3D fluid-structure interaction modelling was utilized for simulation of 13 normal subjects, 11 non-communicating hydrocephalus (NCH) patients at pre-treatment phase, and 3 patients at five post-treatment phases. Evaluation of ventricles volume and maximum CSF pressure (before shunting) following results validation indicated that these parameters were the most proper hydrodynamic indices and the NCH type doesn’t have any significant effect on changes in two indices. The results confirmed an appropriate correlation between these indices although the correlation decreased slightly after the occurrence of disease. NCH raises the intensity of vortex and pulsatility (2.4 times) of CSF flow while the flow remains laminar. On day 18 after shunting, the CSF pressure decreased 81.0% and all clinical symptoms of patients vanished except for headache. Continuing this investigation during the treatment process showed that maximum CSF pressure is the most sensitive parameter to patients’ clinical symptoms. Maximum CSF pressure has decreased proportional to the level of decrease in clinical symptoms and has returned close to the pressure range in normal subjects faster than other parameters and simultaneous with disappearance of patients’ clinical symptoms (from day 81 after shunting). However, phase lag between flow rate and pressure gradient functions and the degree of CSF pulsatility haven’t returned to normal subjects’ conditions even 981 days after shunting and NCH has also caused a permanent volume change (of 20.1%) in ventricles. Therefore, patients have experienced a new healthy state in new hydrodynamic conditions after shunting and healing. Increase in patients’ intracranial compliance was predicted with a more accurate non-invasive method than previous experimental methods up to more than 981 days after shunting. The changes in hydrodynamic parameters along with clinical reports of patients can help to gain more insight into the pathophysiology of NCH patients.

Dataset imbalances pose a significant challenge to predictive modeling in both medical and financial domains, where conventional strategies, including resampling and algorithmic modifications, often fail to adequately address minority class underrepresentation. This study theoretically and practically investigates how the inherent nature of medical data affects the classification of minority classes. It employs ten machine and deep learning classifiers, ranging from ensemble learners to cost-sensitive algorithms, across comparably sized medical and financial datasets. Despite these efforts, none of the classifiers achieved effective classification of the minority class in the medical dataset, with sensitivity below 5.0% and area under the curve (AUC) below 57.0%. In contrast, the similar classifiers applied to the financial dataset demonstrated strong discriminative power, with overall accuracy exceeding 95.0%, sensitivity over 73.0%, and AUC above 96.0%. This disparity underscores the unpredictable variability inherent in the nature of medical data, as exemplified by the dispersed and homogeneous distribution of the minority class among other classes in principal component analysis (PCA) graphs. The application of the synthetic minority oversampling technique (SMOTE) introduced 62 synthetic patients based on merely 20 original cases, casting doubt on its clinical validity and the representation of real-world patient variability. Furthermore, post-SMOTE feature importance analysis, utilizing SHapley Additive exPlanations (SHAP) and tree-based methods, contradicted established cerebral stroke parameters, further questioning the clinical coherence of synthetic dataset augmentation. These findings call into question the clinical validity of the SMOTE technique and underscore the urgent need for advanced modeling techniques and algorithmic innovations for predicting minority-class outcomes in medical datasets without depending on resampling strategies. This approach underscores the importance of developing methods that are not only theoretically robust but also clinically relevant and applicable to real-world clinical scenarios. Consequently, this study underscores the importance of future research efforts to bridge the gap between theoretical advancements and the practical, clinical applications of models like SMOTE in healthcare.

Bruxism is among the habits considered generally as contributory factors for temporomandibular joint (TMJ) disorders and its etiology is still controversial. Three-dimensional models of maxilla and mandible and teeth of 37 patients and 36 control subjects were created using in-vivo image data. The maximum values of stress and deformation were calculated in 21 patients six months after using a splint and compared with those in the initial conditions. The maximum stresses in the jaw bone and head of mandible were respectively 4.4 and 4.1 times higher in patients than in control subjects. Similar values for deformation were 5.8 and 4.9, respectively. The maximum stress in the jaw bone and head of mandible decreased six months after splint application by up to 71.0 and 72.8%, respectively. Similar values for the maximum deformation were 80.7 and 78.7%, respectively. Following the occlusal splint therapy, the approximation of maximum deformation to the relevant values in control subjects was about 2.6 times the approximation of maximum stress to the relevant values in control subjects. The maximum stress and maximum deformation occurred in all cases in the head of the mandible and the splint had the highest effectiveness in jaw bone adjacent to the molar teeth. Splint acts as a stress relaxer and dissipates the extra stresses generated as well as the TMJ deformation and deviations due to bruxism. The splint also makes the bilateral and simultaneous loading possible and helps with the treatment of this disorder through regulation of bruxism by creating a biomechanical equilibrium between the physiological loading and the generated stress.

Three-D head geometrical models of eight healthy subjects and 11 hydrocephalus patients were built using their CINE phase-contrast MRI data and used for computer simulations under three different inlet/outlet boundary conditions (BCs). The maximum cerebrospinal fluid (CSF) pressure and the ventricular system volume were more effective and accurate than the other parameters in evaluating the patients’ conditions. In constant CSF pressure, the computational patient models were 18.5% more sensitive to CSF volume changes in the ventricular system under BC “C”. Pulsatile CSF flow rate diagrams were used for inlet and outlet BCs of BC “C”. BC “C” was suggested to evaluate the intracranial compliance of the hydrocephalus patients. The results suggested using the computational fluid dynamic (CFD) method and the fully coupled fluid-structure interaction (FSI) method for the CSF dynamic analysis in patients with external and internal hydrocephalus, respectively.

Three-dimensional computational models of the cerebrospinal fluid (CSF) flow and brain tissue are presented for evaluation of their hydrodynamic conditions before and after shunting for seven patients with non-communicating hydrocephalus. One healthy subject is also modeled to compare deviated patients data to normal conditions. The fluid-solid interaction simulation shows the CSF mean pressure and pressure amplitude (the superior index for evaluation of non-communicating hydrocephalus) in patients at a greater point than those in the healthy subject by 5.3 and 2 times, respectively.