Machine Learning based Hematologic Disorder Detection

Abstract

Hematologic disorders are serious health problems caused by abnormalities in the production, function or structure of blood cells that can only be controlled with early diagnosis and appropriate intervention. These abnormalities, which manifest themselves in diseases such as anemia, leukemia, lymphoma, hemophilia and thrombocytopenia, directly affect vital functions such as the immune system, oxygen carrying capacity, clotting and tissue repair. Timely diagnosis of these diseases is critical to establishing effective treatment plans and preventing complications. However, crowded data tracking and interpretation difficulties in clinical processes can make physician decisions difficult. Therefore, AI-supported digital health solutions play an important role in treatment management. In this study, a decision support system was developed to facilitate the diagnosis of hematological diseases and provide individualized treatment recommendations. The system aims to perform disease diagnosis with high accuracy based on hemogram (complete blood count) data. While the Random Forest Algorithm is used in the diagnosis process, nutrition, lifestyle and general health recommendations are generated using decision trees. Class imbalance in the dataset was balanced with Synthetic Minority Oversampling Technique (SMOTE) to improve model performance. Flask and Node.js supported backend infrastructure is decomposed with React based interfaces for different user authorizations and data sets are stored in MongoDB database. In the decision support system developed for the detection of hematological disorders, a scalable artificial intelligence solution that can be integrated into clinical applications is presented, aiming to increase the accuracy of diagnosis, accelerate the treatment process and improve the quality of patient care.

Keywords

• Decision support system
• hematologic disorder
• hemogram
• machine learning
• random forest algorithm

Makine Öğrenmesi Tabanlı Hematolojik Bozukluk Tespiti

Abstract

Hematolojik bozukluklar, kan hücrelerinin üretimi, işlevi veya yapısında meydana gelen anormallikler sonucu ortaya çıkan, ancak erken teşhis ve uygun müdahale ile kontrol altına alınabilen ciddi sağlık problemleridir. Anemi, lösemi, lenfoma, hemofili ve trombositopeni gibi hastalıklar ile kendini gösteren bu anormaliler; bağışıklık sistemi, oksijen taşıma kapasitesi, pıhtılaşma ve doku onarımı gibi hayati fonksiyonları doğrudan etkilemektedir. Bu hastalıkların zamanında tanısı, etkili tedavi planlarının oluşturulması ve komplikasyonların önlenmesi açısından kritik önem taşımaktadır. Ancak klinik süreçlerde kalabalık veri takibi ve yorumlama zorlukları, hekim kararlarını zorlaştırabilmektedir. Bu nedenle yapay zeka destekli dijital sağlık çözümleri, tedavi yönetimlerinde önemli bir rol oynamaktadır. Bu çalışmada, hematolojik hastalıkların teşhisini kolaylaştırmak ve bireyselleştirilmiş tedavi önerileri sunmak amacıyla bir karar destek sistemi geliştirilmiştir. Sistem, hemogram (tam kan sayımı) verilerini temel alarak hastalık teşhisini yüksek doğrulukla gerçekleştirmeyi hedeflemektedir. Teşhis sürecinde Random Forest Algoritma’sı kullanılırken; beslenme, yaşam tarzı ve genel sağlık önerileri karar ağaçları yöntemiyle oluşturulmuştur. Veri setindeki sınıf dengesizliği, model performansını artırmak amacıyla Synthetic Minority Oversampling Technique (SMOTE) ile dengelenmiştir. Flask ve Node.js destekli backend altyapısı farklı kullanıcı yetkileri için React tabanlı arayüzler ile ayrıştırılarak, veri setleri MongoDB veritabanında depolanmıştır. Hematolojik bozuklukların tespiti için geliştirilen karar destek sisteminde tanı doğruluğunu artırmayı, tedavi sürecini hızlandırmayı ve hasta bakım kalitesini yükseltmeyi amaçlayan, klinik uygulamalara entegre edilebilir ve ölçeklenebilir bir yapay zeka çözümü sunulmaktadır.

Keywords

• Hematolojik bozukluk
• hemogram
• karar destek sistemi
• makine öğrenimi
• random forest algoritması

