Clinical Research Use Only

Automated Dysarthria Speech Analysis

NeuroClear leverages state-of-the-art acoustic feature extraction and generative AI to provide objective, multi-domain motor speech assessments in seconds.

Start Diagnostic Screening Patient Practice

Rigorous Machine Learning Pipeline

Built on a decoupled architecture combining deep learning classifiers with digital signal processing.

Phonetic Alignment

Utilizes Wav2Vec2 Connectionist Temporal Classification (CTC) to extract log-probability confidence scores, identifying micro-articulatory decay at the phoneme level.

DSP Tremor Detection

SciPy-driven amplitude envelope extraction. Our peak-finding algorithms isolate involuntary motor tremors and false-starts independent of language context.

LLM Clinical Synthesis

Passes objective prosody and regression metrics into a strictly constrained Gemini schema to output reliable, structured clinical diagnostic rationales.