Mizpah Selvam S. Johnson

Bio: Mizpah Selvam S. Johnson is an academic researcher from University of Ontario Institute of Technology. The author has an hindex of 1, co-authored 1 publications receiving 1 citations.

A Review of Photoplethysmography-based Physiological Measurement and Estimation, Part 1: Single Input Methods*

Abstract: Photoplethysmography can be used to estimate many physiological parameters based on features extracted from the measured waveform. Following the single parameter estimations that have been reviewed in part 1 of this paper, we here review methods where the waveform is used in conjunction with other measured physiological signals. Being a low-cost, non-invasive, and user friendly technique, many PPG-based physiological data extraction methods are being researched. The parameters reviewed that can be estimated using the PPG waveform plus additional inputs include cardiac output, blood pressure, venous function assessment, blood oxygen saturation, and fetal heart rate and fetal oxygen saturation. The different processing techniques, algorithms and methods are reviewed in addition to providing a comparison of results with the reference standards to validate the different methods. Future research considerations for each parameter estimation are also discussed. This paper could be helpful for future research on PPG based wearable devices for physiological multi-parameter estimations.

Emerging point-of-care technologies for anemia detection.

Abstract: Anemia, characterized by low blood hemoglobin level, affects about 25% of the world's population with the heaviest burden borne by women and children. Anemia leads to impaired cognitive development in children, as well as high morbidity and early mortality among sufferers. Anemia can be caused by nutritional deficiencies, oncologic treatments and diseases, and infections such as malaria, as well as inherited hemoglobin or red cell disorders. Effective treatments are available for anemia upon early detection and the treatment method is highly dependent on the cause of anemia. There is a need for point-of-care (POC) screening, early diagnosis, and monitoring of anemia, which is currently not widely accessible due to technical challenges and cost, especially in low- and middle-income countries where anemia is most prevalent. This review first introduces the evolution of anemia detection methods followed by their implementation in current commercially available POC anemia diagnostic devices. Then, emerging POC anemia detection technologies leveraging new methods are reviewed. Finally, we highlight the future trends of integrating anemia detection with the diagnosis of relevant underlying disorders to accurately identify specific root causes and to facilitate personalized treatment and care.

Hypertension Detection Based on Photoplethysmography Signal Morphology and Machine Learning Techniques

Abstract: In our modern digitalized world, hypertension detection represents a key feature that enables self-monitoring of cardiovascular parameters, using a wide range of smart devices. Heart rate and blood oxygen saturation rate are some of the most important ones, easily computed by wearable products that are provided by the photoplethysmography (PPG) technique. Therefore, this low-cost technology has opened a new horizon for health monitoring in the last decade. Another important parameter is blood pressure, a major predictor for cardiovascular characterization and health related events. Analyzing only PPG signal morphology and combining the medical observation with machine learning (ML) techniques, this paper develops a hypertension diagnosis tool, named the ANC Test™. During the development process, distinguishable characteristics have been observed among certain waveforms and certain types of patients that leads to an increased confidence level of the algorithm. The test was enchanted by machine learning models to improve blood pressure class detection between systolic normotensive and hypertensive patients. A total of 359 individual recordings were manually selected to build reference signals using open-source available databases. During the development and testing phases, different ML models accuracy of detecting systolic hypertension scored in many cases around 70% with a maximum value of 72.9%. This was resulted from original waveform classification into four main classes with an easy-to-understand nomenclature. An important limitation during the recording processing phase was given by a different PPG acquisition standard among the consulted free available databases.

Assessment of left ventricular systolic function using a single-channel ECG monitor with photoplethysmography based on machine learning models

Abstract:

ЦЕЛЬ ИССЛЕДОВАНИЯ

Разработка моделей машинного обучения для определения снижения систолической функции миокарда левого желудочка (ЛЖ) по данным электрокардиограммы (ЭКГ) и фотоплетизмограммы (ФПГ), зарегистрированным с помощью одноканального ЭКГ-монитора с функцией фотоплетизмографии.

МАТЕРИАЛ И МЕТОДЫ

В исследование были проспективно включены 400 пациентов. Каждому участнику исследования была выполнена эхокардиография, при которой определяли фракцию выброса (ФВ) ЛЖ и VTI выносящего тракта ЛЖ. Затем проводили регистрацию ЭКГ и ФПГ одноканальным ЭКГ-монитором с функцией фотоплетизмографии, который имеет вид чехла для смартфона. Затем все зарегистрированные записи передавали на единый сервер, где проводили расчет параметров ЭКГ и ФПГ. На основе полученных параметров были построены модели для оценки прогнозирования снижения систолической функции ЛЖ с применением регрессии Лассо и алгоритма «случайный лес».

РЕЗУЛЬТАТЫ

Были получены модели для ФВ менее 55, 40, 30% и VTI менее 16 и 13 см соответственно. Для каждой модели рассчитывали площадь под ROC-кривой (AUC), чувствительность, специфичность. Для моделей с применением регрессии Лассо результаты были следующими: для ФВ <55% AUC составила 0,857 (чувствительность 0,818, специфичность 0,860); для ФВ <40% — 0,971; для ФВ <30% — 0,982; для VTI <13 — 0,754, для VTI <16 — 0,746. Для моделей, построенных на основе алгоритма «случайный лес», результаты были также достаточно высокими: для ФВ <55% AUC составила 0,913; для ФВ <40% — 0,955; для ФВ <30% — 0,962; для VTI <13 — 0,776, для VTI <16 — 0,782.

ВЫВОД

Модели на основе машинного обучения, построенные с использованием параметров ЭКГ и ФПГ, показали достаточно высокую точность в оценке снижения систолической функции ЛЖ.