Identification and interpretation of gait analysis features and foot conditions by explainable AI.

Clinical gait analysis is a crucial step for identifying foot disorders and planning surgery. Automating this process is essential for efficiently assessing the substantial amount of gait data. In this study, we explored the potential of state-of-the-art machine learning (ML) and explainable artificial intelligence (XAI) algorithms to automate all various steps involved in gait analysis for six specific foot conditions. To address the complexity of gait data, we manually created new features, followed by recursive feature elimination using Support Vector Machines (SVM) and Random Forests (RF) to eliminate low-variance features. SVM, RF, K-nearest Neighbor (KNN), and Logistic Regression (LREGR) were compared for classification, with a Majority Voting (MV) model combining trained models. KNN and MV achieved mean balanced accuracy, recall, precision, and F1 score of 0.87. All models were interpreted using Local Interpretable Model-agnostic Explanation (LIME) method and the five most relevant features were identified for each foot condition. High success scores indicate a strong relationship between selected features and foot conditions, potentially indicating clinical relevance. The proposed ML pipeline, adaptable for other foot conditions, showcases its potential in aiding experts in foot condition identification and planning surgeries.

Machine learning-based prediction of joint moments based on kinematics in patients with cerebral palsy.

Joint moments during gait provide valuable information for clinical decision-making in patients with cerebral palsy (CP). Joint moments are calculated based on ground reaction forces (GRF) using inverse dynamics models. Obtaining GRF from patients with CP is challenging. Typically developed (TD) individuals' joint moments were predicted from joint angles using machine learning, but no such study has been conducted on patients with CP. Accordingly, we aimed to predict the dorsi-plantar flexion, knee flexion-extension, hip flexion-extension, and hip adduction-abduction moments based on the trunk, pelvis, hip, knee, and ankle kinematics during gait in patients with CP and TD individuals using one-dimensional convolutional neural networks (CNN). The anonymized retrospective gait data of 329 TD (26 years ± 14, mass: 70 kg ± 15, height: 167 cm ± 89) and 917 CP (17 years ± 9, mass:47 kg ± 19, height:153 cm ± 36) individuals were evaluated and after applying inclusion-exclusion criteria, 132 TD and 622 CP patients with spastic diplegia were selected. We trained specific CNN models and evaluated their performance using isolated test subject groups based on normalized root mean square error (nRMSE) and Pearson correlation coefficient (PCC). Joint moments were predicted with nRMSE between 18.02% and 13.58% for the CP and between 12.55% and 8.58% for the TD groups, whereas with PCC between 0.85 and 0.93 for the CP and between 0.94 and 0.98 for the TD groups. Machine learning-based joint moment prediction from kinematics could replace conventional moment calculation in CP patients in the future, but the current level of prediction errors restricts its use for clinical decision-making today.

Stereotactic excision of additional lesions detected with intraoperative ultrasound examination during radiofrequency dissecting sealar (habib®) assisted hepatic metastasectomy: report of 4 cases.

Intraoperative ultrasound has been using to achieve a proper resection strategy in patients undergoing a hepatic colorectal metastasectomy. This study aims to describe and reveal the place of stereotactic metastasectomy in nonpalpable colorectal liver metastases (CLM). A chart review was initiated for all patients underwent resection for CLM between 2006 and 2011. The data concerning perioperative data and intraoperative strategy were abstracted. Among the 58 patients, who underwent a resection for CLM, 4 (6.9 %) (all men, median age 65.5, range 49-72, years) necessitated a stereotactic metastasectomy. Preoperative evaluations showed 1 (n = 1), 2 (n = 2), or 3 (n = 1) lesions, and intraoperative ultrasound (IUS) found an additional lesion in a case. Stereotactic marking was performed for nonpalpable lesions located in segments IVA, II, and VI and at the junction of segments V and VI. The margins were negative for all lesions both resected with conventional and stereotactic techniques. The examinations of the stereotactic resection materials revealed metastatic adenocarcinoma (patients n = 2), focal nodular hyperplasia (n = 1), and abnormal benign liver histology probably induced by chemotherapy (n = 1). The median (range) operation and hospitalization periods were 217.5 (150-310) minutes and 5.5 (2-9) days. No complications were observed except biliary fistula in a case, which spontaneously disappeared within 2 weeks. A patient died due to systemic disease including hepatic metastases 33 months after the liver surgery. Stereotactic metastasectomy may be feasible for the removal of nonpalpable CLM. Further evaluations are necessitated to understand the accurate place of this novel technique.

5-Fluorouracil induces arterial vasoconstrictions but does not increase angiotensin II levels.

