Biosignals are rich, messy, and highly variable. In this tutorial, we offer a practical entry point into AI/ML for biosensing by starting from what makes signals like electroencephalography (EEG) and photoplethysmography (PPG) different from typical datasets: their frequency–magnitude structure, noise and artifacts, and strong variability across people and hardware (often experienced as domain shift). We show how this domain knowledge should shape the way you represent signals and choose models.

We then use statistical analysis as the bridge from raw waveforms to meaningful features, connecting intuitive signal properties to time-domain statistics and frequency-domain summaries that often carry the most useful information. From there, we explain how classical machine learning uses these engineered features, how deep learning can learn representations directly from raw inputs, and how hybrid pipelines can combine both worlds. We also discuss practical, lightweight approaches to interpretability and explainability that help researchers and practitioners trust and debug biosignal models.

Finally, we focus on multi-task learning as an extension for many biosensing problems. We cover how to define tasks that help rather than fight each other, why multi-loss training can become unstable, and how adaptive loss weighting can reduce manual tuning and keep training balanced. Participants will leave with concrete pitfalls to avoid and practical takeaways they can apply immediately.

Rapid decentralized disease diagnosis by non- or minimally- invasive means such as microlitre blood sample or urine/saliva or detection of the newer variety of disease biomarkers such as circulating cell-free DNA demand diagnostic technologies with an attomolar analyte detection capability. The past two decades have seen a tremendous growth in the biosensor strategies that offer such ultra-high sensitivities. A variety of optical and electrochemical transduction techniques aided by nanomaterials (e.g. plasmonic nanoparticles and graphene) and AI/ML have been demonstrated.

This tutorial briefly reviews some of the reports on attomolar labelled and label-free biosensing strategies highlighting the working principles and limitations on the scale-up and high-throughput applications. Subsequently, a detailed discussion will be devoted to the plasmonic fiber optic biosensing solutions in particular. Gratings and modified-geometry based fiber optic sensors and their advantages and limitations will be covered.

Monitoring biological systems is crucial in healthcare, driving the need for reliable and non-invasive solutions. The proliferation of unverified drugs in the market necessitates reliable methods for their detection and identification, especially amidst advancements in pharmaceuticals. Plasmonic biosensors emerge as a great platform for ultra-sensitive detection, identification, and manipulation of biomolecular systems. This tutorial will present the critical need for precise detection and monitoring of biomolecules and drugs, presenting innovative solutions through the design of a plasmonic biosensor to tackle challenges in sensitivity, selectivity, and label-free detection and identification. The tutorial will elaborate a robust and tunable, cavity-integrated plasmonic nanopatterned sensor that exhibits superchiral light in the infrared domain for ultrasensitive detection of chiral molecular concentrations and enantiomeric excesses. The multispectral capability of this system is further harnessed to generate unique chiral fingerprint-based barcodes for the identification of diverse chiral drugs and biomolecules. The tutorial will further discuss and demonstrate results for a surface-modified plasmonic biosensor operating in the visible-near-infrared realm in detecting viral biomarkers and neurotransmitters directly from complex physiological environments. The system, on coupling with a microfluidic flow setup allows sensitive, selective and rapid detection without requiring complex pre-processing or sample preparation steps. We will discuss additional applications of the unique plasmonic sensor, utilizing the property of tunable superchirality to create a dynamic chirality tracking system operating in the near infrared for real-time monitoring of protein dynamics. These techniques aim to revolutionize bio-detection, chiral differentiation, and sorting processes, having extensive applications in medical research and pharmaceutical industries.