Renée Lampe

Background: Adolescent idiopathic scoliosis is a condition that affects up to 3% of adolescents. Although radiographic imaging is the gold standard for diagnosis and monitoring, repeated exposure to ionizing radiation poses significant health risks. Alternative methods, such as estimating spinal alignment and Cobb angle from back surface topography, have been proposed to mitigate these risks. However, validation of these methods is challenging because traditional datasets combine back surface scans and radiographs acquired at different times and in different postures. This introduces technique-dependent errors and makes it difficult to determine whether observed discrepancies stem from the method itself or from inconsistencies in the validation dataset. As a result, this uncertainty in methods accuracy arising from verification errors, limits these alternatives to complementary roles rather than establishing them as reliable diagnostic tools. Methods: To address this limitation, we introduce a fully automated approach for generating verification datasets directly from computed tomography images. This approach enables precise and synchronized extraction of both external back surface and internal spinal structures, thereby eliminating posture-related discrepancies present in commonly used datasets. We assess the utility of our method through a two-step validation: first, by comparing automatically extracted parameters with manual measurements; and second, by comparing the results of an established surface topography method applied to our dataset with reported values from the literature. Results: Our findings suggest that this automatically extracted CT dataset serves as a suitable reference standard for the verification and comparison of surface topography approaches. Conclusion: This work provides a robust framework for future verification studies of different surface topography methods and may facilitate the development of non-radiographic techniques for scoliosis assessment, thereby reducing reliance on ionizing radiation in clinical practice.

Introduction: The likelihood of developing preterm brain injury such as intraventricular hemorrhage (IVH) or cystic white matter injury (cWMI) largely depends on gestational age. However, its pathogenesis is multifactorial, and accounting of additional risk factors is important for preventing brain injury. Glucose is one of the most important energy sources for the brain. Both high and low blood glucose level may be associated with the onset and progression of brain disorders. Methods: To evaluate the relation between preterm brain injury (IVH/cWMI) and blood glucose level, univariate statistical analyses of retrospective data of 109 extremely and very preterm infants was performed. The analyzed data included 14 prenatal and infant diagnoses and 32 regularly measured parameters. Results: Mean values of whole blood glucose levels in preterm infants diagnosed with IVH or cWMI was higher than in controls indicating that elevated blood glucose level is associated with development of preterm brain injury. Mean values were higher both before and after diagnosis of IVH or cWMI, and even significantly exceeded the hyperglycemia threshold of 125 mg/dL following IVH diagnosis. Furthermore, a significant or moderate association was revealed between elevated blood glucose level and 8 medical diagnoses, as well as 21 routinely measured parameters. Conclusion: Our study emphasizes the importance of regularly monitoring blood glucose levels in preterm infants and keeping it within a safe range.

Background: Extremely and very preterm infants are at high risk for adverse neurodevelopmental outcomes. Early prediction remains challenging when relying on static clinical markers or single time-point neuroimaging. Serial cranial ultrasound (CUS) enables repeated bedside assessment of cerebral growth and may provide longitudinal trajectory biomarkers integrable with routine clinical data. Methods: In this retrospective two-center cohort study, 89 preterm infants (<32 weeks’ gestation and/or <1500 g birth weight) were assessed using the Bayley-III at 24 months corrected age. Brain volume trajectory features were derived from serial CUS using a standardized ellipsoid model. A three-level analytical framework was applied as follows: univariate regression (62 models, Bonferroni and Benjamini–Hochberg correction), multivariate SVM classification with five-fold GroupKFold cross-validation, ensuring patient-level data separation and feature importance analysis with interaction characterization using stratified Spearman correlation and two-dimensional partial dependence plots. Results: Multivariate classification yielded modest but above-chance performance (balanced accuracy 0.277–0.463, Cohen’s κ 0.042–0.152). Respiratory morbidity duration—mechanical ventilation and BPD severity—were the most robustly associated univariate predictors, surviving Bonferroni correction. Brain volume trajectory features showed no significant univariate associations but contributed conditionally within the multivariate framework as follows: the interaction between brain volume slope and trajectory linearity was the strongest for cognitive outcome (Δr = 0.47), and postnatal growth restriction showed amplified adverse effects at lower birth weight for motor outcome (Δr = 0.47). Conclusions: This study demonstrates the value of ML methods as structured analytical tools for characterizing predictor–outcome relationships in preterm neurodevelopment; respiratory morbidity and brain volume trajectory features emerged as the most informative predictor classes. Prospective multicenter validation is required before clinical translation.

Introduction: The neonatal period in very preterm born infants is characterized by high vulnerability of the developing brain. Diffuse or cystic white matter injury—cystic periventricular leukomalacia (cPVL)—is one of the most serious complications leading to brain injury in preterm infants, associated with an increased risk of neurological impairments such as cerebral palsy. Methods: In the present study, parameters that may potentially influence the occurrence of cPVL are analyzed. The study is based on retrospective clinical data of 46 preterm infants with and without cPVL, including prenatal factors, birth-related information, neonatal diagnoses and routinely measured clinical monitoring data. The novelty of this work lies in the inclusion in analyses of cerebral blood flow (CBF), CBF fluctuations and partial pressure of oxygen in brain tissue (PtO ₂ ), calculated from routinely measured parameters using a mathematical model. Results: Statistical analyses revealed significant association between cPVL and a group of 5 clinical and 23 regularly measured parameters including mathematically calculated CBF, CBF fluctuations and PtO ₂ . The obtained results may be useful for future risk assessment of cPVL at an early stage.

