Overview / Usage

Acute lymphoblastic leukemia (ALL), also known as Acute Lymphocytic Leukemia, is a cancer that affects the lymphoid blood cell lineage. It is the most common leukemia in children, and it accounts for 10-20% of acute leukemias in adults. The prognosis for both adult and especially childhood ALL has improved substantially since the 1970s. The 5- year survival is approximately 95% in children. In adults, the 5-year survival varies between 25% and 75%, with more favorable results in younger than in older patients.

The Acute Lymphoblastic Leukemia Detection System For Oculus Rift combines Oculus Rift's Virtual Reality technologies with Intel's oneAPI, OpenVINO & Neural Compute Stick to provide real-time classification of Acute Lymphoblastic Leukemia Lymphoblasts in peripheral blood samples within a Virtual Reality environment.

The classifier used in this project is the HIAS Acute Lymphoblastic Leukemia Classifier which is trained on the Acute Lymphoblastic Leukemia Image Database for Image Processing dataset. The classifier achieves 97% accuracy at detecting Acute Lymphoblastic Leukemia Lymphoblasts in images of peripheral blood samples.

Methodology / Approach

ALL_IDB1

In this project, ALL-IDB1 is used, one of the datsets from the Acute Lymphoblastic Leukemia Image Database for Image Processing dataset. We use data augmentation to increase the amount of training data we have.

Classifier

The Acute Lymphoblastic Leukemia (ALL) oneAPI Classifier is an open-source classifier programmed using the Intel® Distribution for Python* and trained using Intel® Optimization for TensorFlow*. The model is deployed on a Raspberry 4 using Intel® Distribution of OpenVINO™ Toolkit and inference is carried out using the Intel® Neural Compute Stick 2 (Intel® NCS2).

The classifier achieves 97 (97.27723)% accuracy. Below are the metrics after training on 150 epochs.

• Loss 0.060633027169964104
• Accuracy 0.9727723
• Precision 0.9727723
• Recall 0.9727723
• AUC 0.9978372
• True Positives: 173
• False Positives: 1
• True Negatives: 220
• False Negatives: 10
• Specificity: 0.995475113122172
• Misclassification: 11

Oculus Rift:

The project provides a virtual room created in Blender inside an Oculus Rift virtual environment.

Technologies Used

• Training: NVIDIA GTX 1050 ti with Ubuntu 18.04
• Model server: Raspberry Pi 4
• VR: Oculus Rift & Oculus Touch

• Python 3
• tensorflow
• oneAPI MKL
• NCS2
• OpenVino
• Unity 3D

Repository

https://github.com/aiial/hias-all-oneapi-classifier

Other links

• HIAS Acute Lymphoblastic Leukemia oneAPI Classifier (https://github.com/aiial/hias-all-oneapi-classifier)

• Official Documentation
• Official Project Page