Automated detection, labelling and radiological grading of clinical spinal MRIs.

Spinal magnetic resonance (MR) scans are a vital tool for diagnosing the cause of back pain for many diseases and conditions. However, interpreting clinically useful information from these scans can be challenging, time-consuming and hard to reproduce across different radiologists. In this paper, we alleviate these problems by introducing a multi-stage automated pipeline for analysing spinal MR scans. This pipeline first detects and labels vertebral bodies across several commonly used sequences (e.g. T1w, T2w and STIR) and fields of view (e.g. lumbar, cervical, whole spine). Using these detections it then performs automated diagnosis for several spinal disorders, including intervertebral disc degenerative changes in T1w and T2w lumbar scans, and spinal metastases, cord compression and vertebral fractures. To achieve this, we propose a new method of vertebrae detection and labelling, using vector fields to group together detected vertebral landmarks and a language-modelling inspired beam search to determine the corresponding levels of the detections. We also employ a new transformer-based architecture to perform radiological grading which incorporates context from multiple vertebrae and sequences, as a real radiologist would. The performance of each stage of the pipeline is tested in isolation on several clinical datasets, each consisting of 66 to 421 scans. The outputs are compared to manual annotations of expert radiologists, demonstrating accurate vertebrae detection across a range of scan parameters. Similarly, the model's grading predictions for various types of disc degeneration and detection of spinal metastases closely match those of an expert radiologist. To aid future research, our code and trained models are made publicly available.

Imaging Fetal Spine Malformations in the Context of In Utero Surgery.

This review covers the embryology, definition, and diagnosis of open spinal dysraphism with a focus on fetal ultrasound and MR imaging findings. Differentiating open versus closed spinal dysraphic defects on fetal imaging will also be discussed. Current fetal surgery practices and imaging findings in the context of fetal surgery are also reviewed.

Is there a connection between spine alignment, chest mobility, shoulder joint and respiratory parameters of patients with ankylosing spondylitis?

INTRODUCTION: Ankylosing spondylitis is chronic progressive disease, which decrease functions of musculoskeletal system including chest area. Those changes influences respiratory mechanics, worsen conditions of proper ventilation of lungs. OBJECTIVES: Rating of functional and respiratory parameters and dependence between them at patients with ankylosing spondylitis. MATERIALS & METHODS: The study included 45 patients with diagnosed ankylosing spondylitis. Chest and upper limbs mobility, resting spinal curvature alignment were assessed, and respiratory parameters were measured in a plethysmographic chamber JAGGER MasterScreen Body. RESULTS: Ankylosing spondylitis patients had lower respiratory parameters especially sReff, and FRC. Restriction of chest and upper limbs mobility was also demonstrated. Forward head extension was observed based on the occipital wall test. Correlations between functional parameters and correlations between functional and respiratory parameters were shown, in particular MIP, MEP, sReff, Rtot, TLC, ERV. CONCLUSIONS: The study confirmed a decrease in functional and respiratory parameters in the examined patients with ankylosing spondylitis compared to the applicable standards. A significant relationship was found between functional parameters in the upper body and respiratory parameters, which worsen with increasing thoracic dysfunction. The obtained results indicate the directions of therapy that should be taken into account to improve respiratory parameters and reduce respiratory dysfunction in these patients. Chest-focused physiotherapy appears to be an important element in improving function in patients with ankylosing spondylitis.

Effects of lower limb length discrepancy on spinopelvic compensation following total hip arthroplasty in patients with developmental dysplasia of the hip.

BACKGROUND: Limited research has examined the impact of lower limb length discrepancy (LLLD) alteration on spinopelvic compensation in individuals with developmental dysplasia of the hip (DDH). This study aimed to investigate the effects of LLLD on spinopelvic compensation following total hip arthroplasty (THA) and elucidate the complex biomechanical adaptations in the spinopelvic structures. METHODS: A retrospective review of DDH patients undergoing THA from January 2014 to December 2021 categorized individuals with Crowe type I and II into the low dislocation group (LDG, n = 94) and those with Crowe type III and IV into the high dislocation group (HDG, n = 43). Demographic data, as well as preoperative, postoperative, and last follow-up imaging data, including lower limb length (LLL), sacral obliquity (SO), iliac obliquity (IO), hip obliquity (HO), Cobb angle, apical vertebral translation (AVT), and coronal decompensation (CD), were collected for analysis. RESULTS: Patients in the LDG had a significantly higher surgical age and shorter disease duration (P<0.05). In LDG, patients exhibited substantial postoperative reductions in LLLD, SO, IO, and HO (P<0.05), while Cobb Angle, AVT, and CD showed no statistically significant changes (P>0.05). The variation in LLLD correlated significantly with the variations in SO, IO, and HO (P<0.05). Postoperative outcomes in the HDG demonstrated marked decreases in LLLD, SO, IO, HO, and CD (P<0.05), with no significant change in Cobb angle and AVT (P>0.05). The variation in LLLD correlated significantly with the variations in SO, IO, HO, and CD (P<0.05). CONCLUSIONS: THA effectively reduces LLLD in patients with DDH, and the variation in LLLD correlates meaningfully with the recovery of spinopelvic compensatory mechanisms.

