Medical Applications

Airborne Contagion Respiratory Analysis

Description: Caregiver-patient interactions present a significant challenge in how healthcare can be delivered effectively and safely under the threat of highly-infectious diseases in the absence of solutions that guarantee protection to both caregivers and patients. These risks have recently become much more pronounced with the recent findings that suggest that the SARS-CoV-2 virus can now be spread through airborne transmission. This means that all individuals within these environments are all directly exposed to exhaled particulates, even with social distancing and Personal Protection Equipment (PPE) use. The societal implication of failing to reduce the risk to employers, caregivers, and their patients will have a profound impact on our ability to combat the COVID-19 threat. To address this problem, we propose to create a multi-modal airborne contagion mapping system that tracks exhaled particulates through visual exhale behaviors and identifies behaviors that contribute to higher transmission rates. Through this new form of analysis, we will be able to objectively evaluate social distancing protocols, PPE use, and create more effective risk mitigation strategies. These solutions can then be adopted into public health safety solutions for nursing homes, daycares, schools, and everyday workplaces including offices and retail stores.
Download Links: [ YouTube ]
Non-contact Medium-based Respiratory Analysis through Reinforced Hybrid Model

Description: The ability to monitor natural respiratory behaviors is essential to pulmonology. Existing contact-based respiratory monitoring methods are known for their accuracy, but aregenerally considered uncomfortable, whereas non-contact methods are comfortable but less accurate. Thin Medium Thermal Imaging (TMTI), a respiration monitoring method that uses a thin medium and thermal imaging to sense breathing activity, was proposed as an alternative non-contact method that monitors respiration directly. However, the accuracy of this method suffered due to lack of information about airflow behaviors occurring between the user’s mouth and the medium. In this paper, we describe a reinforced hybrid model that utilizes a thermal camera with a spectral filter (3-5[m]) to provide breathing behavior information in addition to a low-cost thermal camera to capture images of the medium. The thermal camera with a spectral filter visualizes CO2, giving us the ability to observe the turbulent behavior of exhaled airflows. By further understanding the relationship between human breathing behaviors and the heat signatures on the medium using this hybrid model, we are able to classify breathing mode with at least 91% accuracy.
Download Links: [ PDF ]
Behavioral Analysis of Turbulent Exhale Flows
Authors: Shane Transue, Sayed Mohsin Reza, Ann C. Halbower, Min-Hyung Choi

Description: Dense exhale flow through carbon-dioxide spectral imaging introduces a pivotal trajectory within non-contact respiratory analysis that consolidates several pulmonary evaluations into a single coherent monitoring process. Due to technical limitations and the limited exploration of respiratory analysis through this non-contact technique, this method has not been fully utilized to extract high-level respiratory behaviors through turbulent exhale analysis. In this work, we present a structural foundation for respiratory analysis of turbulent exhale flows through the visualization of dense carbon-dioxide density distributions using precisely refined thermal imaging device to target high-resolution respiratory modeling. We achieve spatial and temporal high-resolution flow reconstructions through the cooperative development of a thermal camera dedicated to respiratory analysis to drastically improve the precision of current exhale imaging methods. We then model turbulent exhale behaviors using a heuristic volumetric flow reconstruction process to generate sparse flow exhale models. Together these contributions allow us to target the acquisition of numerous respiratory behaviors including, breathing rate, exhale strength and capacity, towards insights into lung functionality and tidal volume estimation.
Download Links: [ PDF ] [ YouTube ] [ YouTube2 ]
Thermal-Depth Fusion for Occluded Body Skeletal Posture Estimation
Members: Shane Transue, Phuc Nguyen, Tam Vu, Min-Hyung Choi

Description: Reliable occluded skeletal posture estimation is a fundamentally challenging problem for ision-based monitoring techniques. This is due to several imaging related challenges introduced by existing depth-based pose estimation techniques that fail to provide accurate joint position estimates when the line of sight between the imaging device and the patient is obscured by an occluding material. In this work, we present a new method of estimating skeletal posture in occluded applications using both depth and thermal imaging through volumetric modeling and introduce a new occluded ground-truth tracking method inspired by modern motion capture solutions. Using this integrated volumetric model, we utilize Convolutional Neural Networks to characterize and identify volumetric thermal distributions that match trained skeletal posture estimates which includes disconnected skeletal definitions and allows correct posture estimation in highly ambiguous cases. We demonstrate this approach by correctly identifying common sleep postures that present challenging cases for current skeletal joint estimations, obtaining an average classification accuracy of ~94.45%.
Download Links: [ PDF ] [ IEEE Xplore ] [ YouTube ]
WiKiSpiro: Non-contact Respiration volume Monitoring during Sleep
Members: Phuc Nguyen, Shane Transue, Min-Hyung Choi, Ann C. Halbower, Tam Vu

