Leukemia is a type of cancer that affects the blood and bone marrow, and it is characterized by the uncontrolled growth of abnormal blood cells, including leukocytes [1], [2]. The abnormal leukocytes produced in leukemia are immature and do not function properly, which can lead to a weakened immune system and an increased risk of infection [1]. Leukemia can affect the production and function of leukocytes by producing abnormal and immature leukocytes, interfering with the normal production of leukocytes in the bone marrow, and causing myelosuppression due to chemotherapy treatment [3]-[5]. Studies have shown that patients with leukemia may experience changes in the function and phenotype of myeloid cells, including monocytes and neutrophilic granulocytes, which are critical for innate immunity [4]. Furthermore, the bone marrow function of patients with leukemia may be impacted by chemotherapy treatment, which can cause myelosuppression, a reduction in the production of red and white blood cells and platelets, leading to an increased risk of infections [5].
The paper titled “A YOLO and convolutional neural network for the detection and classification of leukocytes in leukemia” proposes a system for computer-aided detection of leukocytes in leukemia using deep learning techniques [6]. The system uses a modified version of the You Only Look Once (YOLO v2) algorithm and a convolutional neural network (CNN) to detect and classify three types of white blood cells (WBCs) that are fundamental to leukemia diagnosis. The proposed system is trained and evaluated on a dataset created specifically for the addressed problem without any traditional segmentation or preprocessing on microscopic images. The study shows that dividing the addressed problem into sub-problems achieves better performance and accuracy. The results demonstrate that the CAD3 achieved an average precision (AP) of up to 96% in the detection of leukocytes and an accuracy of 94.3% in leukocyte classification. Additionally, the CAD3 provides a report containing complete information about WBCs. Finally, the CAD3 was tested on other datasets such as the acute lymphoblastic leukemia image database (ALL-IBD1) and the blood cell count dataset (BCCD), and it proved its efficiency [6].
The proposed system for the detection and classification of leukocytes in leukemia using a YOLO and convolutional neural network (CNN) has potential applications beyond leukemia diagnosing. Here are some potential applications:
- Automated Blood Cell Analysis:
The system can be utilized for automated analysis of blood cell samples, not limited to leukemia cases. It can assist in the detection and classification of various types of white blood cells, red blood cells, and platelets. This can be valuable in diagnosing other blood disorders and monitoring overall blood health.
- Medical Research and Drug Development:
The system can be used in medical research and drug development studies related to blood disorders. It can aid researchers in analyzing large datasets of blood cell images, identifying patterns, and studying the effects of different treatments or drugs on specific cell types.
- Telemedicine and Remote Healthcare:
With advancements in telemedicine and remote healthcare, the proposed system can be integrated into digital platforms or mobile applications. This would enable healthcare professionals to remotely analyze blood cell samples and provide accurate diagnoses or recommendations to patients in remote or underserved areas.
- Quality Control in Blood Banks:
Blood banks and transfusion centers can benefit from the system’s ability to detect and classify blood cells. It can be used for quality control purposes, ensuring that donated blood is free from abnormalities or infections before it is used for transfusions.
- Education and Training:
The system can be used as a teaching tool in medical education and training programs. It can assist students and healthcare professionals in learning about different types of blood cells, their characteristics, and associated diseases. The system can provide real-time feedback and help improve diagnostic skills.
- Early Detection of Blood Disorders:
By accurately detecting and classifying blood cells, the system can contribute to the early detection of various blood disorders, not limited to leukemia. Early detection can lead to timely interventions and improved patient outcomes.
It is important to note that while these potential applications are based on the capabilities of the proposed system, further research and validation would be necessary to ensure its effectiveness and reliability in different contexts.
The proposed system for the detection and classification of leukocytes in leukemia using a YOLO and convolutional neural network (CNN) has some limitations in terms of scalability and implementation in clinical settings. Here are some of the limitations:
- Limited Dataset:
The proposed system was trained and evaluated on a dataset created specifically for the addressed problem without any traditional segmentation or preprocessing on microscopic images. This means that the system may not perform as well on datasets with different characteristics or from different sources. Therefore, the system would need to be retrained on larger and more diverse datasets to ensure its scalability and generalizability.
- Hardware and Computational Requirements:
The proposed system requires significant computational resources and hardware to train and run. This can be a limitation in clinical settings where resources may be limited. Therefore, the system would need to be optimized to run on less powerful hardware or cloud-based platforms to make it more accessible.
- Integration with Existing Clinical Systems:
The proposed system would need to be integrated with existing clinical systems to be used in clinical settings. This can be a complex process that requires careful consideration of data privacy, security, and regulatory requirements. Therefore, the system would need to be designed with these factors in mind to ensure its successful integration.
- Human Expertise:
The proposed system is designed to assist healthcare professionals in diagnosing leukemia by detecting and classifying leukocytes. However, the system cannot replace the expertise of human pathologists or hematologists. Therefore, the system would need to be used in conjunction with human expertise to ensure accurate diagnoses.
- Cost:
The development and implementation of the proposed system can be costly. This can be a limitation in resource-limited settings or healthcare systems. Therefore, the cost-effectiveness of the system would need to be evaluated to ensure its feasibility and sustainability.
It is important to note that while these limitations exist, they can be addressed through further research and development. The proposed system has shown promising results in detecting and classifying leukocytes in leukemia, and with further refinement, it has the potential to be a valuable tool in clinical settings.
By: D. Ilham
Editor: S. D. Cahyo
References:
Pemodelan perkembangan jumlah sel leukosit penderita leukimia anak di surabaya dengan pendekatan regresi semiparametrik berdasarkan estimator kernel | Oktiriani | Matematika dan Ilmu Pengetahuan Alam
Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia | Bodey | Annals of internal medicine
Pyridoxal phosphate in plasma and leukocytes in patients with leukemia and other diseases | Wachstein | Proceedings of the Society for Experimental Biology and Medicine
Robust discrimination of leukocytes protuberant types for early diagnosis of leukemia | Naz | Journal of Mechanics in Medicine and Biology
Characteristics of patients with acute lymphoblastic leukemia (ALL) at al islam bandung hospital in 2017 | Silva | Prosiding Pendidikan Dokter
Combined ibrutinib and venetoclax changes myeloid phenotype and improves immune function in CLL patients | Svanberg | Blood
A YOLO and convolutional neural network for the detection and classification of leukocytes in leukemia | Abas | Indonesian Journal of Electrical Engineering and Computer Science (IJEECS)