Does spinopelvic alignment affect femoral head cartilage and the proximal femoral physis in slipped capital femoral epiphysis? A finite element analysis.

BACKGROUND: Slipped capital femoral epiphysis is a prevalent pediatric hip disorder. Recent studies suggest the spine's sagittal profile may influence the proximal femoral growth plate's slippage, an aspect not extensively explored. This study utilizes finite element analysis to investigate how various spinopelvic alignments affect shear stress and growth plate slip. METHODS: A finite element model was developed from CT scans of a healthy adult male lumbar spine, pelvis, and femurs. The model was subjected to various sagittal alignments through reorientation. Simulations of two-leg stance, one-leg stance, walking heel strike, ascending stairs heel strike, and descending stairs heel strike were conducted. Parameters measured included hip joint contact area, stress, and maximum growth plate Tresca (shear) stress. FINDINGS: Posterior pelvic tilt cases indicated larger shear stresses compared to the anterior pelvic tilt variants except in two leg stance. Two leg stance resulted in decreases in the posterior tilted pelvi variants hip contact and growth plate Tresca stress compared to anterior tilted pelvi, however a combination of posterior pelvic tilt and high pelvic incidence indicated larger shear stresses on the growth plate. One leg stance and heal strike resulted in higher shear stress on the growth plate in posterior pelvic tilt variants compared to anterior pelvic tilt, with a combination of posterior pelvic tilt and high pelvic incidence resulting in the largest shear. INTERPRETATION: Our findings suggest that posterior pelvic tilt and high pelvic incidence may lead to increased shear stress at the growth plate. Activities performed in patients with these alignments may predispose to biomechanical loading that shears the growth plate, potentially leading to slip.

Chiropractors' perceptions on the use of spinal radiographs in clinical practice: a qualitative study.

BACKGROUND: Radiography is commonly used in the assessment of spinal disorders, despite a lack of high-quality evidence demonstrating improved clinical outcomes or additional benefit to the patient. There is disagreement amongst chiropractors regarding the appropriate use of radiography for clinical management. This study aims to qualitatively explore chiropractors' perceptions on the use of spinal radiographs in clinical practice with respect to how they determine when to order radiographs; and how they use radiographs to inform clinical management. METHODS: Online qualitative semi-structured interviews were conducted with 17 Australian chiropractors who currently manage patients with spinal disorders. Convienence, snowball, and purposive sampling strategies were used to ensure an appropriate breadth and depth of participant characterisitcs and beliefs. Interview data were recorded, transcribed and analysed using framework analysis. RESULTS: Three themes were developed to describe how chiropractors determined when to order radiographs. These themes included specific findings from the clinical encounter that may inform clinical management, their perceptions of radiation risk, and the influence of clinical experience/intuition. Three themes and four subthemes were developed for how chiropractors use radiographs to inform their management. These themes explored the use of radiography for the application of chiropractic technique, as well as the role of radiographs in predicting patient prognosis, and as an educational tool to provide reassurance. CONCLUSION: Australian chiropractors' decision-making around spinal radiography is diverse and can be influenced by a number of clinical and external factors. Previously unexplored uses of spinal radiography in clinical practice were highlighted. Some chiropractors reported potential benefits of radiography that are currently not supported by research evidence. Future research should address how radiographic findings are reported to patients with spinal disorders and how this could be optimised to improve patient outcomes.

Fundamental principle of adult spinal deformity.

