Shuhei Ohyama, Satoshi Maki, Toshiaki Kotani, Yosuke Ogata, Tsuyoshi Sakuma, Yasushi Iijima, Tsutomu Akazawa, Kazuhide Inage, Yasuhiro Shiga, Masahiro Inoue, Takahito Arai, Noriyasu Toshi, Soichiro Tokeshi, Kohei Okuyama, Susumu Tashiro, Noritaka Suzuki, Yawara Eguchi, Sumihisa Orita, Shohei Minami, Seiji Ohtori

Eur Spine. 2025 Feb 4.doi: 10.1007/s00586-025-08680-9. Online ahead of print.

Abstract

Purpose: This study was designed to develop a machine learning (ML) model that predicts future Cobb angle in patients with adolescent idiopathic scoliosis (AIS) using minimal radiographs and simple questionnaires during the first and second visits.

Methods: Our study focused on 887 female patients with AIS who were initially consulted at a specialized scoliosis center from July 2011 to February 2023. Patient data, including demographic and radiographic data based on anterior-posterior and lateral whole-spine radiographs, were collected at the first, second, and final visits. ML algorithms were employed to develop individual regression models for future Cobb angles of each curve type (proximal thoracic: PT, main thoracic: MT, and thoracolumbar/lumbar: TLL) using PyCaret in Python. Multiple models were explored and analyzed, with the selection of optimal models based on the coefficient of determination (R²) and median absolute error (MAE).

Results: For the future curve of PT, MT, and TLL, the top-performing models exhibit R² of 0.73, 0.63, and 0.61 and achieve MAE of 2.3°, 4.0°, and 4.2°.

Conclusions: The ML-based model using items commonly evaluated at the first and second visits accurately predicted future Cobb angles in female patients with AIS.

Keywords: Adolescent idiopathic scoliosis; Classification model; Machine learning algorithms; Prediction model; Regression model.

https://pubmed.ncbi.nlm.nih.gov/39903251/

Siegmund Lang, Jacopo Vitale, Fabio Galbusera, Tamás Fekete, Louis Boissiere, Yann Philippe Charles, Altug Yucekul, Caglar Yilgor, Susana Núñez-Pereira, Sleiman Haddad , Alejandro Gomez-Rice, Jwalant Mehta, Javier Pizones, Ferran Pellisé, Ibrahim Obeid, Ahmet Alanay, Frank Kleinstück, Markus Loibl ; ESSG European Spine Study Group

Spine Deform. 2024 Nov 4. doi: 10.1007/s43390-024-00955-3. Online ahead of print

Abstract

Purpose: Large language models (LLM) have the potential to bridge knowledge gaps in patient education and enrich patient-surgeon interactions. This study evaluated three chatbots for delivering empathetic and precise adolescent idiopathic scoliosis (AIS) related information and management advice. Specifically, we assessed the accuracy, clarity, and relevance of the information provided, aiming to determine the effectiveness of LLMs in addressing common patient queries and enhancing their understanding of AIS.

Methods: We sourced 20 webpages for the top frequently asked questions (FAQs) about AIS and formulated 10 critical questions based on them. Three advanced LLMs-ChatGPT 3.5, ChatGPT 4.0, and Google Bard-were selected to answer these questions, with responses limited to 200 words. The LLMs’ responses were evaluated by a blinded group of experienced deformity surgeons (members of the European Spine Study Group) from seven European spine centers. A pre-established 4-level rating system from excellent to unsatisfactory was used with a further rating for clarity, comprehensiveness, and empathy on the 5-point Likert scale. If not rated ‘excellent’, the raters were asked to report the reasons for their decision for each question. Lastly, raters were asked for their opinion towards AI in healthcare in general in six questions.

Results: The responses among all LLMs were ‘excellent’ in 26% of responses, with ChatGPT-4.0 leading (39%), followed by Bard (17%). ChatGPT-4.0 was rated superior to Bard and ChatGPT 3.5 (p = 0.003). Discrepancies among raters were significant (p < 0.0001), questioning inter-rater reliability. No substantial differences were noted in answer distribution by question (p = 0.43). The answers on diagnosis (Q2) and causes (Q4) of AIS were top-rated. The most dissatisfaction was seen in the answers regarding definitions (Q1) and long-term results (Q7). Exhaustiveness, clarity, empathy, and length of the answers were positively rated (> 3.0 on 5.0) and did not demonstrate any differences among LLMs. However, GPT-3.5 struggled with language suitability and empathy, while Bard’s responses were overly detailed and less empathetic. Overall, raters found that 9% of answers were off-topic and 22% contained clear mistakes.

Conclusion: Our study offers crucial insights into the strengths and weaknesses of current LLMs in AIS patient and parent education, highlighting the promise of advancements like ChatGPT-4.o and Gemini alongside the need for continuous improvement in empathy, contextual understanding, and language appropriateness.

Keywords: Adolescent idiopathic scoliosis (AIS); Artificial intelligence (AI); Large language models (LLMs); Patient education; Spine surgery.

https://pubmed.ncbi.nlm.nih.gov/39495402/

Gözde Yagcı, Halil Gökhan Demirkıran 

J Manipulative Physiol Ther. 2024 Oct 14:S0161-4754(24)00053-8.doi:10.1016/j.jmpt.2024.08.015

Abstract

Objective: This study aimed to investigate the effectiveness of Scientific Exercise Approach to Scoliosis (SEAS) protocol for patients with mild adolescent idiopathic scoliosis (AIS).

