TY - JOUR

T1 - ESTIMATING THE NUMBER OF MALARIA PARASITES ON BLOOD SMEARS MICROSCOPIC IMAGES USING PENALIZED SPLINE NONPARAMETRIC POISSON REGRESSION

AU - Chamidah, Nur

AU - Lestari, Budi

AU - Saifudin, Toha

AU - Rulaningtyas, Riries

AU - Wardhani, Puspa

AU - Budiantara, I. Nyoman

N1 - Publisher Copyright:
© 2024 the author(s).

PY - 2024

Y1 - 2024

N2 - So far, the detection and calculation of the malaria index has been done manually using thick and thin blood smears. Weaknesses of microscopic examination include the inability to detect low parasitaemia (low titre) so that it is not useful in non-endemic areas of malaria, the possibility of misinterpretation of very low or very high parasitaemia, the inability to detect mixed infections requires time and expertise in preparation for reading. Detection and calculation of parasites using digital imaging has begun to be studied in the world, but its application is still limited, especially in Indonesia. Several statistical models can be used to estimate the parasite index and detect parasite morphology microscopically. In this research, we propose an alternative method, called PSNPR method, to estimate the number of malaria parasites precisely by using a statistical modeling approach, namely, penalized spline nonparametric Poisson regression (PSNPR) model. We use image processing techniques for changing image to numeric, then we reduce dimension by using discrete wavelet transform, and principal component analysis. The results show that the proposed alternative method has high ability to detect and calculate the number of malaria parasites on microscopic image of blood smears. In the future, the results of this study can be used for prediction purpose that is to predict duration of time until the malaria parasites death after the patient is given treatment by a doctor who treats the patient.

AB - So far, the detection and calculation of the malaria index has been done manually using thick and thin blood smears. Weaknesses of microscopic examination include the inability to detect low parasitaemia (low titre) so that it is not useful in non-endemic areas of malaria, the possibility of misinterpretation of very low or very high parasitaemia, the inability to detect mixed infections requires time and expertise in preparation for reading. Detection and calculation of parasites using digital imaging has begun to be studied in the world, but its application is still limited, especially in Indonesia. Several statistical models can be used to estimate the parasite index and detect parasite morphology microscopically. In this research, we propose an alternative method, called PSNPR method, to estimate the number of malaria parasites precisely by using a statistical modeling approach, namely, penalized spline nonparametric Poisson regression (PSNPR) model. We use image processing techniques for changing image to numeric, then we reduce dimension by using discrete wavelet transform, and principal component analysis. The results show that the proposed alternative method has high ability to detect and calculate the number of malaria parasites on microscopic image of blood smears. In the future, the results of this study can be used for prediction purpose that is to predict duration of time until the malaria parasites death after the patient is given treatment by a doctor who treats the patient.

KW - WHO

KW - blood smears

KW - malaria parasite

KW - microscopic image

KW - penalized spline nonparametric Poisson regression

UR - http://www.scopus.com/inward/record.url?scp=85199045060&partnerID=8YFLogxK

U2 - 10.28919/cmbn/8578

DO - 10.28919/cmbn/8578

M3 - Article

AN - SCOPUS:85199045060

SN - 2052-2541

VL - 2024

JO - Communications in Mathematical Biology and Neuroscience

JF - Communications in Mathematical Biology and Neuroscience

M1 - 60

ER -