Mirko Kaiser

Last Name

Kaiser

First Name

Mirko

ORCID

0000-0002-8763-5003

Organisational unit

03994 - Taylor, William R. / Taylor, William R.

Show all metadata

Search Results

Publications 1 - 10 of 18

Estimating Spinal Kinematics Using a 3D Shape Model of the Human Back During Dynamic Movement
Kaiser, Mirko; Hügli, Alain; Reng, Jing; et al. (2024)
CMBBE 2024 Abstract Book
Advancing 3D idiopathic scoliosis assessment through optical scans and web-based integration - Validation of the ScolioSIM system
Cukovic, Sasa; Luković, Vanja; Zou, Yiying; et al. (2025)
Computers in Biology and Medicine
Background and objective: The ScolioSIM module, a novel addition to the web-based ScolioMedIS system, was developed specifically for comprehensive 3D scoliosis visualization and assessment. It utilizes 3D optical scans of the back surface, 3D coordinates of anatomical landmarks, internal and external spinal alignment data, and a generic spine (GS) 3D model to create patient-specific (PS) 3D spine models and deformity parameters. The aim of this paper is to 1) describe the novel ScolioSIM module for scoliosis assessment and 2) validate the agreement between patient-specific (PS) 3D models and Cobb angles produced by the ScolioSIM in comparison to the gold standard - Cobb angles obtained clinically from biplanar radiographs. Methods: To generate PS 3D models and deformity parameters, the process includes scanning the patient's back, manually acquiring 3D coordinates of anatomical landmarks and external spinal alignment on the back surface, and determining spinal midline based on X-ray images registered with the surface. Additionally, the process involves uploading the files to ScolioSIM for further analysis, including interpolation, projection, and calculation of spinal deformity parameters, and aligning the GS 3D model to generate the PS 3D model. The deformity parameters are then transferred to ScolioMedIS for inclusion in diagnostic reports alongside the PS 3D spine model. Thirty participants with idiopathic scoliosis (IS) were included in the study, and biplanar EOS radiographs and optical 3D scans were acquired in an upright standing pose using a positioning frame. The study compares Cobb angles generated by ScolioSIM (SCA) with manual assessments on EOS radiographs (MCA) and those computed by sterEOS software (ECA). Results: A total of 80 curves were measured using ScolioSIM. The mean MCA, ECA, and SCA were 22.8 ± 13.0°, 27.9 ± 13.0°, and 24.9 ± 14.3°, respectively. Results showed excellent correlations between SCA and MCA for the thoracic (r = 0.96, 95 % CI: 0.93 to 0.98), lumbar (r = 0.93, 95 % CI: 0.85 to 0.97), and combined thoracic-lumbar regions (r = 0.93, 95 % CI: 0.90 to 0.96). Conclusions: This study introduces ScolioSIM, a web-based module with an optical scan system for 3D assessment of IS, demonstrating a high correlation of Cobb angle measurements with those from the clinical gold standard. Additionally, ScolioMedIS could facilitate multi-center data collection and collaborative decision-making in the management of adolescents with IS. The key advantage of the ScolioSIM module is its ability to compare examinations across different visits longitudinally. ScolioSIM thus holds the potential for monitoring progression and evaluating treatment outcomes in adolescents with IS.
From back shape to spine: A novel radiation-free system to assess posture, motion, and pathologies of the spine
Kaiser, Mirko (2025)
Optical 3D surface scanning combined with “back-shape-to-spine” approaches that estimate deformation parameters solely from the 3D back surface offers a promising method for radiation-free assessment of spinal deformities in patients with adolescent idiopathic scoliosis (AIS). By leveraging back shape asymmetries to estimate the Cobb angle of primary curve (pCA), these approaches avoid the radiation exposure associated with the established radiographic techniques in clinical practice. Despite their potential, existing 3D scanning approaches often underestimate spinal deformities, necessitating deeper understanding of scoliosis and its effect on the back shape. This thesis investigates the capabilities of optical 3D surface scanning in patients with AIS with a focus on back-shape-to-spine approaches to detect spinal deformities and estimate spinal alignment. First, the accuracy and precision required for optical 3D surface scanners were evaluated. This established minimum spatial resolution requirements of 2–5 mm and minimum camera resolution of 640 × 480 pixels. The measurement quality of various optical 3D scanners was identified. These scanners each use a different methodology: single-shot structured light, multishot structured light, and active stereo. This provided guidelines for selecting a cost-effective optical 3D scanning system that satisfies the minimal requirements. A system meeting these requirements was identified, with a resolution of up to 1600 × 1200 @ 30fps and an accuracy under 2 mm. In particular, the spatial resolution, accuracy, and precision of a single-shot and multishot structured-light 3D imaging system in static and dynamic scenarios were evaluated using a novel phantom model. This