Description: Respiration volume has been widely used as an important indication for diagnosis and treatment of pulmonary diseases and other health care related issues such as critically ill patients neonatal ventilation, post-operative monitoring and various others. Most of existing technologies for respiration volume monitoring require physical contact with the human body. While wireless-based approaches have also been discussed in the literature, there are still limitations in terms of estimation accuracy and time efficiency preventing these approaches from being realized in practice. In this paper, we present an automated, wireless-based, vision-supervised system, called WiKiSpiro, for monitoring an individual’s respiration volume. In particular, we present a system design encompassing a wireless device, motion tracking and control system, and a set of methods to accurately derive breathing volume from the reflected signal and to address challenges caused by body movement and posture changes. We present our preliminary results of WikiSpiro, and identify possible challenges for future research and development.
Download Links: [ PDF ] [ ACM Digital Library ] [ YouTube ]
Real-time Tidal Volume Estimation using Iso-surface Reconstruction
Members: Shane Transue, Phuc Nguyen, Tam Vu, Min-Hyung Choi

Description: Breathing volume measurement has long been an important physiological indication widely used for the diagnosis and treatment of pulmonary diseases. However, most of existing breathing volume monitoring techniques require either physical contact with the patient or are prohibitively expensive. In this paper we present an automated and inexpensive non-contact, vision-based method for monitoring an individual's tidal volume, which is extracted from a three-dimensional (3D) chest surface reconstruction from a single depth camera. In particular, formulating the respiration monitoring process as a 3D space-time volumetric representation, we introduce a real-time surface reconstruction algorithm to generate omni-direction deformation states of a patient's chest while breathing, which reflects the change in tidal volume over time. These deformation states are then used to estimate breathing volume through a per-patient correlation metric acquired through a Bayesian-network learning process. Through prototyping and implementation, our results indicate that we have achieved 92.2% to 94.19% accuracy in the tidal volume estimations through the experimentation based on the proposed vision-based method.
Download Links: [ PDF ] [ IEEE Xplore ] [ YouTube ]
Fast Volume Preservation for a Mass-Spring System

Authors: Min Hong, Sunwha Jung, Min-Hyung Choi, Samuel Welch

Description: This article presents a new method to model fast volume preservation of a mass-spring system to achieve a realistic and efficient deformable object animation, without using internal volumetric meshing. With this method the simulated behavior is comparable to a finite-element-method-based model at a fraction of the computational cost.
Download Links: [ PDF ]
Volume-Preserved Human Organs for Surgical Simulation

Authors: Sunwha Jung, Min Hong, Min-Hyung Choi

Description: One of the challenging problems in surgical simulation is to reduce the computational cost to achieve interactive refresh rates for both haptic and visualization devices, while maintaining reasonable behavioural realism. Since human organs are predominantly based on water, they preserve overall volume during deformation. Therefore, representing the volume-preserved behaviour in dynamic system is essential to deliver realistic organ reaction in surgical simulation. Many existing methods for modeling and simulation of human organs often neglect the volume preservation due to its computational complexities. Otherwise, some previous volume preservation methods alter the material properties, resulting in hardened and unnatural dynamic behaviour. This paper presents a novel method to model human organs with volume preservation. It keeps the material properties intact and requires virtually no additional computation cost to address both computational efficiency and visual realism. Our method incorporates an implicit volume constraint on a simple mass-spring system. Experiments show that the object level volume is well maintained even under high pressure. Proposed method makes a realistic human organ simulation possible at an interactive rate with almost no additional computational cost.
Download Links: [ PDF ]
Integrated Term Weighting, Visualization, and User Interface Development for Bioinformation Retrieval

Authors: Min Hong, Anis Karimpour-fard, Steve Russell, Lawrence Hunter

Description: This project implements an integrated biological information website that classifies technical documents, learns about users' interests, and offers intuitive interactive visualization to navigate vast information spaces. The effective use of modern software engineering principles, system environments, and development approaches is demonstrated. Straightforward yet powerful document characterization strategies are illustrated, helpful visualization for effective knowledge transfer is shown, and current user interface methodologies are applied. A specific success of note is the collaboration of disparately skilled specialists to deliver a flexible integrated prototype in a rapid manner that meets user acceptance and performance goals. The domain chosen for the demonstration is breast cancer, using a corpus of abstracts from publications obtained online from Medline. The terms in the abstracts are extracted by word stemming and a stop list, and are encoded in vectors. A TF-IDF technique is implemented to calculate similarity scores between a set of documents and a query. Polysemy and synonyms are explicitly addressed. Groups of related and useful documents are identified using interactive visual displays such as a spiral graph that represents of the overall similarity of documents. K-means clustering of the similarities among a document set is used to display a 3-D relationship map. User identities are established and updated by observing the patterns of terms used in their queries, and from login site locations. Explicit considerations of changing user category profiles, site stakeholders, information modeling, and networked technologies are pointed out.
Download Links: [ PDF ]