Development of a deep learning method for improving diagnostic accuracy for uterine sarcoma cases

A clinical ultrasound algorithm to identify uterine sarcoma and smooth muscle tumors of uncertain malignant potential in patients with myometrial lesions: the MYometrial Lesion UltrasouNd And mRi study

BACKGROUND

Differential diagnosis between benign uterine smooth muscle tumors and malignant counterpart is challenging.

OBJECTIVE

To evaluate the accuracy of a clinical and ultrasound based algorithm in predicting mesenchymal uterine malignancies, including smooth muscle tumors of uncertain malignant potential.

STUDY DESIGN

We report the 12-month follow-up of an observational, prospective, single-center study that included women with at least 1 myometrial lesion ≥3 cm on ultrasound examination. These patients were classified according to a 3-class diagnostic algorithm, using symptoms and ultrasound features. "White" patients underwent annual telephone follow-up for 2 years, "Green" patients underwent a clinical and ultrasound follow-up at 6, 12, and 24 months and "Orange" patients underwent surgery. We further developed a risk class system to stratify the malignancy risk.

RESULTS

Two thousand two hundred sixty-eight women were included and target lesion was classified as benign in 2158 (95.1%), as other malignancies in 58 (2.6%) an as mesenchymal uterine malignancies in 52 (2.3%) patients. At multivariable analysis, age (odds ratio 1.05 [95% confidence interval 1.03-1.07]), tumor diameter >8 cm (odds ratio 5.92 [95% confidence interval 2.87-12.24]), irregular margins (odds ratio 2.34 [95% confidence interval 1.09-4.98]), color score=4 (odds ratio 2.73 [95% confidence interval 1.28-5.82]), were identified as independent risk factors for malignancies, whereas acoustic shadow resulted in an independent protective factor (odds ratio 0.39 [95% confidence interval 0.19-0.82[). The model, which included age as a continuous variable and lesion diameter as a dichotomized variable (cut-off 81 mm), provided the best area under the curve (0.87 [95% confidence interval 0.82-0.91]). A risk class system was developed, and patients were classified as low-risk (predictive model value <0.39%: 0/606 malignancies, risk 0%), intermediate risk (predictive model value 0.40%-2.2%: 9/1093 malignancies, risk 0.8%), high risk (predictive model value ≥2.3%: 43/566 malignancies, risk 7.6%).

CONCLUSION

The preoperative 3-class diagnostic algorithm and risk class system can stratify women according to risk of malignancy. Our findings, if confirmed in a multicenter study, will permit differentiation between benign and mesenchymal uterine malignancies allowing a personalized clinical approach.

The automatic diagnosis artificial intelligence system for preoperative magnetic resonance imaging of uterine sarcoma

OBJECTIVE

Magnetic resonance imaging (MRI) is efficient for the diagnosis of preoperative uterine sarcoma; however, misdiagnoses may occur. In this study, we developed a new artificial intelligence (AI) system to overcome the limitations of requiring specialists to manually process datasets and a large amount of computer resources.

METHODS

The AI system comprises a tumor image filter, which extracts MRI slices containing tumors, and sarcoma evaluator, which diagnoses uterine sarcomas. We used 15 types of MRI patient sequences to train deep neural network (DNN) models used by tumor filter and sarcoma evaluator with 8 cross-validation sets. We implemented tumor filter and sarcoma evaluator using ensemble prediction technique with 9 DNN models. Ten tumor filters and sarcoma evaluator sets were developed to evaluate fluctuation accuracy. Finally, AutoDiag-AI was used to evaluate the new validation dataset, including 8 cases of sarcomas and 24 leiomyomas.

RESULTS

Tumor image filter and sarcoma evaluator accuracies were 92.68% and 90.50%, respectively. AutoDiag-AI with the original dataset accuracy was 89.32%, with 90.47% sensitivity and 88.95% specificity, whereas AutoDiag-AI with the new validation dataset accuracy was 92.44%, with 92.25% sensitivity and 92.50% specificity.

CONCLUSION

Our newly established AI system automatically extracts tumor sites from MRI images and diagnoses them as uterine sarcomas without human intervention. Its accuracy is comparable to that of a radiologist. With further validation, the system could be applied for diagnosis of other diseases. Further improvement of the system's accuracy may enable its clinical application in the future.

Diagnostic accuracy of MRI in the differential diagnosis between uterine leiomyomas and sarcomas: A systematic review and meta-analysis

BACKGROUND

Differential diagnosis between uterine leiomyomas and sarcomas is challenging. Magnetic resonance imaging (MRI) represents the second-line diagnostic method after ultrasound for the assessment of uterine masses.

OBJECTIVES

To assess the accuracy of MRI in the differential diagnosis between uterine leiomyomas and sarcomas.

SEARCH STRATEGY

A systematic review and meta-analysis was performed searching five electronic databases from their inception to June 2023.

