Microcontroller based in vitro hematocrit measurement system

Abstract

<span>The hematocrit (HCT) is the most important measurement in the blood profile. It has been used for early diagnose of the specific blood diseases such as anaemia, leukaemia and malaria. The microhematocrit is the conventional method of measurement of HCT manually which is time-consuming and uncertain due to human error. An automated system for measuring hematocrit will minimize the human-error and the time which will give the ability for medical staff to serve more patients. This paper aims to demonstrate an automated system for measuring the HCT based on microcontroller. The designed system based on Arduino Atmega 2560 microcontroller and combination array of lighting emitting diode and photodetectors. The transmission and the absorption characteristics of the red light (660nm) through the centrifuged blood sample in a capillary tube are calculated and used to determine the HCT. The outputs are analyzed to determine the haemoglobin (HB) and packed cell volume (PCV). The significant correlation (r=0.9856, p=3.106*10-4) between the PCV readings of the proposed system and the conventional method has been observed. The most important finding is the precise of PCV and HB readings for the proposed system compared with previous automated methods as well as the conventional method have been obtained.</span>

The application of instrumentation system on a contactless robotic triage assistant to detect early transmission on a COVID-19 suspect

“…It is also used as the core for measuring the vital signs of the user's since its specifications are suitable for real-time processing [19], [20]. Microcontroller boards such as Arduino have also been widely used for instrumentations and control applications in the medical field [21]- [30]. User interactions, triage processes, and report generations are controlled by desktop application, from which a triage report in portable document format (PDF) format will be produced and emailed to the medical operator.…”

Cotton-wool spots, red-lesions and hard-exudates distinction using CNN enhancement and transfer learning

“…Moving from human inspection to machine inspection will impose a new approach in the medical field to analyze the multilane classification and the disease conditions [8]. The strategy of implementing deep learning (DL) approach for the disease recognition and image distinction requires a primary training set which is annotated with those lesion symptoms.…”