The management of adult spinal deformity (ASD) requires a personalized, multidisciplinary approach. Effective treatment hinges on thorough assessment using advanced imaging to understand the severity and impact of the spinal curvature. This paper underscores the importance of tailoring treatment plans to individual patient factors such as age, health, and psychological well-being, weighing both surgical and non-surgical options.Non-surgical treatments like pain management and physical therapy are preferred initially. If surgery is necessary, candidate selection and the choice of surgical technique are crucial. Minimally invasive procedures and advanced technologies like robotics enhance precision and reduce risks.Postoperative care and continuous monitoring are essential to assess the success of the intervention and manage any complications. This comprehensive strategy aims to improve overall functionality and quality of life, ensuring that treatment addresses both the physical deformity and its broader impacts. (Presented at the 2010th Meeting, May 20, 2024).

Sdc4 deletion perturbs intervertebral disc matrix homeostasis and promotes early osteopenia in the aging mouse spine.

Syndecan 4 (SDC4), a cell surface heparan sulfate proteoglycan, is known to regulate matrix catabolism by nucleus pulposus cells in an inflammatory milieu. However, the role of SDC4 in the aging spine has never been explored. Here we analyzed the spinal phenotype of Sdc4 global knockout (KO) mice as a function of age. Micro-computed tomography showed that Sdc4 deletion severely reduced vertebral trabecular and cortical bone mass, and biomechanical properties of vertebrae were significantly altered in Sdc4 KO mice. These changes in vertebral bone were likely due to elevated osteoclastic activity. The histological assessment showed subtle phenotypic changes in the intervertebral disc. Imaging-Fourier transform-infrared analyses showed a reduced relative ratio of mature collagen crosslinks in young adult nucleus pulposus (NP) and annulus fibrosus (AF) of KO compared to wildtype discs. Additionally, relative chondroitin sulfate levels increased in the NP compartment of the KO mice. Transcriptomic analysis of NP tissue using CompBio, an AI-based tool showed biological themes associated with prominent dysregulation of heparan sulfate GAG degradation, mitochondria metabolism, autophagy, endoplasmic reticulum (ER)-associated misfolded protein processes and ER to Golgi protein processing. Overall, this study highlights the important role of SDC4 in fine-tuning vertebral bone homeostasis and extracellular matrix homeostasis in the mouse intervertebral disc.

Rib index is a strong surrogate of scoliometric reading in idiopathic scoliosis.

INTRODUCTION AND AIM: Scoliometry is not always included in the examination protocol of IS patients. The aim of this report is to examine the degree of correlation of Segmental Rib Index (SRI) to scoliometry, in order for SRI to be used as a surrogate of scoliometric angle of trunk rotation (ATR). MATERIAL AND METHOD: 66 Idiopathic Scoliosis (IS) subjects were studied, with a mean age 12.2 ± 2.9 years, 18 boys and 48 girls: 20 thoracic, 22 thoracolumbar and 24 lumbar curves. The standing lateral spine radiographs (LSR) were obtained and the Segmental Rib Index (SRI) from T1 to T12 were assessed. The ATR was documented. RESULTS: In all 66 cases with IS the scoliometer readings (ATR) were significantly correlated to the SRI at the T6, T7 and T8 levels. In the thoracic curves SRI and ATR correlations were significant for the levels T6-T12. DISCUSSION: It was suggested that as long as the patients doesn't have scoliometer measurements, the SRI, could be used as a surrogate for scoliometry. It was also found that in thoracic, thoracolumbar and lumbar level, in both genders, changing from the flexed position to the standing position, the mean trunk asymmetry (TA) decreases. Therefore, if these patients had their TA measured using a scoliometer during the Adams test, their body asymmetry would have been greater than that measured using the SRI method on standing LSR. Consequently, it is evident that the significantly correlated SRI used as a surrogate for the scoliometric assessment of TA is reasonably a strong surrogate.

Comprehensive assessment of global spinal sagittal alignment and related normal spinal loads in a healthy population.

