Sensor glove: A textile glove worn by the therapist to send commands and adjust glove settings remotely.

Actuating glove: A pneumatic soft robotic glove worn by the patient to receive real-time feedback or haptic feedback through actuators embedded in the glove.

Gloves:

1. Have flexible sensors and inertial measurement units that detect joint angles, movement patterns, and grip force.

2. The sensor data are transmitted wirelessly a a cloud platform.

3. Stores data that enables monitoring progress and adjusting therapy protocols.

A Machine learning dataset was trained from data collected from 12 subjects, and the actuating T‐IoT glove moves according to these states.

📊 Results

Movement recognition: It was highly accurate across multiple finger movements.

Latency: Average system response 48 milliseconds. Accuracy 93.4%. This makes it suitable for real-time telerehabilitation.

High usability.

The gloves have the potential to support telerehabilitation, allowing clinician-guided hand therapy remotely.

The system was fast, accurate, and reliable.

It can be beneficial for patients with limited mobility or those living in remote areas.

Timeframe: 2015–2025.

Studies included: 62 publications.

Technologies reviewed:

Multi-sensor platforms and advanced smart insoles can monitor humidity, microcirculation, vascular status, and pH at the foot surface or within wounds, supporting comprehensive risk assessment and wound monitoring.

Pressure/temperature sensors: Demonstrated high accuracy (80–95%) for detecting abnormal plantar loading and hotspots associated with ulcer risk.

Smart socks/insoles: Can be effective for continuous monitoring of plantar pressure, shear forces, monitor blood oxygen saturation, and risk of ischemia.

Smart bandages: These are integrated with temperature and moisture sensors for wound monitoring.

AI integration: Machine learning algorithms improved risk prediction and personalized alerts.

These devices often also track step count and adherence to offloading regimens.

Smart wearable devices can:

Evidence for effectiveness is encouraging, but larger randomized clinical trials need to be done.

Wearable sensors can prevent limb loss and enhance the quality of life of patients by protecting them from preventable complications.

NextStep Robotics: a wearable device for angle impairment, like foot drop from stroke and other pathologies. Assistance can be graded (Link).

Impakt: An app that uses your smartphone camera to track body movements and guides you through the workouts. It analyzes your performance in real-time and gives you feedback (Link).

TutorAIme: Do you want to learn something and don’t know where to start? This AI platform allows users to learn about any topic. Has a database of millions of articles, videos, and resources to help create a course (Link).

Echodocs.ai Turn your audio into documents instantly. This tool makes transcription and documentation easy, and works in over 50 languages (Link).

ThinkRare: an algorithm to identify patients with undiagnosed rare genetic disease, checking thousands of electronic medical records (Link).

Google developed an AI-powered personal health coach for the Fitbit. Works as a fitness trainer, a sleep coach, and a health advisor (Link).

Hospitals invest in AI despite doubts about readiness: Report (Link).

A meta-analysis that found AI systems had high diagnostic accuracy for ischemic and hemorrhagic strokes on MRI and CT scans, almost like human radiologists (Link).

HLTH AI & Innovation Healthcare Event in Las Vegas, October 19-22 (Link).