Method: (A controlled clinical trial) This preliminary study included 32 patients with AIS, aged 10 to 16 years, with Cobb angles ranging from 12° to 25°. The participants were divided into 2 groups: the SEAS group (n = 21) and standard care (n = 11). We assessed Cobb angles via radiographs, axial trunk rotation using the Adams test, spinal deformity from various perspectives with the Spinal Appearance Questionnaire (SAQ), and health-related quality of life with the Scoliosis Research Society-22 (SRS-22) questionnaire, both before and after 1 year of treatment.

Results: Repeated measure-ANOVA revealed significant group x time interactions for various parameters: Cobb angle (F_1,30 = 35.757; P < .001), axial trunk rotation (F_1,30 = 39.595; P < .001), SAQ appearance domain for the physiotherapist (F_1,30 = 53.943; P < .001), SAQ appearance (F_1,30= 14.054; P = .001), expectations (F_1,30 = 9.103; P = .005) domains, the total score (F_1,30 = 23.574; P < .001) for the child, and the SAQ appearance domain for the parent (F_1,29 = 5.412; P = .027). There were no significant differences in SRS-22 scores between the 2 groups. The success rates for the SEAS group and the standard care group were 100% and 72.7%, respectively.

Conclusion: After 1 year, the group of patients with mild AIS who received the SEAS method showed greater improvements in the Cobb angle, axial trunk rotation, and spinal appearance than those who received standard care. This preliminary study suggests that larger trials should be completed.

Keywords: Body; Exercise Therapy; Physical Appearance; Quality of Life; Scoliosis.

Copyright © 2024. Published by Elsevier Inc.

https://pubmed.ncbi.nlm.nih.gov/39412452/

Lening Li, Man-Sang Wong 

Skeletal Radiol. 2024 Aug;53(8):1517-1528.doi: 10.1007/s00256-024-04592-9. Epub 2024 Feb 21.

Abstract

Predicting curve progression during the initial visit is pivotal in the disease management of patients with adolescent idiopathic scoliosis (AIS)-identifying patients at high risk of progression is essential for timely and proactive interventions. Both radiological and clinical factors have been investigated as predictors of curve progression. With the evolution of machine learning technologies, the integration of multidimensional information now enables precise predictions of curve progression. This review focuses on the application of machine learning methods to predict AIS curve progression, analyzing 15 selected studies that utilize various machine learning models and the risk factors employed for predictions. Key findings indicate that machine learning models can provide higher precision in predictions compared to traditional methods, and their implementation could lead to more personalized patient management. However, due to the model interpretability and data complexity, more comprehensive and multi-center studies are needed to transition from research to clinical practice.

Keywords: Adolescent idiopathic scoliosis; Machine learning; Prediction.

https://pubmed.ncbi.nlm.nih.gov/39118179/

Prudence Wing Hang Cheung, Oi Kiu Olivia Chan, Hao Wu, Marcus Kin Long Lai, Lester Po Kwan Wong, Shiyu Tang, Jason Pui Yin Cheung 

JAMA Pediatr. 2024 Jul 1;178(7):657-668.doi: 10.1001/jamapediatrics.2024.1484

Abstract

Importance: Lack of evidence and consensus for brace weaning protocol in adolescent idiopathic scoliosis (AIS) results in clinicians prescribing gradual weaning in the hope of avoiding curve deterioration after weaning. However, gradual weaning contributes to prolonged brace wear, which can affect spinal stiffness and health-related quality of life (HRQoL).

Objective: To determine whether gradual weaning results in better curve magnitude and truncal balance maintenance after brace weaning vs immediate brace removal for patients with AIS.

Design, setting, and participants: This was an open-labeled randomized clinical trial commenced in April 2017 with 24-month follow-up completed in January 2023. Outcome assessors were masked to weaning protocol assigned. The study took place at a territory-wide tertiary scoliosis clinic serving the largest number of referrals in the local population. Patients with AIS ready to wean off of brace wear were eligible (402 were screened; 33 were excluded [15 for <18 hours/day of brace-wear compliance before weaning, 11 were treated with Milwaukee brace, and 7 declined to participate]; and 369 were included), and those who were treated with a custom molded thoracolumbosacral orthosis and had reached skeletal maturity were consecutively recruited.

Interventions: Patients were randomized to gradual weaning protocol (n = 176) with an additional 6 months of nighttime wear before completely stopping or immediate weaning protocol (n = 193) with immediate brace removal at recruitment.

Main outcomes and measures: Changes in major curve Cobb angle and truncal balance from the time of weaning to 6-month, 12-month, and 24-month follow-up. HRQoL was also assessed using the refined Scoliosis Research Society 22-item and EuroQol 5-dimension questionnaires.

Results: A total of 369 patients (mean [SD] age, 14.9 [1.1] years; 304 [83.4%] girls) were randomized with 284 (77.0%) completing 24-month longitudinal follow-up. Immediate and gradual weaning groups had no significant differences in change of major Cobb angle at postweaning 6-month (difference, -0.6°; 95% CI, -1.4 to 0.2; P = .17), 12-month (difference, -0.3°; 95% CI, -1.2 to 0.6; P = .47), and 24-month (difference, -0.3°; 95% CI, -1.2 to 0.7; P = .60) follow-up. The number of curve progression, nonprogression, and rebound cases were comparable (χ22 = 2.123; P = .35). Postweaning changes in truncal balance and HRQoL demonstrated no significant differences between groups.

Conclusions: Gradual weaning did not demonstrate superiority to immediate weaning with predefined criteria of Cobb angle and truncal balance maintenance and HRQoL after brace weaning. Gradual and immediate weaning achieved very similar maintenance of brace outcomes in AIS. We therefore recommend the consideration of immediate brace weaning, which aims to benefit patients with earlier time for increased exercises and activity level.

Trial registration: ClinicalTrials.gov Identifier: NCT03329716.

https://pubmed.ncbi.nlm.nih.gov/38829664/