Abnormal postoperative global sagittal alignment (GSA) is associated with an increased risk of mechanical complications after spinal surgery. Typical assessment of sagittal alignment relies on a few selected measures, disregarding global complexity and variability of the sagittal curvature. The normative range of spinal loads associated with GSA has not yet been considered in clinical evaluation. The study objectives were to develop a new GSA assessment method that holistically describes the inherent relationships within GSA and to estimate the related spinal loads. Vertebral endplates were annotated on radiographs of 85 non-pathological subjects. A Principal Component Analysis (PCA) was performed to derive a Statistical Shape Model (SSM). Associations between identified GSA variability modes and conventional alignment measures were assessed. Simulations of respective Shape Modes (SMs) were performed using an established musculoskeletal AnyBody model to estimate normal variation in cervico-thoraco-lumbar loads. The first six principal components explained 97.96% of GSA variance. The SSM provides the normative range of GSA and a visual representation of the main variability modes. Normal variation relative to the population mean in identified alignment features was found to influence spinal loads, e.g. the lower bound of the second shape mode (SM2-2σ) corresponds to an increase in L4L5-compression by 378.64 N (67.86%). Six unique alignment features were sufficient to describe GSA almost entirely, demonstrating the value of the proposed method for an objective and comprehensive analysis of GSA. The influence of these features on spinal loads provides a normative biomechanical reference, eventually guiding surgical planning of deformity correction in the future.

Clinical applications of AI-prediction tools in spine surgery: a narrative review.

Spine surgery has grown into a wide, complex field encompassing trauma surgery to deformity to tumours. Artificial intelligence (AI) based technology has been particularly useful in improving imaging-reporting and detection of predictive patterns. The purpose of this narrative review is to present practical approaches towards implementing upcoming AI spine research for clinicians to help improve practices, clinical throughput, and surgical decision-making.

Age- and sex-related changes in vertebral trabecular bone architecture in Neolithic and Mediaeval populations from Poland.

This paper investigates trabecular bone ontogenetic changes in two different Polish populations, one prehistoric and the other historical. The studied populations are from the Brzesc Kujawski region in Kujawy (north-central Poland), one from the Neolithic Period (4500-4000 BC) and one from the Middle Ages (twelfth-sixteenth centuries AD), in total 62 vertebral specimens (32 males, 30 females). Eight morphometric parameters acquired from microCT scan images were analysed. Two-way ANOVA after Box-Cox transformation and multifactorial regression model were calculated. A significant decrease in percentage bone volume fraction (BV/TV; [%]) with age at death was observed in the studied sample; Tb.N (trabecular number) was also significantly decreased with age; trabecular separation (Tb.Sp) increased with advancing age; connectivity density (Conn.D) was negatively correlated with biological age and higher in the Neolithic population. These data are found to be compatible with data from the current biomedical literature, while no loss of horizontal trabeculae was recorded as would be expected based on modern osteoporosis.

MRI-based vertebral bone quality score as a novel bone status marker of patients with adolescent idiopathic scoliosis.

To investigate the application of MRI-based vertebral bone quality (VBQ) score in assessing bone mineral density (BMD) for patients with adolescent idiopathic scoliosis (AIS). We reviewed the data of AIS patients between January 2021 and October 2023 with MRI, whole-spine plain radiographs, quantitative computed tomography (QCT) and general information. VBQ score was calculated using T1-weighted MRI. Univariate analysis was applied to present the differences between variables of patients with normal BMD group (QCT Z-score > - 2.0) and low BMD group (QCT Z-score ≤ - 2.0). The correlation between VBQ score and QCT Z-score was analyzed with Pearson correlation test. A multivariate logistic regression model was used to determine the independent factors related to low BMD. Receiver operating characteristic curve (ROC) was drawn to analyze the diagnostic performance of VBQ score in distinguishing low BMD. A total of 136 AIS patients (mean age was 14.84 ± 2.10 years) were included, of which 41 had low BMD. The low BMD group had a significantly higher VBQ score than that in normal group (3.48 ± 0.85 vs. 2.62 ± 0.62, P < 0.001). The VBQ score was significantly negative correlated with QCT Z score (r = - 0.454, P < 0.001). On multivariate analysis, VBQ score was independently associated with low BMD (OR: 4.134, 95% CI 2.136-8.000, P < 0.001). The area under the ROC curve indicated that the diagnostic accuracy of the VBQ score for predicting low BMD was 81%. A sensitivity of 65.9% with a specificity of 88.4% could be achieved for distinguishing low BMD by setting the VBQ score cutoff as 3.18. The novel VBQ score was a promising tool in distinguishing low BMD in patients with AIS and could be useful as opportunistic assessment for screening and complementary evaluation to QCT before surgery.

Artificial Intelligence Assistance for the Measurement of Full Alignment Parameters in Whole-Spine Lateral Radiographs.

