An algorithm to detect overlapping red blood cells for sickle cell disease diagnosis

Vicent, Mabirizi; Kawuma, Simon; Yonasi, Safari

doi:10.1049/ipr2.12439

Cited by 7 publications

(4 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The concept of transfer learning has been widely applied in the detection of sickle cells for example a study by (de Haan et al, 2020) utilised deep learning techniques in their framework for detecting sickle cell using blood smears taken with smartphone microscope. The framework achieved 98% accuracy, also (Vicent et al, 2022) reported 98.18% accuracy automation detection of overlapping red blood cells for sickle cells diagnosis. (Arishi, Alhadrami, & Zourob, 2021) conducted a review study on current and emerging techniques for sickle cell disease detection and highlighted potential methods for early diagnosis of sickle cell disease.…”

Section: Table 1: Model Performance On Training Dataset Table 2: Mode...mentioning

confidence: 72%

“…(Das, Meher, Panda, & Abraham, 2019) provides a detailed methodological review about deep learning technique and tools used to detect sickle cell disease. (Vicent et al, 2022) developed an algorithm to detect presence of sickle cells in overlapping red blood cells. In their method, canny edge and double threshold machine learning techniques were used to separate overlapping cells of digital blood smears.…”

Section: Literature Reviewmentioning

confidence: 99%

“…It is a comprehensive and tiresome task. On the other hand, application of Artificial Intelligence (AI) specifically deep learning techniques are amongst the recent alternative technologies to provide faster, cheap and more accurate diagnosis of sickle cell anemia (Kumar, Koul, Singla, Ijaz, & Computing, 2022) as compared to the clinical haemoglobin electrophoresis machine which is too expensive to be enrolled in every hospital (Vicent, Simon, & Yonasi, 2022). As a result, a number of models and algorithms for sickle cell disease detection have been developed.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Comparison of Deep Learning Techniques in Detection of Sickle Cell Disease

Simon,

Vicent,

Addah

et al. 2023

AIA

View full text Add to dashboard Cite

Recently transfer learning technique has proved to be powerful in enhancing development of deep learning methods for sickle cell disease (SCD) detection as a complement to the clinical method where a haemoglobin electrophoresis machine is used. This is evidenced by a number of models and algorithms with ≥ 90% prediction accuracy. From literature, most of the proposed methods are trained and tested on pre-trained deep learning models like VGG16, VGG19, ResNet, Inception_V3 and ReNet. However, training and testing of these methods is limited on one model and separate dataset which may lead to biased results due to implementation in variation of these models which affects results produced. To this end, there exist a need to evaluate the SCD models using the same dataset. Thus, in this research study, we carried out a comparative investigation and evaluated predominate pretrained models used to detect Sickle Cell Disease using the same dataset to ascertain which one has the best accuracy. We used secondary dataset obtained from an online dataset. In our study, we have discovered that , Inception V3 yielded the highest accuracy of 97.3% followed by VGG19 at 97.0%, VGG16 at 91%, ResNet50 at 82% and ReNet at 67%, and the CNN-scratch model achieved 81% accuracy. Results from our study will aid researchers and industry practitioners to make decision on the best deep learning model to use while detecting SCD.

show abstract

Section: Table 1: Model Performance On Training Dataset Table 2: Mode...mentioning

confidence: 72%

Section: Literature Reviewmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Comparison of Deep Learning Techniques in Detection of Sickle Cell Disease

Simon,

Vicent,

Addah

et al. 2023

AIA

View full text Add to dashboard Cite

show abstract

“…Azimuthal Walsh filters have also been used to modify the far-field diffraction pattern in a customized way [12,13]. Deep Convolutional Neural Networks and image processing techniques have also been applied to sickle cell detection [14,15]. Also, it is known that often any irregularity in the geometry of a cell implies the advent of some disease in it.…”

Section: Introductionmentioning

confidence: 99%

Application of geometrical irregularities identification using the cubic phase masks

Chatterjee

2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

The geometrical analysis of the healthy and the deformed biological cells is a very helpful aide to the generally more expensive and complicated pathological methods for disease detection. The methods demonstrated in this paper have successfully identified the healthy and the distorted cells using the cubic phase masks as the Fourier domain filters and a 4-f setup architecture of the Fourier Optics principles and the orientation wise weighted tuning of the edges of the cell. For two orthogonal arrangements of the mask, two images are generated. For the healthy and regular shaped cells, the maximum intensity values of these two images are the same whereas they are different for a distorted cell – for example, an ellipse or a sickle cell. Moreover, the maximum intensity values of the two halves of an image frame for a particular arrangement of the mask will be different for a tear-drop cell. These maximum brightness values are automatically read and the appropriate marking of these recognized cells has been done in an automated process.

show abstract