SELECTION CRITERIA

All peer-reviewed observational or randomized clinical trials that reported an unbiased postoperative histologic diagnosis of uterine leiomyoma or uterine sarcoma, which also comprehended a preoperative MRI evaluation of the uterine mass.

DATA COLLECTION AND ANALYSIS

Sensitivity, specificity, positive and negative likelihood ratios, diagnostic odds ratio, and area under the curve on summary receiver operating characteristic of MRI in differentiating uterine leiomyomas and sarcomas were calculated as individual and pooled estimates, with 95% confidence intervals (CI).

RESULTS

Eight studies with 2495 women (2253 with uterine leiomyomas and 179 with uterine sarcomas), were included. MRI showed pooled sensitivity of 0.90 (95% CI 0.84-0.94), specificity of 0.96 (95% CI 0.96-0.97), positive likelihood ratio of 13.55 (95% CI 6.20-29.61), negative likelihood ratio of 0.08 (95% CI 0.02-0.32), diagnostic odds ratio of 175.13 (95% CI 46.53-659.09), and area under the curve of 0.9759.

CONCLUSIONS

MRI has a high diagnostic accuracy in the differential diagnosis between uterine leiomyomas and sarcomas.

Comparison of diagnostic performance of radiologist- and AI-based assessments of T2-FLAIR mismatch sign and quantitative assessment using synthetic MRI in the differential diagnosis between astrocytoma, IDH-mutant and oligodendroglioma, IDH-mutant and 1p/19q-codeleted

PURPOSE

This study aimed to compare assessments by radiologists, artificial intelligence (AI), and quantitative measurement using synthetic MRI (SyMRI) for differential diagnosis between astrocytoma, IDH-mutant and oligodendroglioma, and IDH-mutant and 1p/19q-codeleted and to identify the superior method.

METHODS

Thirty-three cases (men, 14; women, 19) comprising 19 astrocytomas and 14 oligodendrogliomas were evaluated. Four radiologists independently evaluated the presence of the T2-FLAIR mismatch sign. A 3D convolutional neural network (CNN) model was trained using 50 patients outside the test group (28 astrocytomas and 22 oligodendrogliomas) and transferred to evaluate the T2-FLAIR mismatch lesions in the test group. If the CNN labeled more than 50% of the T2-prolonged lesion area, the result was considered positive. The T1/T2-relaxation times and proton density (PD) derived from SyMRI were measured in both gliomas. Each quantitative parameter (T1, T2, and PD) was compared between gliomas using the Mann-Whitney U-test. Receiver-operating characteristic analysis was used to evaluate the diagnostic performance.

RESULTS

The mean sensitivity, specificity, and area under the curve (AUC) of radiologists vs. AI were 76.3% vs. 94.7%; 100% vs. 92.9%; and 0.880 vs. 0.938, respectively. The two types of diffuse gliomas could be differentiated using a cutoff value of 2290/128 ms for a combined 90th percentile of T1 and 10th percentile of T2 relaxation times with 94.4/100% sensitivity/specificity with an AUC of 0.981.

CONCLUSION

Compared to the radiologists' assessment using the T2-FLAIR mismatch sign, the AI and the SyMRI assessments increased both sensitivity and objectivity, resulting in improved diagnostic performance in differentiating gliomas.

Preoperative Imaging Evaluation of Endometrial Cancer in FIGO 2023

The staging of endometrial cancer is based on the International Federation of Gynecology and Obstetrics (FIGO) staging system according to the examination of surgical specimens, and has revised in 2023, 14 years after its last revision in 2009. Molecular and histological classification has incorporated to new FIGO system reflecting the biological behavior and prognosis of endometrial cancer. Nonetheless, the basic role of imaging modalities including ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography, as a preoperative assessment of the tumor extension and also the evaluation points in CT and MRI imaging are not changed, other than several point of local tumor extension. In the field of radiology, it has also undergone remarkable advancement through the rapid progress of computational technology. The application of deep learning reconstruction techniques contributes the benefits of shorter acquisition time or higher quality. Radiomics, which extract various quantitative features from the images, is also expected to have the potential for the quantitative prediction of risk factors such as histological types and lymphovascular space invasion, which is newly included in the new FIGO system. This article reviews the preoperative imaging diagnosis in new FIGO system and recent advances in imaging analysis and their clinical contributions in endometrial cancer. EVIDENCE LEVEL: 4 TECHNICAL EFFICACY: Stage 3.