Background: Very Low Birth Weight preterm infants are at elevated risk for disrupted brain growth and later neurodevelopmental impairment. Bedside-accessible tools for monitoring cerebral development remain limited. Methods: In this retrospective pilot cohort study, 153 Very Low Birth Weight infants (3 to 275 cm³ across the hospital stay, corresponding to a median growth rate of 2.3 cm³/day (95% CI: 1.5–3.1). Duration of invasive mechanical ventilation was associated with reduced cerebral growth (p < 0.01, R² = 0.26). Cerebral volume growth showed a weak but statistically significant correlation with head circumference percentile progression (p < 0.05, ρ = 0.16). Conclusions: Sonographic brain volumetry is a feasible and non-invasive method for tracking cerebral development in Very Low Birth Weight infants. These findings confirm significant associations between cerebral growth and head growth and identify prolonged invasive ventilation as a risk factor for impaired cerebral development.

As the prevalence of machine-written texts grows, it has become increasingly important to distinguish between human- and machine-generated content, especially when such texts are not explicitly labeled. Current artificial intelligence (AI) detection methods primarily focus on human-like characteristics, such as emotionality and subjectivity. However, these features can be easily modified through AI humanization, which involves altering word choice. In contrast, altering the underlying grammar without affecting the conveyed information is considerably more challenging. Thus, the grammatical characteristics of a text can be used as additional indicators of its origin. To address this, we employ a newly developed rule-based parser to analyze the grammatical structures in human- and machine-written texts. Our findings reveal systematic grammatical differences between human- and machine-written texts, providing a reliable criterion for the determination of the text origin. We further examine the stability of this criterion in the context of AI humanization and translation to other languages.

Background/Objectives: Patients with physical disabilities, like cerebral palsy, the most common movement disorder in childhood, can benefit from instrumental therapy using piano. Playing the piano promotes the interaction between different brain regions and integrates motor skills, sensory skills, musical hearing, and emotions. A pilot music study examined the effects of six months of piano lessons on hand motor skills and musical hearing in groups of children with motor disabilities. Methods: The allocation to the group was not randomized. Various tests, including the standardized Box and Block Test (BBT) and piano tests, assessed hand motor skills. Musical hearing was evaluated, and a questionnaire was used to determine the participants’ enjoyment and experience with the piano lessons. The regularity, tempo of keystrokes, and synchronization between the two hands were assessed and compared to evaluate the effects of six months of piano training. Results: After six months of piano training, statistically significant improvements were observed in the BBT, as well as in the regularity and tempo of the non-dominant hand. The children showed significant improvement in hand-motor control, moving 27.3% more cubes in the BBT. Regularity and tempo in piano playing, especially in the non-dominant hand, also improved. Moreover, 55% of the children better recognized the correct pitches of notes. Conclusions: Thus, this study supports the concept that piano lessons are an effective form of physical therapy for the development of hand motor skills and musical hearing.

The “Easy Language” standard was created to help individuals with cognitive disabilities understand texts more easily. Typically, text simplification is performed by language experts and is available for limited materials. We introduce a new software tool designed to analyze and simplify any text according to the “Easy Language” rules. This tool uses a rule-based system, conducting a full grammatical analysis of each sentence and then simplifying it into a grammatically correct form. Unlike neuronal approaches, which are based on statistics and are very popular today, our rule-based approach explicitly addresses language ambiguities by examining all possible interpretations and eliminating the incorrect ones. The purpose of the present study is to compare the performance of our rule-base parser with two state-of-the-art statistical parsers, one based on dependencies between words (SpaCy parser) and the other based on linguistic constituents (Stanford parser). Although large language models (LLMs), which are the technical basis of the software ChatGPT, were not designed specifically for grammatical parsing, because of their popularity, users, especially language learners, often ask them grammatical questions as well. Therefore, we use LLMs as supplementary models for comparison. LMMs produce grammatically correct text on any topic; however, their grammar knowledge is implicit within the trained weights. To evaluate how well state-of-the-art methods can perform a grammatical analysis, we parse ten sentences with our tool, the statistical parsers from SpaCy and Stanford, and ask two LLMs equivalent grammar questions. The results show that our rule-based method provides a more informative and reliable grammatical analysis compared to these two parsers and outperforms LLMs in that specific task.

Intraventricular hemorrhage (IVH)4 is one of the most threatening neurological complications associated with preterm birth which can lead to long-term sequela such as cerebral palsy. Early recognition of IVH risk may prevent its occurrence and/or reduce its severity. Using multivariate logistic regression analysis, risk factors significantly associated with IVH were identified and integrated into risk scales. A special aspect of this study was the inclusion of mathematically calculated cerebral blood flow (CBF) as an independent predictive variable in the risk score. Statistical analysis was based on clinical data from 254 preterm infants with gestational age between 23 and 30 weeks of pregnancy. Several risk scores were developed for different clinical situations. Their efficacy was tested using ROC analysis, and validation of the best scores was performed on an independent cohort of 63 preterm infants with equivalent gestational age. The inclusion of routinely measured clinical parameters significantly improved IVH prediction compared to models that included only obstetric parameters and medical diagnoses. In addition, risk assessment with numerically calculated CBF demonstrated higher predictive power than risk assessments based on standard clinical parameters alone. The best performance in the validation cohort (with AUC = 0.85 and TPR = 0.94 for severe IVH, AUC = 0.79 and TPR = 0.75 for all IVH grades and FPR = 0.48 for cases without IVH) was demonstrated by the risk score based on the MAP, pH, CRP, CBF and leukocytes count.