Deep Learning Prediction of Ovarian Malignancy at US Compared with O-RADS and Expert Assessment

• Hui Chen*, 

• Bo-Wen Yang*, 

• Le Qian, 

• Yi-Shuang Meng, 

• Xiang-Hui Bai, 

• Xiao-Wei Hong, 

• Xin He, 

• Mei-Jiao Jiang, 

• Fei Yuan, 

• Qin-Wen Du, 

• Wei-Wei Feng 

* H.C. and B.W.Y. contributed equally to this work.

Published Online:Apr 12 2022

A deep learning model that used feature fusion to classify benign and malignant ovarian tumors on gray scale and color Doppler US images achieved a similar performance to clinical expert assessment.

Background

Deep learning (DL) algorithms could improve the classification of ovarian tumors assessed with multimodal US.

Purpose

To develop DL algorithms for the automated classification of benign versus malignant ovarian tumors assessed with US and to compare algorithm performance to Ovarian-Adnexal Reporting and Data System (O-RADS) and subjective expert assessment for malignancy.

Materials and Methods

This retrospective study included consecutive women with ovarian tumors undergoing gray scale and color Doppler US from January 2019 to November 2019. Histopathologic analysis was the reference standard. The data set was divided into training (70%), validation (10%), and test (20%) sets. Algorithms modified from residual network (ResNet) with two fusion strategies (feature fusion [hereafter, DLfeature] or decision fusion [hereafter, DLdecision]) were developed. DL prediction of malignancy was compared with O-RADS risk categorization and expert assessment by area under the receiver operating characteristic curve (AUC) analysis in the test set.

Results

A total of 422 women (mean age, 46.4 years ± 14.8 [SD]) with 304 benign and 118 malignant tumors were included; there were 337 women in the training and validation data set and 85 women in the test data set. DLfeature had an AUC of 0.93 (95% CI: 0.85, 0.97) for classifying malignant from benign ovarian tumors, comparable with O-RADS (AUC, 0.92; 95% CI: 0.85, 0.97; P = .88) and expert assessment (AUC, 0.97; 95% CI: 0.91, 0.99; P = .07), and similar to DLdecision (AUC, 0.90; 95% CI: 0.82, 0.96; P = .29). DLdecision, DLfeature, O-RADS, and expert assessment achieved sensitivities of 92%, 92%, 92%, and 96%, respectively, and specificities of 80%, 85%, 89%, and 87%, respectively, for malignancy.

Conclusion

Deep learning algorithms developed by using multimodal US images may distinguish malignant from benign ovarian tumors with diagnostic performance comparable to expert subjective and Ovarian-Adnexal Reporting and Data System assessment.

© RSNA, 2022

U S offers alternative to MRI for bone stress

 

By Kate Madden Yee, AuntMinnie.com staff writer

March 24, 2022 -- Ultrasound appears to be an effective point-of-care alternative to MRI for the evaluation of bone stress injury, according to a study published March 22 in the Journal of Ultrasound in Medicine.

MRI is typically used to assess bone injuries, but the study results suggest that clinicians may have a more easily accessible tool for this application, wrote a team led by Dr. Isaac Syrop of Columbia University in New York City and Dr. Yaeko Fukushima, PhD, of Kansai Medical University in Osaka, Japan.

"There are many advantages to ultrasound imaging over MRI, including its dynamic practicality, which provides the treating clinician with an opportunity to evaluate local soft tissue sites in real-time," the group wrote. "Ultrasound imaging takes significantly less time than MRI to perform ... and can be done as part of the clinical evaluation."

Bone stress injuries are common in young adult athletes, and can negatively affect performance over the long term, the investigators noted. It's crucial to diagnose these injuries early so that athletes can be treated effectively and return to play.

X-ray and CT aren't very effective for early diagnosis of bone stress injuries and expose patients to radiation; since the 1980s, MRI has been used as the diagnostic standard. But MRI is expensive and time-consuming.

"There is good reason to believe that the more affordable and accessible diagnostic option of musculoskeletal ultrasound imaging may help to address the shortcomings of MRI," the authors noted.

The researchers sought to compare the sensitivity and specificity of ultrasound to MRI in the diagnosis of bone stress injury through a study that included 37 young adult athletes presenting in an academic sports medicine clinic between 2016 and 2020 with suspected lower extremity bone stress injury. Sports included everything from crew, field hockey, gymnastics, and running to soccer, tennis, track, and volleyball. All patients underwent MRI and ultrasound exams.

Of the 37 study participants, 30 (81%) had bone stress injuries. The most common injuries were in the metatarsal bones (54%) and the tibia (32%).

Ultrasound scored relatively high across a range of performance measures relative to the gold standard of diagnosis on MRI scans, the group found.

Performance of ultrasound for diagnosing bone stress injuries compared with baseline assessment MRI
Measure	Performance
Sensitivity	80%
Specificity	71%
Positive predictive value	92%
Negative predictive value	45%

"In summary, ultrasound imaging may be a point-of-care tool for the current practicing sports medicine provider to combine with their clinical evaluation in the diagnosis of bone stress injuries of the lower extremity," the team concluded.