PB1107 - AI classification model to discriminate causes of prolonged clotting time through the application of machine learning to clot waveform analysis parameters

Tuesday, June 27, 2023

18:30 – 19:30 ET

Room: Exhibition

Presenting Author(s)

Masato Matsuda, PhD

Assistant professor
Fukushima Medical University
Fukushima, Fukushima, Japan

Co-Author(s)

Daiki Shimomura, PhD

Vice-Head Medical technologist
Department of Laboratory Medicine, Tenri Hospital
Tenri, Nara, Japan
TM

Tomoko Matsumoto

Associate Professor
Department of Clinical Laboratory , Tenri Health Care University
Tenri, Nara, Japan
YT

Yuka Tabuchi

Employee
Sysmex Corporation
Kobe, Hyogo, Japan
Ak

Aya kohno

Employee
Sysmex Corporation
Kobe, Hyogo, Japan
TS

Takeshi Suzuki

Employee
Sysmex Corporation
Kobe, Hyogo, Japan
KN

Keisuke Nishi

Employee
Sysmex Corporation
Kobe, Hyogo, Japan
NA

Nobuo Arai

Employee
Sysmex Corporation
Kobe, Hyogo, Japan
HK

Hiroshi Kurono

Employee
Sysmex Corporation
Kobe, Hyogo, Japan
YH

Yoshiki Hoshiyama

Head Medical technologist
Medical Laboratory Division, Niigata University Medical and Dental Hospital
Niigata, Niigata, Japan
MK

Mikio Kamioka

Director
Department of Laboratory Medicine, Tenri Hospital
Tenri, Nara, Japan
ST

Shuji Terai

Professor
Division of Gastroenterology and Hepatology, Graduate School of Medical and Dental Sciences, Niigata University
Niigata, Niigata, Japan
MM

Masato Moriyama

Associate Professor
Department of Medical Oncology, Niigata University Graduate School of Medical and Dental Sciences.
Niigata, Niigata, Japan

Background: Activated partial thromboplastin time (APTT) and prothrombin time (PT) are routine screening tests for coagulation factor deficiency and/or lupus anticoagulant (LA). Although prolongation may indicate these pathologies, other factors, such as liver dysfunction and anticoagulant use, can cause prolongation. Further testing is usually required to discriminate between causes, which is costly and time-consuming. Clot waveform analysis (CWA) in APTT and PT tests could overcome these limitations. Unfortunately, even though a large number of CWA parameters have been available, it is yet to be determined how to effectively combine them to discriminate between the causes of prolonged clotting time.

Aims: To develop an artificial intelligence (AI) model to classify the cause of prolonged clotting time through the application of machine learning (ML) to CWA parameters.

Methods: This study used 719 clinical samples, all of which showed APTT prolongation, including 100 heparin-containing, 258 direct oral anticoagulants (DOACs)-containing, 104 warfarin-containing, 104 with liver dysfunction, 77 LA-positive, and 76 hemophilia A- or B-derived samples. APTT and PT were measured for these samples on an automated coagulation analyzer (CN-6000) using Revohem™ APTT and Revohem™ PT reagents (Sysmex, Japan). A total of 22 and 21 APTT- and PT-based CWA parameters, respectively, were obtained from clot waveform, first- and second-derivative curves. Using the data and the ML algorithm, a 6-category classification model was developed for APTT prolongation cause (heparin, DOACs, warfarin, liver dysfunction, LA, and hemophilia), followed by cross-validation to assess the algorithm.

Results: With ≥95% specificity, the sensitivity in heparin, DOACs, warfarin, liver dysfunction, LA, and hemophilia samples was 87%, 87.6%, 98.1%, 75%, 85.7%, and 100%, respectively.

Conclusion(s): This is the first study to develop an AI model of CWA parameters for coagulopathy screening. ML is a useful tool to effectively incorporate numerous CWA parameters into this model.