revealed submillimeter accuracy at velocities up to 0.6 m/s. Second, the thesis explores the relationship between spinal alignment and back shape, with improvements in back-shape-to-spine approaches using the PCdare research platform, which integrates 3D surface scans and biplanar radiographs. In addition, the impact on the accuracy of estimates of internal parameters was investigated with various postures, including upright standing and Adam’s forward bending test, and dynamic measurements, such as forward bending and lateral bending. Results showed very strong correlations between estimated and gold-standard parameters, with median Cobb angle errors of less than 3.4° and median root mean square errors between the lines through the centroids of vertebral bodies of 6.9 mm. These findings increase confidence in the use of 3D scanning with a back-shape-to-spine approach for clinical applications in scoliosis screening and monitoring. However, the accuracy and long-term viability of these methods require further evaluation with a larger dataset and a longitudinal study involving repeated scans over several years. Continued data collection, software development, and model refinement will likely enhance the clinical applicability of optical 3D surface scanning in scoliosis management.
Investigating the relationship between internal spinal alignment and back shape in patients with scoliosis using PCdare: a comparative, reliability and validation study
Kaiser, Mirko; Mudavamkunnel, Meby; Bertsch, Martin; et al. (2025)
medRxiv
Optical 3D surface scanning has emerged as a valuable modality for assessing spinal deformity in patients with scoliosis, avoiding radiation exposure. However, correlations remain moderate between deformation parameters obtained from radiographs and those estimated solely from the 3D back surface, referred to as the “back-shape-to-spine” approach. To improve the accuracy with which the back-shape-to-spine approach can estimate the internal spinal alignment (ISL) from 3D surface scanning, deeper understanding is required of the effect of scoliosis on the back shape. The PCdare software, which enables semi-automatic registration of 3D surface scans with the corresponding biplanar radiographs, has been used by students in a previous study to validate study protocols, generate references for estimated ISL, and evaluate correlations between the spinous process line (SPL) and ISL. This study explored the potential of the PCdare software to investigate the underlying relationship between the ISL and the 3D back shape, conducted a comparative study with 3 study protocols, and conducted an inter- and intrarater reliability (IIR) study with 6 clinicians and 10 students as raters to evaluate the applicability of PCdare when used by students. The comparative study involved 252 patients with idiopathic scoliosis from 3 studies that compared the back-shape-to-spine approach with radiography. The quality of study protocols and the relationship between internal spinal alignment and 3D back shape were both investigated by evaluating the posture alignment errors and correlations between Cobb angles. The inter- and intrarater reliability study involved 7 patients with idiopathic scoliosis and was conducted using PCdare and validated with PACS. The median Cobb angle difference (interquartile range: IQR) between students and clinicians (interclass) was 0.06° (1.5°). The ICC [confidence interval] between Cobb angles (interrater) was 0.94 [0.7,0.98]. The median absolute Cobb angle difference (IQR) between 3 repetitions (intrarater) were 4.2° (5.3°) or lower. The median Cobb angle difference (IQR) between PCdare and PACS was 1.5° (8.4°) for clinicians and 1.4° (6.9°) for students, whereas the corresponding correlation [confidence interval] was 0.94 [0.92,0.96] and 0.95 [0.93,0.96], respectively. The median RMSE (median SD) of posture alignment error ranged between 8.1 mm (5.2 mm) and 5 mm (3.5 mm), whereas the median PCC (IQR) between ISL and SPL ranged between 0.64 (0.58) and 0.99 (0.02). Students achieve outcomes comparable to clinicians when using PCdare, which underlines its reliability and ease of use. In addition, the application of PCdare to examine the quality of study protocols revealed the necessity of markers and posture alignment and delivered correlation coefficients for the relationship between internal spinal alignment and 3D back shape. These findings highlight the potential of the PCdare software to advance the non-ionizing assessment of spinal deformities and thus improve understanding of scoliosis.
Noninvasive estimation of internal spinal alignment in patients with adolescent idiopathic scoliosis using PCdare and back shape asymmetry
Kaiser, Mirko; McLaughlin, Emily; Bertsch, Martin; et al. (2025)
Scientific Reports
