Artery ultrasound image recognition model training method and device, equipment and medium

The invention relates to an arterial ultrasound image recognition model training method and device, equipment and a storage medium. The method comprises the steps of preprocessing an obtained arterial ultrasound image to obtain a training image sample; classifying the training image samples by using...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: WANG WEI, CHEN HUAISHENG, HONG CHENGYING
Format: Patent
Sprache:chi ; eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The invention relates to an arterial ultrasound image recognition model training method and device, equipment and a storage medium. The method comprises the steps of preprocessing an obtained arterial ultrasound image to obtain a training image sample; classifying the training image samples by using a preset plaque gray scale double-peak gamma distribution model to obtain low-echo image samples, mixed-echo image samples and/or high-echo image samples; generating a low-echo image sample set, a mixed-echo image sample set and/or a high-echo image sample set by using a preset boundary smooth adversarial generative network based on the classified image samples; and respectively training a convolutional neural network by using the image sample sets of various types to obtain an arterial ultrasound image recognition model. According to the invention, the obtained arterial ultrasound image recognition model is stable, reliable and high in accuracy. In conclusion, the method can be widely applied to artery ultrasonic