BACKGROUND: Measuring spinal alignment with radiological parameters is essential in patients with spinal conditions likely to be treated surgically. These evaluations are not usually included in the radiological report. As a result, spinal surgeons commonly perform the measurement, which is time-consuming and subject to errors. We aim to develop a fully automated artificial intelligence (AI) tool to assist in measuring alignment parameters in whole-spine lateral radiograph (WSL X-rays). METHODS: We developed a tool called Vertebrai that automatically calculates the global spinal parameters (GSPs): Pelvic incidence, sacral slope, pelvic tilt, L1-L4 angle, L4-S1 lumbo-pelvic angle, T1 pelvic angle, sagittal vertical axis, cervical lordosis, C1-C2 lordosis, lumbar lordosis, mid-thoracic kyphosis, proximal thoracic kyphosis, global thoracic kyphosis, T1 slope, C2-C7 plummet, spino-sacral angle, C7 tilt, global tilt, spinopelvic tilt, and hip odontoid axis. We assessed human-AI interaction instead of AI performance alone. We compared the time to measure GSP and inter-rater agreement with and without AI assistance. Two institutional datasets were created with 2267 multilabel images for classification and 784 WSL X-rays with reference standard landmark labeled by spinal surgeons. RESULTS: Vertebrai significantly reduced the measurement time comparing spine surgeons with AI assistance and the AI algorithm alone, without human intervention (3 minutes vs. 0.26 minutes; P < 0.05). Vertebrai achieved an average accuracy of 83% in detecting abnormal alignment values, with the sacral slope parameter exhibiting the lowest accuracy at 61.5% and spinopelvic tilt demonstrating the highest accuracy at 100%. Intraclass correlation analysis revealed a high level of correlation and consistency in the global alignment parameters. CONCLUSIONS: Vertebrai's measurements can accurately detect alignment parameters, making it a promising tool for measuring GSP automatically.

Diagnostic efficacy and clinical impact of image-guided core needle biopsy of suspected vertebral osteomyelitis.

OBJECTIVES: The diagnostic yield and clinical impact of image-guided core needle biopsy (ICNB) of suspected vertebral osteomyelitis in adults is heterogenous in published studies owing to small sample sizes, indicating the need for large cohort studies. METHODS: A retrospective analysis of ICNBs was performed from 2010 to 2021 for patients with imaging findings consistent with vertebral osteomyelitis. For each biopsy, a series of factors were analyzed, as well as if histopathology was diagnostic of osteomyelitis and if microbiological cultures were positive. In addition, it was recorded in what way biopsy influenced clinical management regarding antimicrobial treatment. A multivariate statistical analysis was performed to evaluate the factors associated with yield. RESULTS: A total of 570 biopsies performed on 527 patients were included. A histopathologic diagnosis of osteomyelitis was made in 68.4% (359 of 525) of biopsies, and microbiological cultures were positive in 29.6% (169 of 570). Elevated erythrocyte sedimentation rate was positively associated with a histopathologic diagnosis of osteomyelitis (odds ratio [OR] =1.96, P = 0.007) and positive cultures from bone cores (OR = 1.02, P ≤0.001) and aspirate (OR = 1.02, P ≤0.001). Increased total core length was positively associated with a histopathologic diagnosis of osteomyelitis (OR = 1.81, P = 0.013) and positive cultures from bone cores (OR = 1.65, P = 0.049). Clinical management was affected by ICNB in 37.5% (214 of 570) of cases. CONCLUSIONS: In this large cohort, ICNB yielded approximately 30% positive cultures and changed clinical management in over one-third of the patients.

From 2D to 3D: automatic measurement of the Cobb angle in adolescent idiopathic scoliosis with the weight-bearing 3D imaging.

