A Comparative Study of Machine Learning Models for Early Detection of Acute Lymphoblastic Leukemia (ALL)
Khalid Mahboob, Umme Laila, Mahvash Arsalan Lodhi · 2024
It is critical to identify acute lymphoid leukemia (ALL) at the onset to give the patients a chance at recovery and lead healthier lives with proper management. Although effective, traditional methods of diagnostics are long and situational when it comes to the involvement of people. Development in system cognition has led to the availability of automated, accurate, and diagnostic equipment that can analyze clinical images and data and enhance early diagnosis. Using the mode recognition system as the evaluation tool for this study, it chooses among the pre-trained models like Inception V3, VGG-16, and VGG-19, along with machine learning (ML) models for diagnosing ALL at its early stage. It was found that the Inception V3 outperformed the VGG models in all the aspects recognized with the added feature of high accuracy and low computation time. The two models built, NN and SGD, reached about 98% accuracy. Inception V3 has a well-thought-out architecture, so the authors believe it suits medical applications.