Vol 7, No 2 (2025)

BACKGROUND: Development of a universal predictive tool for patients with stroke remains a challenge.

AIM: The study aimed to develop machine learning–based models that could predict functional outcomes from the first day after stroke. The models were trained using clinical and anamnestic predictors, and functional outcomes were evaluated using the National Institutes of Health Stroke Scale (NIHSS) and the modified Rankin Scale (mRS) at hospital discharge.

METHODS: Models based on artificial neural network (ANN) and random forest (RF) algorithms were developed using a database created from 5,225 records of patients with stroke discharged from neurological departments. Twenty-one parameters were used, including patient demographics; baseline National Institutes of NIHSS and mRS scores; stroke type; time from stroke onset to hospitalization; comorbidities; and emergency revascularization. Outcomes were predicted using NIHSS and mRS scores. The algorithms solved the classification problem for multiple sets of outcome values. Model I had 26 classes (NIHSS score pf 0–25), while model II had 6 classes (mRS score of 0–5). The quality of the models was evaluated using the area under the receiver operating characteristic curve (ROC-AUC). The contribution of each predictor was evaluated using the SHapley Additive exPlanations (SHAP).

RESULTS: The predictive value of the ANN was determined based on the AUC-ROC: 0.771 for model I and 0.844 for model II. The RF AUC-ROC was 0.778 for model I and 0.845 for model II. The RF algorithm was chosen for further work due to its better interpretability. The most significant features that influenced the predicted outcomes were baseline NIHSS and mRs scores, patient age, time from stroke onset to admission, and stroke type. When RF performance was tested on an external validation set of 783 records, ROC-AUC values were 0.786 for model I and 0.774 for model II. A calculator was developed for practical use.

CONCLUSION: The proposed RF-based models can reliably predict an early functional outcome within the first day of stroke onset, using NIHSS and mRs scores and clinical and anamnestic predictors. This tool can be used to develop personalized therapeutic and rehabilitation strategies in the acute phase of a stroke. These models are versatile to be used in rural and remote healthcare organizations that lack specialized staff and diagnostic equipment.

BACKGROUND: After a stroke, patients often have impaired knee flexion of the paretic limb, which is one of the most common issues affecting walking ability. Biofeedback on joint kinematics is a relatively new technology that requires further research to understand its capabilities.

AIM: The study aimed to evaluate the potential of using biofeedback to improve knee flexion range of motion in patients with hemiparesis during the early recovery period after ischemic cerebral stroke.

METHODS: The main group of patients (n = 11; hemiparesis; early recovery period after ischemic cerebral stroke) completed a biofeedback training program for knee flexion range of motion, as well as a personalized rehabilitation program. The comparison group (n = 11) had the same selection criteria but received only standard rehabilitation. A biomechanical gait evaluation and clinical evaluation using various scales were performed before and after treatment. The control group included practically healthy volunteers (n = 34).

RESULTS: A clinical evaluation using the scales showed significant improvement (p < 0.05), as did two domains of the International Classification of Functioning, Disability, and Health. General functional patterns typical of hemiparetic gait were observed: decreased ability to support weight on the paretic limb, impaired reciprocity, and step-to-step asymmetry. In addition, a decrease in joint flexion range of motion was observed due to the paretic limb becoming relatively longer. Muscle activity decreased in nearly all muscle groups, particularly the calf muscles. This decrease reflected a reduction in ankle flexion range of motion and a condition known as drop foot. The absence of high-amplitude flexion was due to the weakness of the posterior thigh muscles. Analysis of step cycle phases in the main group revealed significant decreases (p < 0.05) in the contralateral support and total double support phases. Reduced support on the healthy side led to a decreased load because both limbs provide more uniform support. This type of biofeedback training significantly increased (p < 0.05) the swing range of motion of the paretic knee during weight transfer. This range of motion was the target of biofeedback training. The comparison group showed no significant change in the range of motion.

CONCLUSION: Biofeedback training lasting 20 minutes for 10 days helps to reduce step asymmetry, increase endurance, and improve safety when walking. Biofeedback training is a promising noninvasive, non-drug treatment option for poststroke disorders. It has minimal contraindications and can be used to restore knee joint flexion range of motion and improve the support capacity of the paretic limb in patients during the early recovery period after ischemic stroke.

Patients with motor and/or cognitive disorders, a history of stroke or head/spinal cord injury, chronic pain, neurodegenerative diseases, or neurological complications of severe COVD-19 may not benefit from pharmacotherapy or traditional rehabilitation. Virtual reality is an innovative therapeutic approach that restores lost functions. This interactive technology uses computer modeling to create a virtual world. Virtual reality provides multisensory stimulation, activates restorative neural mechanisms, and produces an analgesic effect. In recent years, virtual reality has become more prevalent in neurology. It improves motor and cognitive outcomes in patients with various neurological and neuropsychological dysfunctions. These improvements are the result of neuroplasticity and neurogenesis in brain lesions. Virtual reality–based rehabilitation can be used alone or in combination with other treatment options. The most effective treatment combines exercise therapy, physical therapy, psychological counseling, standard cognitive training, and training with immersive virtual reality technology. This technology creates a sense of presence in a three-dimensional digital environment. With head-mounted displays and body monitoring sensors, users can interact naturally with virtual objects. Systematic reviews and meta-analyses have shown that depending on the initial condition, most patients demonstrate improvement in voluntary range of motion, balance, cognitive function, mood, quality of life, anxiety, and pain. Therefore, virtual reality–based therapy is a promising rehabilitation option for patients with neurological symptoms. Virtual reality–based neurological rehabilitation allows for the personalized selection of exercise complexity and intensity, thereby increasing patient compliance through gamification. However, further research is needed to standardize virtual reality–based modalities, clarify inclusion and exclusion criteria for clinical trials, optimize protocols, and evaluate long-term effects.

The term “chronic critical illness” was first coined in 1985 by researchers at Stanford University School of Medicine in an article with the provocative headline “Save or let die?” The authors described a unique category of patients that stood out among the general population of patients in intensive care units. These patients, who received a full range of intensive care, did not die, but also did not achieve recovery for weeks, months, or even years after an acute Injury. This research opened doors for further study.

National and foreign sources show that patients with chronic critical illness comprise 5–20% of intensive care unit patients. Despite intensive treatment and rehabilitation measures, most of these patients eventually develop homeostasis disorders and complications, leading to multiple organ failure and death.

This article presents an intermittent hypoxia-hyperoxia therapy case study for a patient with the chronic disorder of consciousness. The patient completed two weeks of intensive care therapy at the Scientific Research Institute of rehabilitation named after Prof. Pryanikova I.V. of the Federal Research and Clinical Center of Intensive Care Medicine and Rehabilitology. Metered-dose intermittent hypoxia-hyperoxia therapy training improved the patient’s adaptation and reduced neurological deficits as a part of comprehensive rehabilitation.

Intermittent hypoxic-hyperoxia therapy shows potential in the intensive care unit patient treatment and in rehabilitation.