Optical 3D surface scanning is used increasingly to assess spinal deformity in patients with adolescent idiopathic scoliosis (AIS), largely because it avoids additional radiation burden. However, such approaches generally underestimate the extent of the abnormality. Improving the accuracy of such estimates requires a deeper understanding of AIS and its effect on back shape. We present a unique platform with publicly available code that contains noninvasive and nonionizing approaches to estimating the Cobb angle of the primary curve, called primary Cobb angle (pCA) and internal spinal alignment (ISL) in patients with AIS. Our approaches use asymmetries of the back shape during upright standing, the Adam’s forward bending test, bending forward, and lateral bending. The results have revealed strong (0.75 [0.53, 0.87]) to excellent (0.91 [0.81, 0.96]) correlations [95% confidence interval] and a median pairwise absolute error (IQR) of 3.4° (6.8°) between the estimated pCAs and clinical gold-standard assessments in 30 patients. The correlations (IQR) between estimated shape of ISLs and their references were very strong (0.87 (0.24)) to excellent (0.94 (0.03)), and the median root mean square error (IQR) between estimated and reference ISL was 6.9 mm (3.3 mm). These results indicate confidence both in the use of 3D scanning using a “back-shape-to-spine” approach and in the establishment of optical 3D surface scanning approaches for scoliosis screening and monitoring in clinical practice.
Automating spinal alignment analysis with machine learning and PCdare software: focus on idiopathic scoliosis
Stoican, Kübra; Simsar, Enis; Bertsch, Martin; et al. (2025)
Proceedings of SPIE ~ Medical Imaging 2025: Clinical and Biomedical Imaging
Optical 3D scanning is an emerging technology for evaluating spinal deformities in patients with idiopathic scoliosis. Despite the potential of surface scanning to avoid radiation exposure, its reliability and accuracy remain moderate compared to radiography. PCdare is an open-source software that is used to analyze and compare the optical 3D surface scans of the patients’ back with the corresponding biplanar radiographic images of the patients’ spine to better understand the relationship between internal spinal alignment and 3D back shape. It includes two important components: the line through the centroids of the vertebral bodies, the internal spinal line (ISL), and the line connecting all spinous processes on the back, the spinous process line (SPL). However, this software requires users to manually draw the ISL, which can be time consuming and subjective. Here we present a fully automatic approach for ISL estimation to improve the workflow of PCdare. We implemented and trained a convolutional neural network-based U-Net on 330 anterior-posterior radiographs of the spine and subsequently detected the centroids of vertebral bodies to estimate the ISL. We evaluated the model’s performance on a test set of 63 images, obtaining a median dice similarity coefficient of 0.93. The median root mean square error between ISL estimations and ground truth was 1.16 mm. Our method demonstrated promising results and will improve the applicability of the PCdare software.
Exploring radiation-free scoliosis monitoring: systematic review and meta-analysis of non-ionizing methods
Bertsch, Martin; Mulatero, Lucrezia; Salehpour, Sina; et al. (2025)
BMC Musculoskeletal Disorders
Background: Idiopathic scoliosis is a three-dimensional spinal deformity that typically develops during childhood or adolescence but may affect individuals across the lifespan. Regular monitoring is often necessary to detect progression and assess treatment effectiveness. Radiography remains the clinical gold standard; however, repeated ionizing radiation exposure is associated with increased cancer risks, highlighting the need for reliable, non-invasive, and radiation-free assessment methods. This systematic review and meta-analysis evaluated the diagnostic accuracy and criterion validity of emerging radiation-free scoliosis monitoring techniques compared to radiographic standards. Methods: A comprehensive literature search across six databases (Cochrane, EMBASE, IEEE Xplore, PUBMED, Scopus, Web of Science) identified 56 eligible studies involving 4,774 patients diagnosed with idiopathic scoliosis (median number of patients per study: 38; range: 5 to 952, mean patient age: 15.2 years, female-to-male ratio: 3:1). Criterion validity was assessed by pooling Pearson correlation coefficients between radiographic and non-ionizing measurements. Measurement accuracy was assessed by pooling their mean absolute differences in Cobb angles. Additionally, sensitivity and specificity for detecting deformity progression were assessed. Statistical analyses employed multilevel linear mixed-effects models, introducing moderators to explain study heterogeneity. Results: Ultrasonography demonstrated the highest overall validity, consistently correlating strongly (r≈0.9) with radiographic Cobb angles. Surface topography also showed robust correlation (r > 0.8), although evidence remains insufficient for patients with higher body mass indices or more severe spinal curvatures for both methods. Magnetic resonance imaging exhibited a very strong correlation (r = 0.93) with radiographic measurements; however, correlation varied significantly depending on patient positioning. Upright MRI provided more consistent results compared to supine positioning. Conclusions: Ultrasonography and surface topography represent promising radiation-free alternatives that could significantly reduce radiographic assessment frequency, minimizing radiation exposure, particularly in suitable patient groups. While MRI also demonstrates excellent validity, its broader clinical applicability remains constrained by substantial costs, limited availability, and extended examination durations. Although these non-ionizing modalities are not yet viable replacements for routine radiography, their demonstrated validity and accuracy supports their potential as complementary technologies, particularly for screening or supplementary monitoring of scoliosis.