BACKGROUND CONTEXT: Adolescent idiopathic scoliosis (AIS) necessitates accurate spinal curvature assessment for effective clinical management. Traditional two-dimensional (2D) Cobb angle measurements have been the standard, but the emergence of three-dimensional (3D) automatic measurement techniques, such as those using weight-bearing 3D imaging (WR3D), presents an opportunity to enhance the accuracy and comprehensiveness of AIS evaluation. PURPOSE: This study aimed to compare traditional 2D Cobb angle measurements with 3D automatic measurements utilizing the WR3D imaging technique in patients with AIS. STUDY DESIGN/SETTING: A cohort of 53 AIS patients was recruited, encompassing 88 spinal curves, for comparative analysis. PATIENT SAMPLE: The patient sample consisted of 53 individuals diagnosed with AIS. OUTCOME MEASURES: Cobb angles were calculated using the conventional 2D method and three different 3D methods: the Analytical Method (AM), the Plane Intersecting Method (PIM), and the Plane Projection Method (PPM). METHODS: The 2D cobb angle was manually measured by 3 experienced clinicians with 2D frontal whole-spine radiographs. For 3D cobb angle measurements, the spine and femoral heads were segmented from the WR3D images using a 3D-UNet deep-learning model, and the automatic calculations of the angles were performed with the 3D slicer software. RESULTS: AM and PIM estimates were found to be significantly larger than 2D measurements. Conversely, PPM results showed no statistical difference compared to the 2D method. These findings were consistent in a subgroup analysis based on 2D Cobb angles. CONCLUSION: Each 3D measurement method provides a unique assessment of spinal curvature, with PPM offering values closely resembling 2D measurements, while AM and PIM yield larger estimations. The utilization of WR3D technology alongside deep learning segmentation ensures accuracy and efficiency in comparative analyses. However, additional studies, particularly involving patients with severe curves, are required to validate and expand on these results. This study emphasizes the importance of selecting an appropriate measurement method considering the imaging modality and clinical context when assessing AIS, and it also underlines the need for continuous refinement of these techniques for optimal use in clinical decision-making and patient management.

Differences in vertebral bone density between African apes.

OBJECTIVES: Low-energy vertebral fractures are a common health concern, especially in elderly people. Interestingly, African apes do not seem to experience as many vertebral fractures and the low-energy ones are even rarer. One potential explanation for this difference is the lower bone density in humans. Yet, only limited research has been done on the vertebral bone density of the great apes and these have mainly included only single vertebrae. Hence the study aim is to expand our understanding of the vertebral microstructure of African apes in multiple spinal segments. MATERIALS: Bone density in the vertebral body of C7, T12, and L3 was measured from 32 Pan troglodytes and 26 Gorilla gorilla using peripheral quantitative computed tomography (pQCT). RESULTS: There was a clear difference between the three individual vertebrae and consequently the spinal segments in terms of trabecular density and cortical density and thickness. The variation of these bone parameters between the vertebrae differed between the apes but was also different from those reported for humans. The chimpanzees were observed to have overall higher trabecular density, but gorillas had higher cortical density and thickness. Cortical thickness had a relatively strong association with the vertebral size. DISCUSSION: Despite the similarity in locomotion and posture, the results show slight differences in the bone parameters and their variation between spinal segments in African apes. This variation also differs from humans and appears to indicate a complex influence of locomotion, posture, and body size on the different spinal segments.

CT-derived vertebral bone mineral density is a useful biomarker to predict COVID-19 outcome.

The low vertebral bone computed tomography (CT) Hounsfield unit values measured on CT scans reflect low bone mineral density (BMD) and are known as diagnostic indicators for osteoporosis. The potential prognostic significance of low BMD defined by vertebral bone CT values for the coronavirus disease 2019 (COVID-19) remains unclear. This study aimed to assess the impact of BMD on the clinical outcome in Japanese patients with COVID-19 and evaluate the association between BMD and critical outcomes, such as high-flow nasal cannula, non-invasive and invasive positive pressure ventilation, extracorporeal membrane oxygenation, or death. We examined the effects of COVID-19 severity on the change of BMD over time. This multicenter retrospective cohort study enrolled 1132 inpatients with COVID-19 from the Japan COVID-19 Task Force database between February 2020 and September 2022. The bone CT values of the 4th, 7th, and 10th thoracic vertebrae were measured from chest CT images. The average of these values was defined as BMD. Furthermore, a comparative analysis was conducted between the BMD on admission and its value 3 months later. The low BMD group had a higher proportion of critical outcomes than did the high BMD group. In a subanalysis stratifying patients by epidemic wave according to onset time, critical outcomes were higher in the low BMD group in the 1st-4th waves. Multivariable logistic analysis of previously reported factors associated with COVID-19 severity revealed that low BMD, chronic kidney disease, and diabetes were independently associated with critical outcomes. At 3 months post-infection, patients with oxygen demand during hospitalization showed markedly decreased BMD than did those on admission. Low BMD in patients with COVID-19 may help predict severe disease after the disease onset. BMD may decrease over time in patients with severe COVID-19, and the impact on sequelae symptoms should be investigated in the future.