2020-416 Automated Screening of Sickle Cells Using a Smartphone-Based Microscope and Deep Learning

SUMMARY

UCLA researchers in the Department of Electrical & Computer Engineering have developed an automated method for detecting sickle cell disease (SCD) using a smartphone-based microscope and deep learning.

BACKGROUND

Sickle cell disease (SCD) is a preventable disease that affects millions of people world-wide. Current screening methods involve the use of costly laboratory-based methods that require expertise to operate and analyze. In resource-limited regions, suspected SCD patients are individually screened by very limited, specially trained personnel who manually analyze blood smear samples. For better diagnosis, there is a need for an automated, cost-effective point-of-care platform for the inspection of blood smear samples that does not require specially trained personnel and can be employed in a resource-limited setting.

INNOVATION

UCLA researchers in the Department of Electrical & Computer Engineering have developed a smartphone-based microscopy method for analyzing blood samples using deep learning to detect sickle cell disease (SCD). The smartphone-based method requires minimal sample prep and is cost-effective and portable for rapid (<7 sec) sickle cell screening. The method has been successfully tested to achieve ~98% accuracy of detection, with an area-under-the-curve (AUC) of 0.998 using blood smears from 96 unique patients (including 32 SCD patients). With its high accuracy, this mobile and cost-effective method has the potential to be used as a screening tool for SCD and other blood cell disorders in resource-limited settings.

POTENTIAL APPLICATIONS

• Sickle cell diagnosis
• Sickle beta thalassemia diagnosis
• Hemoglobin diseases

ADVANTAGES

• Cost-effective
• Mobile
• High Accuracy

RELATED MATERIALS

• de Haan, K., Ceylan Koydemir, H., Rivenson, Y. et al. Automated screening of sickle cells using a smartphone-based microscope and deep learning. npj Digit. Med. 3, 76 (2020). https://doi.org/10.1038/s41746-020-0282-y

STATUS OF DEVELOPMENT

Successful development of method for detecting sickle cell disease using smartphone microscope and testing using blood smears from 96 unique patients (including 32 SCD patients) with 6.85 second analysis time per patient, ~98% accuracy of detection and with a 0.998 area-under-the-curve (AUC).