References

1. M. N. Garcia-Casal, O. Dary, M. E. Jefferds, and S. R. Pasricha, "Diagnosing anemia: Challenges selecting methods, addressing underlying causes, and implementing actions at the public health level," Ann. N. Y. Acad. Sci., vol. 1524, no. 1, pp. 37-50, 2023.
2. Y. El Alaoui, A. Elomri, M. Qaraqe, R. Padmanabhan, R. Y. Taha, H. El Omri, and O. Aboumarzouk, "A review of artificial intelligence applications in hematology management: current practices and future prospects," J. Med. Internet Res., vol. 24, no. 7, p. e36490, 2022.
3. A. Srisuwananukorn, M. E. Salama, and A. T. Pearson, "Deep learning applications in visual data for benign and malignant hematologic conditions: a systematic review and visual glossary," Haematologica, vol. 108, no. 8, p. 1993, 2023.
4. Y. Zhang, Y. Weng, and J. Lund, "Applications of explainable artificial intelligence in diagnosis and surgery," Diagnostics, vol. 12, no. 2, p. 237, 2022.
5. N. B. Noor, U. A. Oyshi, A. Das, and K. Iqbal, "A systematic approach to predict anemia from eye conjunctiva images," in Proc. 26th Int. Conf. Comput. Inf. Technol. (ICCIT), Cox's Bazar, Bangladesh, Dec. 13-15, 2023, pp. 1-5.
6. U. Ponnusamy, D. Darshan, and N. Sampathila, "Approaching explainable artificial intelligence methods in the diagnosis of iron deficiency anemia using blood parameters," in Proc. Int. Conf. Recent Adv. Inf. Technol. Sustain. Dev. (ICRAIS), Manipal, India, Nov. 6-7, 2023, pp. 201-206.
7. A. Tummala and K. Parvataneni, "Comparative study of AI-based anemia prediction using eye conjunctiva images and blood tests," in Proc. Int. Workshop Artif. Intell. Image Process. (IWAIIP), Yogyakarta, Indonesia, Dec. 1-2, 2023, pp. 87-91.
8. A. G. Mugdha, F. T. Pinki, and S. K. Talukdhar, "Hemoglobin estimation and anemia severity prediction using machine learning algorithms," in Proc. 5th Int. Conf. Sustain. Technol. Ind. 5.0 (STI), Dhaka, Bangladesh, Dec. 9-10, 2023, pp. 1-6.

9. M. M. Rahman, M. U. Mojumdar, H. A. Shifa, N. R. Chakraborty, N. P. Stenin, and M. A. Hasan, "Anemia disease prediction using machine learning techniques and performance analysis," in Proc. 11th Int. Conf. Comput. Sustain. Global Dev. (INDIACom), New Delhi, India, Feb. 28-Mar. 1, 2024, pp. 1276-1282.
10. E. Aboelnaga, "Anemia types classification," Kaggle, Dataset, 2022. [Online]. Available: https://www.kaggle.com/datasets/eboelnaga/anemia-types-classification
11. M. Levi, M. Simonetti, E. Marconi, O. Brignoli, M. Cancian, A. Masotti, and F. Lapi, "Gender differences in determinants of iron-deficiency anemia: a population-based study conducted in four European countries," Ann. Hematol., vol. 98, no. 7, pp. 1573-1582, 2019.
12. R. Krisnanda, "Vitamin C helps in the absorption of iron in iron deficiency anemia," J. Penelit. Perawat. Prof., vol. 2, no. 3, pp. 279-286, 2020.
13. J. Deng, L. Ramelli, P. Y. Li, A. Eshaghpour, G. E. M. Schuenemann, and M. A. Crowther, "Efficacy of vitamin C with iron supplementation in iron deficiency anemia patients: A systematic review and meta-analysis," Blood, vol. 142, p. 1091, 2023.
14. J. Deng, L. Ramelli, P. Y. Li, A. Eshaghpour, G. E. M. Schuenemann, and M. A. Crowther, "Efficacy of vitamin C with Fe supplementation in patients with iron deficiency anemia: A systematic review and meta-analysis," Blood Vessels Thromb. Hemost., vol. 1, no. 4, p. 100023, 2024.
15. Y. Qin, A. Melse-Boonstra, X. Pan, B. Yuan, Y. Dai, J. Zhao, and Z. Shi, "Anemia in relation to body mass index and waist circumference among Chinese women," Nutr. J., vol. 12, no. 10, pp. 1-3, 2013.
16. A. C. Cepeda-Lopez, S. J. Osendarp, A. Melse-Boonstra, I. Aeberli, F. Gonzalez-Salazar, E. Feskens, and M. B. Zimmermann, "Sharply higher rates of iron deficiency in obese Mexican women and children are predicted by obesity-related inflammation rather than by differences in dietary iron intake," Am. J. Clin. Nutr., vol. 93, no. 5, pp. 975-983, 2011.
17. A. Peña-Fernández, M. Higueras, E. Segura, M. D. Evans, and M. D. L. Á. Peña, "Inadequacies of dietary iron intake in normal- and overweight young university students from Leicester, England," Biol. Life Sci. Forum, vol. 38, no. 1, p. 6, 2025.
18. N. V. Chawla, K. W. Bowyer, L. O. Hall, and W. P. Kegelmeyer, "SMOTE: synthetic minority over-sampling technique," J. Artif. Intell. Res., vol. 16, pp. 321-357, 2002.
19. L. Breiman, "Random forests," Mach. Learn., vol. 45, pp. 5-32, 2001.
20. A. Liaw and M. Wiener, "Classification and regression by randomForest," R News, vol. 2, no. 3, pp. 18-22, 2002.
21. M. Fernández-Delgado, E. Cernadas, S. Barro, and D. Amorim, "Do we need hundreds of classifiers to solve real world classification problems?," J. Mach. Learn. Res., vol. 15, no. 1, pp. 3133-3181, 2014.
22. H. Chen, et al., "A random forest model based classification scheme for medical diagnosis," Comput. Math. Methods Med., 2017.