Marta García-Díez

Ultrasound evaluation of the median nerve in carpal tunnel syndrome (CTS) is inherently operator-dependent. Automating CTS severity grading from ultrasound using deep learning may improve objectivity, reproducibility, and scalability in clinical practice. Background/Objectives: To develop convolutional neural network (CNN) models for automated CTS severity classification using B-mode ultrasound and electrodiagnostic results as the reference standard, and to compare performance between ultrasound scans. Methods: Fifty participants with suspected CTS provided 94 wrists, each classified into four severity levels (normal, mild, moderate, severe) through standardized nerve conduction studies. Ultrasound videos in transverse and longitudinal planes were acquired under a uniform protocol. From 11,895 frames, expert quality control produced a dataset of 2,518 valid images. Two CNN classifiers (transverse and longitudinal) were trained on nerve-centred crops using a 75/25 training–validation split, Adam optimization, and categorical cross-entropy loss. Performance was assessed through accuracy, class-wise precision, and confusion matrices. Results: Both models showed stable convergence. The transverse classifier achieved 0.92 validation accuracy, with precision values of 0.88 (normal), 0.94 (mild), 0.93 (moderate), and 0.97 (severe). The longitudinal classifier achieved 0.94 accuracy, with precision values of 0.98 (normal), 0.94 (mild), 0.96 (moderate), and 0.85 (severe). Misclassifications occurred mainly between adjacent categories, and reduced performance for severe longitudinal cases reflected limited sample representation. Conclusions: CNN-based classifiers can automatically predict CTS severity from B-mode ultrasound with high agreement to electrodiagnostic labels. These operator-independent models may support standardized diagnostic pathways and ultrasound-based screening. Future work should expand and balance datasets, incorporate multi-center data, and conduct external validation.

Background/Objectives: Electrical threshold testing (ETT) offers a promising method for assessing somatosensory function. Despite its growing use, fundamental aspects such as the physiological recovery time required between repeated threshold measurements remain poorly understood. This gap is critical when evaluating sensory, motor, or pain thresholds (EST, EMT, EPT) in pre–post designs or rapid intra-session protocols. The aim is to investigate the short-term recovery dynamics of electrical thresholds following electrical threshold testing, and to determine the minimum interval required for values to return to a stable baseline. Methods: In this pilot, repeated-measures study, 10 healthy adults (20 upper limbs) underwent three progressive stimulation trials (sensory, motor, and pain). Electrical thresholds were assessed at fixed recovery intervals (0–120 s), with duplicate measurements at each time point. Stability was defined as the absence of significant differences between repeated measures. Results: EST stabilized rapidly after sensory or motor stimulation, showing no significant differences beyond 0 and 15 s, respectively. Within pain stimulation, EST recovered at 60 s. EMT showed immediate recovery with motor stimulation and required longer recovery with pain stimulation, with stabilization observed at 90 s. EPT exhibited the highest variability, with the smallest time-dependent differences observed immediately after the first assessment. Conclusion: Recovery time after electrical stimulation varies by threshold type and intensity of the stimuli. EST and EMT can be reliably reassessed immediately after sensory and motor stimulation, respectively. However, when stimulation reaches EPT level, EST requires 60 s to recover and EMT needs 90 s. EPT demonstrates higher variability, indicating the need for further investigation. These findings support the implementation of standardized recovery intervals in ETT and underscore the importance of interpreting EPT results with caution during rapid assessments.