A human-interpretable machine learning pipeline based on ultrasound to support leiomyosarcoma diagnosis

The preoperative evaluation of myometrial tumors is essential to avoid delayed treatment and to establish the appropriate surgical approach. Specifically, the differential diagnosis of leiomyosarcoma (LMS) is particularly challenging due to the overlapping of clinical, laboratory and ultrasound features between fibroids and LMS. In this work, we present a human-interpretable machine learning (ML) pipeline to support the preoperative differential diagnosis of LMS from leiomyomas, based on both clinical data and gynecological ultrasound assessment of 68 patients (8 with LMS diagnosis). The pipeline provides the following novel contributions: (i) end-users have been involved both in the definition of the ML tasks and in the evaluation of the overall approach; (ii) clinical specialists get a full understanding of both the decision-making mechanisms of the ML algorithms and the impact of the features on each automatic decision. Moreover, the proposed pipeline addresses some of the problems concerning both the imbalance of the two classes by analyzing and selecting the best combination of the synthetic oversampling strategy of the minority class and the classification algorithm among different choices, and the explainability of the features at global and local levels. The results show very high performance of the best strategy (AUC = 0.99, F1 = 0.87) and the strong and stable impact of two ultrasound-based features (i.e., tumor borders and consistency of the lesions). Furthermore, the SHAP algorithm was exploited to quantify the impact of the features at the local level and a specific module was developed to provide a template-based natural language (NL) translation of the explanations for enhancing their interpretability and fostering the use of ML in the clinical setting.

Clinical features of uterine sarcomas presenting mainly with uterine masses: a retrospective study

BACKGROUND

Uterine sarcomas are uncommon mesenchymal tumors of the uterus. The clinical problem is that the features of uterine sarcomas can sometimes mimic uterine fibroids. This study aims to investigate the clinical characteristics of patients with uterine sarcomas who were preoperative presenting mainly with uterine masses.

METHODS

A retrospective analysis of patients who underwent gynecological surgery for uterine sarcomas at the Obstetrics & Gynecology Hospital of Fudan University, between January 2016 and December 2021.

RESULTS

Over the 5-year period, 277 patients were final diagnosed of uterine sarcomas. A total of 162 patients were preoperatively diagnosed as uterine fibroids for surgical treatment, the majority of whom were diagnosed of uterine leiomyosarcoma (uLMS) (49/162) and low-grade endometrial stromal sarcoma (LG-ESS) (100/162). Ninety people underwent total hysterectomy and bilateral salpingo-oophorectomy (TH + BSO), while 72 underwent myomectomy followed by supplemental TH + BSO. The group with direct hysterectomy had a higher average age than the group with prior myomectomy (47.20 ± 8.94 vs. 40.86 ± 5.88, p < 0.001). Among patients preoperatively diagnosed as uterine fibroids, patients with uLMS had a higher proportion of previous myomectomy (26.53% vs. 5.00%, p < 0.001), a larger uterine mass diameter on ultrasound (8.38 ± 3.39 cm vs. 6.41 ± 1.92 cm, p < 0.001), and richer hypervascularity (34.69% vs. 18%, p = 0.024) compared with LG-ESS.

CONCLUSIONS

Analysis of our data showed that a large proportion of uterine sarcomas, especially uLMS and LG-ESS, present mainly with uterine masses. Ultrasound features including a large uterine mass diameter and rich hypervascularity, and with a history of myomectomy may alert clinicians in suspicion of uLMS when compared with LG-ESS.

Large Cervical Leiomyoma of the Uterus: A Rare Cause of Chronic Pelvic Pain Associated With Obstructive Uropathy and Renal Dysfunction: A Case Report

Large cervical leiomyomas (≥10cm) are extremely rare. Our case report concerns the surgical treatment of a patient with a large cervical leiomyoma associated with chronic pelvic pain, bilateral hydroureteronephrosis and significant impairment of renal function. A 47-year-old patient of reproductive age with a normal menstrual cycle and a medical history of chronic pelvic pain presented to the gynecology clinic for examination. Clinically, the presence of a large pelvic mass was found, the upper margins of which were palpable at the level of the umbilicus. A preoperative assessment revealed bilateral hydroureteronephrosis due to obstructive uropathy and renal dysfunction. Hydroureteronephrosis, as a consequence of the large pelvic mass, probably originating from the cervix of the uterus, was evaluated as the main cause of renal dysfunction. Tumor markers were negative. The imaging studies confirmed the clinical diagnosis of uterine leiomyoma, and the surgical treatment of the patient with laparotomy was decided. Intraoperatively, the presence of a large uterine cervical fibroid was detected, and a total abdominal hysterectomy and bilateral adnexectomy were performed. Operating was difficult, with significant surgical difficulties. The postoperative course was uneventful, without immediate complications. The patient's symptom relief began gradually, immediately after surgery. Three months after surgery, the patient reported complete relief of her pelvic pain. A re-examination of the urinary tract revealed complete recovery of renal morphology and function. In the paper, after the presentation of the case, a brief review of cervical leiomyomas is attempted based on the literature, mainly regarding the diagnostic and therapeutic approach.