A Physics-Flavored Transformer Network for Parametrizing Contraction Dynamics of Engineered Skeletal Muscle Tissues

The paper introduces a Physics-Flavored Neural Network (PFNN) that embeds a stretched-exponential physical model in a CNN-Transformer to extract interpretable kinetic parameters from force-time profiles of engineered skeletal muscle tissues. To mitigate limited labeled biological data, the system first learns from synthetic signals and then performs unsupervised self-alignment on unlabeled real measurements. The supplied abstract claims high-fidelity parameterization across several contractile phenotypes and cell lines, including Duchenne muscular dystrophy models, but reports no sample counts, numerical errors, baseline comparisons, or external-validation results.