbar curvatures. In addition, Bland–Altman plots demonstrated an acceptable agreement between ultrasound and radiographic Cobb measurements. In this study, very good correlations and agreement were demonstrated between the ultrasound and X-ray measurements of the scoliotic curvature. Judging from the promising results of this study, patients with AIS with different severity of curves can be evaluated and monitored by ultrasound imaging, reducing the usage of radiation during follow-ups. This method could also be used for scoliosis screening. The Translational potential of this article: Ultrasound curve angle (UCA) obtained from 3D ultrasound imaging system can provide reliable and valid evaluation on coronal curvature for patients with AIS, without the need of radiation.

To diagnose scoliosis, the standing radiograph with Cobb’s method is the gold standard for clinical practice. Recently, three-dimensional (3D) ultrasound imaging, which is radiation-free and inexpensive, has been demonstrated to be reliable for the assessment of scoliosis and validated by several groups. A portable 3D ultrasound system for scoliosis assessment is very much demanded, as it can further extend its potential applications for scoliosis screening, diagnosis, monitoring, treatment outcome measurement, and progress prediction. The aim of this study was to investigate the reliability of a newly developed portable 3D ultrasound imaging system, Scolioscan Air, for scoliosis assessment using coronal images it generated. The system was comprised of a handheld probe and tablet PC linking with a USB cable, and the probe further included a palm-sized ultrasound module together with a low-profile optical spatial sensor. A plastic phantom with three different angle structures built-in was used to evaluate the accuracy of measurement by positioning in 10 different orientations. Then, 19 volunteers with scoliosis (13F and 6M; Age: 13.6 ± 3.2 years) with different severity of scoliosis were assessed. Each subject underwent scanning by a commercially available 3D ultrasound imaging system, Scolioscan, and the portable 3D ultrasound imaging system, with the same posture on the same date. The spinal process angles (SPA) were measured in the coronal images formed by both systems and compared with each other. The angle phantom measurement showed the measured angles well agreed with the designed values, 59.7 ± 2.9 vs. 60 degrees, 40.8 ± 1.9 vs. 40 degrees, and 20.9 ± 2.1 vs. 20 degrees. For the subject tests, results demonstrated that there was a very good agreement between the angles obtained by the two systems, with a strong correlation (R² = 0.78) for the 29 curves measured. The absolute difference between the two data sets was 2.9 ± 1.8 degrees. In addition, there was a small mean difference of 1.2 degrees, and the differences were symmetrically distributed around the mean difference according to the Bland–Altman test. Scolioscan Air was sufficiently comparable to Scolioscan in scoliosis assessment, overcoming the space limitation of Scolioscan and thus providing wider applications. Further studies involving a larger number of subjects are worthwhile to demonstrate its potential clinical values for the management of scoliosis.

Volume Projection Imaging from ultrasound data is a promising technique to visualize spine features and diagnose Adolescent Idiopathic Scoliosis. In this paper, we present a novel multi-task framework to reduce the scan noise in volume projection images and to segment different spine features simultaneously, which provides an appealing alternative for intelligent scoliosis assessment in clinical applications. Our proposed framework consists of two streams: i) A noise removal stream based on generative adversarial networks, which aims to achieve effective scan noise removal in a weakly-supervised manner, i.e., without paired noisy-clean samples for learning; ii) A spine segmentation stream, which aims to predict accurate bone masks. To establish the interaction between these two tasks, we propose a selective feature-sharing strategy to transfer only the beneficial features, while filtering out the useless or harmful information. We evaluate our proposed framework on both scan noise removal and spine segmentation tasks. The experimental results demonstrate that our proposed method achieves promising performance on both tasks, which provides an appealing approach to facilitating clinical diagnosis.