Spine Detection and Vertebral Body Segmentation in the Pcdare Software with Machine Learning
Stoican, Kübra; Bertsch, Martin; Laux, Christoph J.; et al. (2025)
SSRN
Background and Objective: Optical 3D scanning is an emerging technology for evaluating back shapes in patients with spinal deformities. Despite the potential of surface scanning to avoid radiation exposure, its reliability and accuracy remain moderate compared to radiography, which is the current gold standard in clinical practice. PCdare is an open-source software that is used to analyze and compare the optical 3D surface scans of the back with the corresponding biplanar radiographic images of the spine to better understand the relationship between internal spinal alignment and 3D back shape. It considers the internal spinal line (ISL), which is the line through the centroids of the vertebral bodies. However, PCdare requires users to manually draw the ISL, which can be time consuming and subjective. Methods: Here we present a fully automatic approach for ISL estimation to improve the workflow of PCdare. We implemented and trained a pipeline of two convolutional neural networks, based on the well-known U-Net and YOLO, on 328 anterior-posterior radiographs of the spine and subsequently detected the centroids of vertebral bodies to estimate the ISL. Results: We evaluated the model’s performance on a test set of 63 images, obtaining a median dice similarity coefficient of 0.93 and the median root mean square error between ISL estimations and ground truth was 0.89 mm. Conclusions: Our method demonstrated promising results and improved the applicability of the PCdare software. These findings may support the integration of 3D optical scanning for scoliosis screening in clinical settings.
Pcdare Software Registers 3d Back Surface with Biplanar Radiographs: Application to Patients with Scoliosis
Kaiser, Mirko; Bertsch, Martin; Laux, Christoph J.; et al. (2024)
SSRN
Background and ObjectiveOptical 3D surface scanning is used increasingly in medical applications, including in the assessment of spinal deformity of patients with scoliosis. Despite the potential of surface scanning to avoid radiation burden on patients, correlations between the deformation parameters estimated solely from the 3D back surface in a “back-shape-to-spine” approach and the results obtained from radiographs remain moderate, with a tendency to underestimate deformity from optical 3D scanning. To improve the back-shape-to-spine approach and the accuracy with which internal spinal alignment (ISL) can be estimated from 3D surface scanning, deeper understanding is required of scoliosis and its effect on the back shape. Investigating this relationship requires registration of 3D surface scans with biplanar radiographic images. To the best of our knowledge, no publicly available software currently provides fully automated registration of optical 3D surface scans with any radiographic images and includes evaluations of the relationship between the ISL and the 3D back surface.MethodsWe present the PCdare software, with publicly available code, which registers 3D surface scans with the corresponding biplanar radiographs semi-automatically and facilitates investigations into the relationship between surface and internal modalities. The PCdare software was used to calculate the correlation between the spinous process line (SPL) and ISL. In addition, PCdare was used to generate reference lines for the ISL estimated solely from the 3D back surface. To validate the reference lines, the Cobb angles calculated with PCdare were compared with the gold standard in clinical practice, manual annotation in a picture archiving and communication system (PACS).ResultsThe median Pearson correlation coefficient (PCC) between SPL and ISL, calculated using PCdare, ranged from very strong (> 0.8) to excellent (> 0.9) for 30 patients with scoliosis. The median difference between the Cobb angles of frontal curvature calculated using PCdare and manually annotated in PACS was -0.9°.ConclusionsThese results strengthen confidence in the use of 3D scanning combined with the back-shape-to-spine approach and confirm the applicability of the PCdare software to investigations of the relationship between internal spinal alignment and 3D back surface.
Optimizing 3D Registration With the PCDicom App to Analyse the Relation Between Spinal Alignment and Back Shape
Kaiser, Mirko; Bertsch, Martin; Cukovic, Sasa; et al. (2024)
29th Congress of the ESB - Abstracts

Publications 1 - 10 of 18

Mirko Kaiser

Last Name

First Name

ORCID

Organisational unit

Filters

Search Results