Machine learning algorithms to predict the 1 year unfavourable prognosis for advanced schistosomiasis

Jiang, Honglin; Deng, Weicheng; Zhou, Jie; Ren, Guang-hui; Xin-Ting, Cai; Sheng-ming, LI; Hu, Beibei; Li, Chunlin; Shi, Ying; Zhang, Na; Zheng, Yingyan; Chen, Yue; Jiang, Qingwu; Zhou, Yibiao

doi:10.1016/j.ijpara.2021.03.004

Cited by 16 publications

(14 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, although ultrasound is generally accepted as playing a significant role in the treatment of unstable patients, its use with stable patients is still controversial because it cannot rule out abdominal organ lesions. For patients with blunt trauma, the inability to demonstrate a lack of bleeding or delayed bleeding in the abdominal cavity has proven to be a primary limitation of US, often requiring patients to undergo contrast‐enhanced computed tomography 21,29,59 . However, as demonstrated in this study, the inclusion of AI could lead to earlier intervention, more precise treatment, and improved outcomes for trauma patients, particularly in emergency situations.…”

Section: Discussionmentioning

confidence: 81%

“…Specific areas in the abdomen are often examined for the presence of abdominal free fluid, a strong indication of severe intra‐abdominal injuries that may require an emergency laparotomy 16 . Several studies have shown that FAST can guide clinical decision‐making and determine the need for angiography or surgery, particularly for children, pregnant women, and patients exhibiting hemodynamic instabilities 3,10,17,19–25 . FAST can be performed at the bedside in 3‐4 min, repeatedly if need be, and avoids the risks associated with transporting patients and radiation 3,14,26 .…”

Section: Introductionmentioning

confidence: 99%

“…The use of artificial intelligence (AI) for medical applications has been expanding across multiple fields in recent years and deep learning (DL), specifically, has become one of the most popular computer‐aided diagnosis (CAD) techniques for ultrasonography (US) images 32–38 . DL has previously been applied to the CAD of ascites 21 . It is a data‐driven methodology that can be used to extract and learn nonlinear features from data, without requiring domain expertise 39 .…”

Section: Introductionmentioning

confidence: 99%

“…16 Several studies have shown that FAST can guide clinical decision-making and determine the need for angiography or surgery,particularly for children,pregnant women,and patients exhibiting hemodynamic instabilities. 3,10,17,[19][20][21][22][23][24][25] FAST can be performed at the bedside in 3-4 min, repeatedly if need be, and avoids the risks associated with transporting patients and radiation. 3,14,26 FAST identifies the presence of free fluid in the abdominal cavity, which is typically thought to be secondary to serious abdominal trauma.…”

mentioning

confidence: 99%

“…[32][33][34][35][36][37][38] DL has previously been applied to the CAD of ascites. 21 It is a data-driven methodology that can be used to extract and learn nonlinear features from data, without requiring domain expertise. 39 This is particularly beneficial for first-response and emergency situations, where access to specialists is limited.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Deep learning for emergency ascites diagnosis using ultrasonography images

Lin

Cao

et al. 2022

J Applied Clin Med Phys

View full text Add to dashboard Cite

Purpose The detection of abdominal free fluid or hemoperitoneum can provide critical information for clinical diagnosis and treatment, particularly in emergencies. This study investigates the use of deep learning (DL) for identifying peritoneal free fluid in ultrasonography (US) images of the abdominal cavity, which can help inexperienced physicians or non‐professional people in diagnosis. It focuses specifically on first‐response scenarios involving focused assessment with sonography for trauma (FAST) technique. Methods A total of 2985 US images were collected from ascites patients treated from 1 January 2016 to 31 December 2017 at the Shenzhen Second People's Hospital. The data were categorized as Ascites‐1, Ascites‐2, or Ascites‐3, based on the surrounding anatomy. A uniform standard for regions of interest (ROIs) and the lack of obstruction from acoustic shadow was used to classify positive samples. These images were then divided into training (90%) and test (10%) datasets to evaluate the performance of a U‐net model, utilizing an encoder–decoder architecture and contracting and expansive paths, developed as part of the study. Results Test results produced sensitivity and specificity values of 94.38% and 68.13%, respectively, in the diagnosis of Ascites‐1 US images, with an average Dice coefficient of 0.65 (standard deviation [SD] = 0.21). Similarly, the sensitivity and specificity for Ascites‐2 were 97.12% and 86.33%, respectively, with an average Dice coefficient of 0.79 (SD = 0.14). The accuracy and area under the curve (AUC) were 81.25% and 0.76 for Ascites‐1 and 91.73% and 0.91 for Ascites‐2. Conclusion The results produced by the U‐net demonstrate the viability of DL for automated ascites diagnosis. This suggests the proposed technique could be highly valuable for improving FAST‐based preliminary diagnoses, particularly in emergency scenarios.

show abstract

Section: Discussionmentioning

confidence: 81%