Because the mechanisms of 5-Fluorouracil (5-FU) cardiotoxicity have not yet been completely identified, prophylactic options are not available. To our knowledge, there are no published data investigating the use of angiotensin converting enzyme (ACE) inhibitors for 5-Fluorouracil-associated cardiotoxicity. In this study, we aimed to evaluate the influence of 5-FU administration on the diameter of the brachial artery and the levels of angiotensin II. The patients were administered bolus 5-FU/leucovorin in the study group. Angiotensin II and troponin T assays, complete blood cell counts, hepatic and renal function tests were analyzed in five consecutive blood samples in the initiation, just after termination, and on 24, 48, and 72 h after termination of the regimen. Pre- and post-treatment angiotensin II and troponin T assays, complete blood cell counts, hepatic and renal function tests were also analyzed in the control group. Brachial arterial diameters were measured and recorded in all patients before and after the treatment. A total of 59 patients were included in this study. Thirty one out of 59 patients (52.5%) were in the 5-FU study group and the remaining 28 patients (47.5%) were in the control group. Basal and post-treatment brachial artery diameters in the 5-FU study group were 0.436 +/- 0.51 and 0.423 +/- 0.50 cm, respectively (P = 0.001). The corresponding values in the controls were 0.3954 +/- 0.50 and 0.3957 +/- 0.49 cm, basal and post-treatment, respectively (P = 0.979). Angiotensin II levels were not changed significantly at serial measurements (P = 0.496). Moreover, the corresponding measurements were not statistically different in both two groups treated with and without 5-FU (P = 0.372). The pathophysiology of 5-FU-induced cardiac toxicity has not yet been elucidated. In the present study, 5-FU-associated vasoconstriction was not dependent on angiotensin II levels, thus we suggest that the prophylactic administration of ACE inhibitors cannot prevent cardiotoxicity in these patients. The underlying mechanisms of cardiotoxicity related to 5-FU might be multifactorial; nevertheless, further prospective investigation for the toxic effects of fluoropyrimidines on the coronary endothelium and myocardium are needed.

Spontaneous migration of a bullet into the brain.

We report the case of a 20-year-old man with a gunshot injury as an example of spontaneous migration of a metallic foreign body within the brain. Computed tomography (CT) showed the bullet in the left temporoparietal region. At 10 days follow-up, CT revealed that the bullet had migrated posteriorly, due to the effect of gravity, lodging in the occipital lobe. Although there are a few literature reports of spontaneous migration of a bullet within the brain, this case was unique as the patient was fully conscious without any neurological deficit.

[Cranial MRI findings in epileptic children]. / Epilepsili çocuklarda kranyal MRG bulgulari.

PURPOSE: To investigate the cranial MRI findings in childhood epilepsy and to detect the most frequent cerebral lesions in epileptic children in our region. MATERIALS AND METHODS: In this study, 98 epileptic children were examined with 1 T MRI equipment between January 1997 and March 2001. T1-weighted sagittal, T2 and proton weighted axial and T2-weighted coronal MR images were obtained. Contrast-enhanced T1-weighted axial, sagittal and coronal MR images were also examined in patients with suspicion of infection or intracranial mass. RESULTS: Among a total of 98 patients, 65 (66.3%) had no pathological findings on MRI. The pathological findings were diffuse cerebral atrophy (18.4%), delay in myelination and dismyelination (12.4%), ischemia and infarct (8.2%), arachnoid cysts (6.2%), tumoral mass (5.1%), mesial temporal sclerosis (1.0%) and sequelae of kernicterus (1.0%). CONCLUSION: MRI has been found to be effective in the evaluation of cerebral structures and imaging the pathologies in the etiology of epilepsy and other lesions in childhood focal epilepsy. However, MRI does not always appear to be successful in detecting the epileptic focus.

Timing of shunt surgery in childhood tuberculous meningitis with hydrocephalus.

Hydrocephalus is a common complication of tuberculous meningitis (TBM) in children. The aims of this study are to review our experience of hydrocephalus in childhood TBM and to evaluate the effect of the timing of ventriculoperitoneal shunting (VPS) on the final outcome. In this study, 156 patients with TBM and hydrocephalus were reviewed retrospectively between 1990 and 2000. Patients' ages ranged from 6 months to 15 years, with a mean age of 4.1 years. There were 85 boys, and the male-to-female ratio was 1.19:1.0. Sixty-two percent of the children were younger than 6 years old. VPS was performed 2 days after the diagnosis in 100 patients, and in the remaining 56 patients, 3 weeks after the diagnosis. The average follow-up period was 8.5 months. Good recovery or minor sequelae was seen in 82 patients (52.6%), and 51 died (12.3%). The timing of the VPS procedure and cerebral complications had an effect on the final outcome. Early VPS gave a better outcome in mild and moderate hydrocephalus (p = 0.040). This study has shown that early surgical procedure for mild/moderate hydrocephalus has a positive effect on the morbidity and mortality of hydrocephalus in childhood TBM (p = 0.014, p = 0.040, respectively). In severe hydrocephalus, there was a tendency for early shunting to have a positive effect on morbidity, although this did not reach statistical significance.

Analysis of 107 civilian craniocerebral gunshot wounds.

In this study, we present a retrospective analysis of 107 cases due to civilian craniocerebral gunshot wounds that were treated by the medical faculty of Dicle University during a period of 7 years (January 1993 to January 2000). Twenty patients died at the hospital, and the deaths were determined to result from direct effects of brain damage. Coma was the best prognostic guideline. Diffuse brain damage and ventricular injury, particularly infections, were associated with poor outcome.