At Telemed Solutions, we've established a simple approach to medical device development: Design clinically and technologically superior medical devices, and place them in the hands of caring medical professionals.
Our company is comprised of individuals who possess an average of 25 years experience developing products for the cardiology field. Our specialties are software development and embedded device design.
We are working with the University of Glasgow in a joint venture to combine our algorithm technologies for use in Resting 12-Lead ECG. We also plan on using this joint technology in products currently in development. We have a contract with Glasgow who has granted us permission to sublicense this joint technology to others. We can deploy this technology on embedded, desktop, and enterprise systems. This innovative technology is based on:
Advanced Signal Processing - Time-frequency analysis: Stationary, discrete, and continuous wavelets (SWT, DWT, CWT), matching pursuit, and wavelet packet analysis. Filters: Finite Impulse Response (FIR), Infinite Impulse Response (IIR), Polyphase Multirate, Adaptive (Wiener, Kalman, etc.). Estimation and Statistical Signal Processing: Analysis and feature extraction of information from signals and noise based on stochastic properties.
Nonlinear Signal Processing – Maximum Likelihood and Robust Estimation: Parameter Estimation (restricted and search control), Linear Regression in nonlinear spaces with kernel transformations, Quantile-based estimation, U-statistic, Chi squared, Kolmogorov-Smirnov (KS), etc. Stable Distributions: Series approximations (stable pdf, cdf), Discrete distributions, Multivariate stable functions. Weighted Myriads: Smoothers, Heavy-tailed noise filters, and Linear combinations (order statistics). Chaos theory (attractors in embedded dimensions): State space reconstruction.
Unsupervised Machine Learning – Latent Variable Models: Principle Component Analysis (PCA), Independent Component Analysis (ICA), Factor Analysis (FA), Gaussian Mixtures, and K-means. Time Series Models: State-Space Models (SSM) and Hidden Markov Models (HMM). Nonlinear Models: PCA, FA, SSM, and HMM.
Supervised Machine Learning – Neural Networks (NN): Feed-Forward, Radial Basis Function (RBF), Kohonen Self-organizing, Learning Vector Quantization, and Neuro-fuzzy. Wavelet Linear Regression: Linear, Exponential, and Quadratic. Support Vector Machines (SVM): Linear, Soft Margin, Nonlinear, Multiclass, Structured, and Regression. Logistic Regression: Binary, Ordinal, and Nominal. Naïve Bayes: Pairwise Threshold and Singleton Threshold. Linear Discriminant Analysis (LDA): Class-Dependent Transformation and Class-Independent Transformation. K-Nearest Neighbor: Weighting and Self Organizing Maps (SOM). Decision Trees: Classification, Regression, Boost, Forest, K-Means. Bag & Boosting: AdaBoost, LP Boost, GentleBoost, and Bootstrap Aggregating.
We support a unique hookup solution that incorporates a vector transform. In addition to the traditional 12-lead hook-up, this will allow for a 5-wire or 7-wire hookup to produce 12 channels of ECG data. By using fewer wires, technicians and clinicians will benefit from a faster, simpler, and accurate diagnostic process.
What Makes Us Technologically Superior?
Our primary development focus is on medical devices with wireless radios. We believe the future of cardiology will rely heavily on getting important medical data from remote locations to physicians in a timely, reliable manner. We believe our wireless technology can accomplish this. Our flagship device, the battery powered, TM-12, is a 12-Lead ECG system with a Bluetooth wireless radio. The TM-12 is capable of streaming twelve channels of ECG up to 100ft. This product uses the world renowned Glasgow Resting 12-Lead ECG Analysis Program. The TM-12 can also email reports to physician’s phone.